Experiment ledger

Every idea and test, its stage and why. idea 103 · killed 491 · paper 48 · parked 97 · replay 15 · verified 2. Updated 23 Oct 02:30 Bangkok.

idideastagefamilyownertradesP&L $10verdict / reasonupdated
#43Grid: live mc5 rule replay with variants
verifiedv3-mc5/replay-gridtestmgr158VERIFIED as fill replay, NOT as profit forecast: rests on ~15 runners12 Oct 16:20
#15-Min Cap live rule (v3 mc5, cap 250k-1M at +5 min)
Buy v3 mc5 coins with cap 250k-1M at first swap +300 s, stop 0.5x, 4 h; profit comes from rare 10x+ runners		verifiedv3-mc5/5min-capmanager158+2860Replay matches 29 live fills (auditor 9 Oct); profit rests on ~15 runners; live since 6 Oct at $10 (Ben)10 Oct 08:39
#756Cap-continuation entry 6-10M ULTRAWIDE stop .4 trail 60% (v3), hunt 20M+
papermisstestmgr3replay 1.59x n30; cron pending ops22 Oct 08:30
#755Cap-continuation entry 150-300k ULTRAWIDE stop .4 trail 60% (v3), hunt 3M+
papermisstestmgr3replay 0.92x n303; cron pending ops22 Oct 08:30
#738Cap-continuation entry 600k-1M ultrawide stop .4 trail 60% (v3), hunt 3M+
papermisstestmgr3mature22 Oct 02:28
#737Cap-continuation entry 3-6M ultrawide stop .4 trail 60% (v3), hunt 10M+
papermisstestmgr3mature22 Oct 02:28
#734Cap-continuation 1.5-3M ultrawide stop .4 trail 60%, hunt 3M+
papermisstestmgr3mature21 Oct 19:30
#733Cap-continuation 300-600k ultrawide tighter stop .4 trail 60%, hunt 3M+
papermisstestmgr3mature21 Oct 19:30
#732Cap-continuation entry 3-6M ULTRAWIDE stop .5 trail 70% (v3), hunt 10M+
papermisstestmgr3paper v3capw cron */10, decide 25 Oct21 Oct 12:35
#731Cap-continuation entry 300-600k ULTRAWIDE stop .5 trail 70% (v3), hunt 3M+
papermisstestmgr3paper v3capz cron */10, decide 25 Oct21 Oct 12:35
#730Cap-continuation entry 1.5-3M ULTRAWIDE stop .5 trail 70% (v3), hunt 5-20M
papermisstestmgr3paper v3capy cron */10, decide 25 Oct21 Oct 12:35
#729Cap-continuation 300-600k VERY FRESH age<=60m (v3), hunt 1-3M
floor #3633 variant of v3cny engine, test v3cx3		papermisstestmgr4mature21 Oct 12:29
#728Cap-continuation 1-1.5M FRESH WIDE (v3), hunt 3-5M
floor #3633 variant of v3cny engine, test v3cx1		papermisstestmgr4mature21 Oct 12:29
#727Cap-continuation 600k-1M FRESH WIDE (v3), hunt 3M+
floor #3633 variant of v3cny engine, test v3cx6		papermisstestmgr4mature21 Oct 12:29
#721Cap-continuation entry 600k-1M ULTRAWIDE stop .5 trail 70% (v3), hunt 3M+
replay n190 1.16x med .46x		papermisstestmgr3mature21 Oct 06:29
#720Cap-continuation entry 1-1.5M WIDE trail 35% (v3), stop 30%, hunt 3-5M
replay n126 1.06x med .65x		papermisstestmgr3mature21 Oct 06:29
#703Cap-continuation entry 150-300k WIDE trail 35% (v3), stop 30%, hunt 3M+
papermisstestmgr3mature20 Oct 10:36
#702Cap-continuation entry 300-600k WIDE trail 35% (v3), stop 30%, hunt 3M+
papermisstestmgr3mature20 Oct 10:36
#701Cap-continuation entry 150-300k (v3), trail 50% stop 30%, hunt 1-3M
papermisstestmgr3mature20 Oct 10:36
#706Cap-continuation 1.5-6M FRESH + WIDE exit, hunt 5-20M
combines winning arms of v3capw/v3capy; v3cwy, engine v3capf with lo-reach prefilter (no 1M look-ahead)		papermisstestmgr4registered, cron */10, testmgr4 #348920 Oct 10:28
#705Cap-continuation 300-600k FRESH only (age<=180m), trail 50% stop 30%, hunt 1-3M
combines winning arms of v3capw/v3capy; v3cny, engine v3capf with lo-reach prefilter (no 1M look-ahead)		papermisstestmgr4registered, cron */10, testmgr4 #348920 Oct 10:28
#704Cap-continuation 300-600k WIDE exit (stop .5, trail 70%), hunt 1-3M
combines winning arms of v3capw/v3capy; v3cnw, engine v3capf with lo-reach prefilter (no 1M look-ahead)		papermisstestmgr4registered, cron */10, testmgr4 #348920 Oct 10:28
#651Kit ramp is deterministic (70% reach 100k, 55% 200k from 32k): early entry + exit at 150-200k? needs swap tape for the 1100 kit pools (r6nat4 ev empty)
early entry cap 40-60k, exit 150-200k catches ramp with real depth; needs tape		paperunhooked-v4/blind-spotanalyst2mature18 Oct 08:01
#63916 Oct misses: 4 hooked v4 runners (tokq hook x3 cap_first 49-55k, usdg hook 2.9M; 1.0-1.4M ATH) seen by radar as 'hooked' but mc5 MISSED; 15 Oct had AGEM+WIN same kit. Test: v4 tokq-hook pool entry at 50-60k first print, hunt 1M tier; needs base rate of non-runners with same hook/cap_first (blocked on v4tape >25min)
papermissanalyst2running18 Oct 06:20
#541mc5 buy only at entry cap >=600k (4/22 ran vs 0/41 below; n=63 live)
Entry cap proxies for confirmed momentum; size did not matter. Needs shadow/holdout on 11-13 Oct (shadow n=11 mean 0.43).		paperv3-mc5/5min-capanalyst2paper mc5hi600 (/root/tm/mc5hi600) from 17 Oct, derived from mc5 mirror; all-time HI n42 2.38x, since 16 Oct HI n8 0.26x; kill fwd n20 <1.018 Oct 05:25
#650DEADKIT32: native nohook pools born cap 32-34k (1100 since 14 Oct) ramp deterministically to ~300k in 5-15 min then stop: 0/173 reach 1M. Skip filter for every native 300k chaser: hit rate 20% -> 38% (halves 40/36%)
Chasing the 300k touch on these buys the farm's exit; ex-kit native 300k touchers P1M 38% n=188		paperunhooked-v4/blind-spotanalyst2mature18 Oct 03:16
#632SPARSE60: v3 +60s entry only if >=4 swaps by 60s and no two buys in the same second
Lens: order-flow cadence (bot-batch vs organic). v3hist 6-16 Oct, entry first print >=68s after birth (8s latency, +7% slip), cap60 3k-150k, tp5 sl.5 2h, LP pulls (>=50% share removed after entry) for		paperv3-mc5/first-minute-filtersanalyst3mature17 Oct 03:06
#624MOM60: v3live entry at +60s only if cap grew >=1.2x between +30s and +60s (cap<150k), tp5 stop.5
mom>=1.2 n=136 mean 3.75 (base 1.74), halves 3.23/3.88, 9 runners; script m3.py; no slippage/rug model, base 1.74		papermissanalyst1136mature17 Oct 01:37
#623v3 net-inflow early F (age<=300 cap>=10k net>=1.5ETH)
Replay v3live 6-16 Oct n=716 mean=1.74, 15s latency, 7% cost, trail50. EV: ~20/day x (1.74-1) x $10 well above 0.005 ETH/day. Sell sim pending.		papermissanalyst2716mature17 Oct 01:37
#617v3 very-early cap-cross 10k within 300s (entry cap<30k, pool ETH>=1, trail 50%): missed v3 organic runners enter at birth caps 2.5-10k; replay 6-15 Oct
Enter first swap >=10k cap at age<=300s; exit 0.6x stop or 50% trail; hunt 1M+		papermissanalyst2384mature16 Oct 16:43
#44USDG/SUSD-quoted v4 pools, entry +66s tp5
biggest missed runners are USDG/SUSD v4		paperstable-v4/baselineanalyst25466baseline 1.35x, 14% reach 5x, price from tick only (no depth/sell sim); no pre-entry rule separates sub-groups; hooked worse 1.1716 Oct 10:23
#106NHW-S6: v4 nohook WETH enter first 300k touch with >=6 distinct senders, tp5 sl.5 6h
95% of nohook WETH pools are 1-5 wallet scripts; the ones with >=6 senders at 300k are organic and run. Kill: OOS mean<1.2 over 30 triggers		paperunhooked-v4/first-minute-filtersanalyst165sent to testmgr16 Oct 10:23
#506A3-V2Q-S3: v2 0x8bce trigger only if >=3 sell-like Syncs (r0 down, r1 up) before trigger; dead-adjusted mean 4.08 n16 (halves 3.85/4.32, zeros 0) vs 2.54 n32 unguarded
Same trigger/entry/exit as #501 (m>=8, reserve0>=2 WETH, enter 2nd Sync after, tp5 sl.5 LP-pull exit) plus guard sells_pre>=3 counted from pair Syncs before the trigger. ppfeeds sell probe #2028 shows		paperunknown-venue/v2-factoryanalyst316mature14 Oct 03:42
#563Run 106: v2 0x8bce forward n=18: sells_pre>=3 n=3 (2 hit 5x, 1 pulled in 7 min); winners keep half reserve 43 min / 3.2 h after 5x
A3-V2Q-S3 gate		paperunknown-venue/v2-factoryanalyst3info14 Oct 03:42
#535Run 123: H3000 WETH-quote 1M-cross chaser (cap_first<150k, depth>=20k at cross): tp2 sl.5-.6 n19 mean 1.58-1.64 (halves 1.86/1.33), ~5/day; WETH sellability unprobed; second leg to #483
paperhooked-v4/short-tpanalyst1mature13 Oct 07:20
#249Create-tx value (ETH sent at launch) as t=0 runner signal: >=12.5 ETH 1M 58-100%, 3M 36-57%; first-time sender 73%/43%/24% (n=74); born ~690k cap; replay t+60 tp10 sl.5 1.37 (0.88/1.84) n=71
paperv3-mc5/launch-metadataanalyst2mature11 Oct 15:27
#98mc5 band: top single buyer share of first-5-min buy volume <0.11 -> tp5 1.58 n=146 vs 0.99 n=74 (>=7 buyers)
paperv3-mc5/5min-capanalyst3mature11 Oct 15:27
#91mc5 band residual: swaps before +5min (n_pre) after dust guard
Skip band pools with <=10 swaps by +5min even after the dust guard passes		paperv3-mc5/5min-capanalyst2330mature11 Oct 15:27
#220All-caps symbol on v3 mc5 band: upper+vol5>=49k tp5sl.5 1.35 n=151 (halves 1.56/1.18) vs mixed 0.95
mixed-case symbols are joke copies that do not run		paperv3-mc5/first-minute-filtersanalyst3151mature11 Oct 15:27
#496TOP1-MC5: largest-buyer share <0.11 in first 300 s on mc5 band (read-only log)
paperv3-mc5analyst2mature11 Oct 15:26
#392NET-FLOW proxy of INFLOW: net buy USD (buys-sells) in +60..+305s >0 on v3hist band 250k-1M, 28 days (Sep12-Oct10), entry +305s tp5 sl.5 4h
net buying 1-5 min proxies reserve inflow; extends #387 beyond Oct 6 (reserve field exists only from Oct 6)		paperv3-mc5/first-minute-filtersanalyst2151mature11 Oct 14:26
#387INFLOW: pool reserve at +300s > reserve at +60s (net buying 1-5 min) on v3 band 250k-1M: tp5 sl.5 1.80 n23 (6 fivers) vs 0.79 n71; halves 1.83/1.68; same sign at 30-100k, 100-250k, 1-3M caps; perm p .0005 (3 features tried); 4 days data, no OOS, overlap with BREADTH untested
Band entries after net reserve inflow between 60s and 300s beat the rest; wrong if forward n>=30 shows <=1.0		paperv3-mc5/first-minute-filtersanalyst2mature11 Oct 14:26
#333A3-COPY OOS on token_born names (to 9 Oct): band copy 5x 30% vs 18% (late half 33 vs 15, n=36); gap since last runner of symbol <6h 38% vs >48h 25%; copies on v4 WETH/native nohook n=12/6 flat, hooked copy worse
paperv3-mc5/first-minute-filtersanalyst2mature11 Oct 14:26
#34MC5 5-min cap rule (cap 250k-1M at +5min)
paperv3-mc5/5min-capanalyst1see #1 (live)11 Oct 14:25
#42P204 V3 new pools nightly replay 5x
paperv3-new-pools/tp5testmgr22 closed at audit, nothing to audit10 Oct 08:40
#41ENG2 older coins Solana/BSC/Base 4h
paperother-chains/eng2analyst100 closed trades at audit10 Oct 08:40
#40H3N native unique-hook sell at 10 min
paperhooked-v4/tp-ladderanalyst1running10 Oct 08:40
#36A1mc late buyers / no repeat deployer arms
paperv3-mc5/sizing-deployeranalyst1running, LATEB vs flat after 60 trades10 Oct 08:40
#10HOOK11689 + high init-price template, tp20 hold 40min
paperhooked-v4/short-tpanalyst1152hi-init 3.4x vs 2.2; decay 3.8->2.8 in thirds, not significant; needs OOS10 Oct 08:40
#9QUIET60 refined (<=17 swaps, >=2 sells) -> merged into hooked filter
paperhooked-v4/short-tpanalyst1302322/343 cohort is hooked; merged as hook filter, 2.2-2.510 Oct 08:40
#594v3 early ramp entry +100s (SEALS), tp3/sl.5
enter at +100s on buyer growth, ride to runner		replayfirst-minute-filtersanalyst2killed15 Oct 09:09
#214Counterfactual audit of live mc5 decisions: runners lost to 429/confirm errors
49 band pools, 47 bought; the 2 missed were 72x and 22x runners lost to HTTP 429 and swap-confirmation errors (exec #593)		replayv3-mc5/5min-capanalyst249parked11 Oct 15:27
#374Insider re-buy share at 1M cross (buys 300k-1M from wallets that bought <300k, ladder_from tx.from): >=24% reaches 2M 28% vs 58% (n179, halves 38/68, 18/48)
replayv3-mc5/late-runneranalyst1179parked11 Oct 14:26
#397Round-number buy share in first 300s (human preset sizes) on v3 band n262: terciles 0.95/1.43/1.74, >=0.10 1.67 n119 vs 1.12 p .001, not a buy-count proxy, 3 of 4 quarters; mature, needs forward + v4 nohook amounts
replayv3-mc5/first-minute-filtersanalyst2parked11 Oct 14:26
#407Campaign heat: cell trailing 24h fast-1M rate (3h embargo): P5M cold/mid/hot v3 .1/.8/1.2%, hooked WETH .3/.5/1.9%; inside v3 band no lift (1.34/1.27/1.18); hooked WETH band hot 2.69 n44 vs cold 1.33 n7
heat gate for h3k		replayhooked-v4/first-minute-filtersanalyst2parked11 Oct 14:26
#422Token name clone-flood (prior same-name tokens as-of birth) vs P1M on tokmeta+radar crosses
replayunhooked-v4/blind-spotanalyst2103parked11 Oct 14:26
#446Creator sells own token in first 300s (tx.from=creator_from, first5.db) on v3 band cap5m 250k-1M: P1M 56% n50 vs 35% n279, coarse 3x/stop.5 sim 1.85 (sell>=$1000, n43) vs 1.33; halves 54/64% vs 33/37%; fee100 only; distinct one-pool creators
Creator selling into demand = real demand, not an operator pump		replayv3-mc5/walletsanalyst2parked11 Oct 14:26
#383Request ppdata: pool_births table (all venues, true create time) from 13 Sep, to retest quiet-chain regime gate
replayscout/regimeanalyst2info11 Oct 14:25
#349Permutation test of our 7 band flags: SEED/VOL5/SERIAL/BORNHI survive (p .001, hold last 40%); CAPS/COPY are noise
replayv3-mc5/first-minute-filtersanalyst3info11 Oct 14:25
#275Late-ladder entry v3 at 2M/3M/5M cross: 3M tp3 sl.5 combined 1.24 n=81 (quarters .79/.98/1.59/1.58), bootstrap 5% 0.98, win 25%
replayv3-mc5/late-runneranalyst2replay10 Oct 17:25
#252Band 5-min volume heat: vol5>175k 0.64 (0.65/0.62) n=52 vs 1.16; skip flag; band edge decays by quarter 1.64/.94/.77/.51
replayv3-mc5/first-minute-filtersanalyst2mature10 Oct 16:46
#55Skip/flag Pons V2 launches with snipe-tax-exempt first-second bundles and crew funding links
Exempt-wallet bundles + shared funder chain mark rug crews (53-56 launches, 467 other groups per Bitquery); filter lifts runner precision, or reveals decoy pumps rideable		replayscout/pons-v2-insider-bundlescoutdataeng ponsexempt.db: >=10 first-second exempt grad 1.87% vs 0.64% none (~3x); runner tiers not yet tested, analyst3 tasked10 Oct 11:10
#65Launch fee as metadata: USDG fee>=100k 0 real runners of 7794
skip fee 50k-2M at birth		replaystable-v4/first-minute-filtersanalyst1folded into #6010 Oct 10:26
#60USDG-EARLY-B: USDG v4 fee 1-50k, first 4 swaps all buys, enter swap 4, stop .5 trail .6@2x
all-buy start shows real demand; sim n=259 mean 2.21 halves 2.19/2.23 w/o top5 1.51; needs sell check		replaystable-v4/first-minute-filtersanalyst1sent to testmgr #46710 Oct 10:26
#45Hook storage slot1 at creation as filter
slot1 1e12/1e13 n=106 tp10 4.19 vs 2.09 rest; look-ahead trap if read latest		replayhooked-v4/short-tpanalyst1forward test needed10 Oct 09:10
#75221 Oct miss review (2nd view): 25 runners, 19 v4:native organic born cap ~9.8k depth 30-80k (Kimchi x6, FINLEY x3, CUMROCKET x2 serial symbols), rest v3 kit:B/other + 2 hooked; all = known seed-10k/serial-sym families (#663 #662 #684 #650 #751 killed/paper); no new shared feature, no new test
ideamissanalyst222 Oct 08:21
#725v4:native exit-side wide trail on aipaper5-scored subset: PARKED (exits already killed: #435 stress trails/ladders worse than flat tp3/s.5, #687 wide exit killed; scored subset too small, 63/80 never scored)
ideamissanalyst221 Oct 09:05
#723MISS20-AIPAPER-LINE: 20 Oct, ~12 v4:native organic runners (Dacity/Laura/Pokemon/KOLSHI 1.1-4.4M) scored 0.53-0.70 by aipaper, just below line; test lowering line for v4:native organic only, with pull sim (replay needed, kit pools pulled ~99.8% so likely killed like #698)
ideamissanalyst121 Oct 08:20
#722MISS7D-CLUSTERS 13-19 Oct: 208 real 1M+ runners, all in already-killed clusters (v3 organic 77 #605/#640/#592, v4 native 58+12 fee100 #561/#577/#529, v3 thin 18, hooked 14 #510, jump 13 #674); no new fingerprint
Re-clustered misses.csv by radar venue/class; every cluster maps to a numbered killed idea; no replay with mean>=1.1 possible without repeating them		ideamissanalyst121 Oct 06:31
#717DEVBUY-REQ: deployer own first-buy share and deployer-swap field per pool (data request to ppdata)
Deployer buying >X% of supply in own pool block vs P1M/rug in 60s. Needs per-pool deployer swap list from v4tape swaps.jsonl joined on tx_from=deployer; not in live_paths.		ideaunhooked-v4/blind-spotanalyst121 Oct 04:02
#697STRATEGY: runner in stock-token-quoted pool
Pools quoted in tokenized stocks (MSTR) can run to $7M; check if our feeds watch non-WETH quote pools		ideascout / stock-quote-runnerscout20 Oct 07:01
#679Venue base rates: Pons grad 1/16, Flap 0.9% reach 3 traders
Per-venue first-hour trader-count gate filters Flap noise; entry filters must beat 1/16 Pons base rate		ideascout / venue-base-ratesscout19 Oct 07:01
#678V3VE-PULLPOL: live pull policy for v3vearly: base-rate stop (>=17 pulls of last 20) + 15-stake drawdown hold instead of rug-count stop
58-64% pulls are base rate; zero-streak median 6/p95 11 in 60 trades so 3-rug stop trips daily		ideav3-new-poolsanalyst119 Oct 06:21
#630Data request: v3 LP-remove (Burn/Decrease) events per pool to separate LP pull from seller drain
rug-aware bound 1.42 vs naive 1.74 differs by 0.3; true LP pulls resolve it		ideamissanalyst217 Oct 02:19
#610NHN6: #514 rule (300k touch, >=6 wallets) on v4 nohook NATIVE pools
ideaunhooked-v4/first-minute-filterstestmgr416 Oct 10:23
#603Same-name decoy tokens: did we buy the copy of a runner?
Missed runners had earlier same-name decoys; bots bought decoys		ideascout / decoy-clonesscout16 Oct 07:01
#591Only 1 runner born 14 Oct (quiet day): aipaper5 never scored it (NOT SEEN) while aipaper/30 scored 0.58/0.66 below line; check whether a fast-ramp bypass of the score line (cap>1M by +300s, born<15k) beats threshold; replay needs v4tape ext (16 Oct 21:00 UTC)
ideamissanalyst215 Oct 08:20
#589Flap Stocks Vault tokens: graduation as runner signal
Vault tokens (S&P500) graduating Flap curve run more often than other Flap tokens		ideascout / venuescout15 Oct 07:01
#588MadeOnSol deployer tier + KOL first-touch as runner/rug features
Deployer reputation and multi-KOL first-touch separate runners/rugs before $1M		ideascout / datascout15 Oct 07:01
#586Run 15 Oct: v3hist.swaps pool_eth_after is empty for 1-5 Oct (2.5M rows) -> depth-to-exit untestable for 40% of history; ask dataeng backfill from reserves
ideadataanalyst2needs-data15 Oct 02:55
#57913 Oct misses: 10/27 are the native fee100 nohook seed-25k kit; blocked on tape window (v4tape 25 min); ask ppfeeds 90 min for kit pools, then replay #561 with loser leg
ideamissanalyst114 Oct 08:20
#574Slow-start v3/v2/v4 runners (ppfeeds #2398): consumer-side gap; v3 late 1M entry already killed #265, v3 organic misses #499 discard; no new test
Feeds are fine; aipick drops v4 pools unpicked after 1h and never scores v3/v2. Prior replays (#265 late v3 1M, oos 2.32 but killed; #348 breadth) cover it; a v3 scorer would need new evidence.		ideamissanalyst114 Oct 07:25
#573Idle run 148: signer.db 887,732 tx (+10k), 245k blocks still needed, fetcher HTTP 500; origin-EOA test still waits; honest stop
ideaunhooked-v4/blind-spotanalyst114 Oct 07:15
#571Run 14 Oct: buyer-overhang lens on v3hist band n=291: entry cap / 300s eth-weighted VWAP <=1.12 (price barely above avg buyer cost) mean 0.98 (P1M 44%) vs >1.12 1.49 (P1M 54%), halves 0.84/1.10 vs 1.68/1.30; share of buy-eth bought <0.7x entry non-monotone (Q3 best). Weak, partly overlaps FADE/momentum (#363); mature only as a 2nd skip flag inside FADE forward arm, no new test
ideav3-mc5/first-minute-filtersanalyst214 Oct 05:28
#570Run 14 Oct: name/symbol shape on v3hist band n=291 (non-ASCII, digits, multiword, symbol length, name==symbol, fee, creator, mint-to-first-trade lag): all flat/non-monotone, halves flip; unreadable-symbol pools n=32 mean 0.65 vs 1.30 (coverage artifact, P1M same) -> discard
ideav3-mc5/launch-metadataanalyst214 Oct 05:28
#559Run 16 Oct L5: deployer-funder cohort fingerprint at t=0 (deployer funded by same wallet as smart early buyers)
Fleet funder of the deployer is visible before first trade and predicts runner		ideawalletsanalyst2needs-data14 Oct 03:20
#551Run 133: LP-remove in first 300s (lp_remove.db x live_paths, cap_5m 100k-1.5M) as runner/rug flag: nohook 32 pools with early remove, 6 hit 1M (18.8%) vs 7.8% without; hooked n4. Early remove is not a warning, n tiny, adds no filter: discard
ideaunhooked-v4/blind-spotanalyst114 Oct 01:55
#532Run 121: deployer pull repeat (lpowner tx_from x lpratio alarms, 12259 pools): prior-pull deployers re-pull 6/10 vs 5.8% base; alarms.jsonl is a 500-row window, n too small, not decidable
ideaunhooked-v4/blind-spotanalyst213 Oct 06:32
#531Run 120: new roundtrip.jsonl feed (rtd, 393 pools, 774 probes): A1 native nohook 415/415 probes roundtrip 0.9998 (sellable at 5 s and 60 s age); H3000 native hooked 26/45 (58%) rt 0.0001 = unsellable. Confirms v4chase (#525) has no honeypot risk, H3000 does; not an edge, supports sellability gate. Discard as new entry.
ideaunhooked-v4/blind-spotanalyst113 Oct 06:10
#530Run 122 microstructure lens (Amihud price impact, buy-size gini, round-size share, inter-arrival CV, net flow) on v3hist band pools (cap 250k-1M at +300s, n=286): only net-flow share >0.327 AND median impact/ETH <=0.153 separates: P1M 66% vs 45% (n65/221, early/late halves 66/66 vs 44/46), tp5 sl.5 slip7 replay 1.46 vs 1.18. Cuts chosen post hoc on same data, overlaps #502 sell-share; gini/round/CV/arrival flat. Replay-to-forward shrink per #522 => mature only, not test-ready
Low price impact per ETH plus buy-dominated first 300s marks real depth, not scripted crowd		ideav3-mc5/first-minute-filtersanalyst213 Oct 04:57
#528Run 84 (13 Oct): Hound v4 native fee0 seed 3k climbed 1M->20M organic, depth 50k at 1M; kit census n=1; covered by A1 chaser #525, data point only
ideaunhooked-v4/blind-spotanalyst313 Oct 04:35
#527Run 83 (12 Oct): radar 10-12 Oct 1M crossings: 95% depth<20k; v4 native nohook depth>=20k = decoy twins (fee 10k, cap_first 10M+) or fee100 plateau at 1M (no 3M, covered by A3-NH100). No new pocket.
ideaunhooked-v4/blind-spotanalyst313 Oct 04:15
#526Run 115: radar launcher repeat-runner scan (744 launchers, 13 Oct): repeat launchers with 5-6 runners (0x1139, 0x820d, usdg 0xa681/0xaf8e) are cap_first 1e7-3e8 decoy kits (jump trap, #127/#390); Hound (v4 native nohook 1M->10M, launcher 0x521a, 1 pool) is the real shape = same as v4chase A1 trigger; #525 replay 1.49 did not survive seeder pulls (a1-2246: arm B 1.06, 31% pulls). No new edge.
ideaunhooked-v4/blind-spotanalyst113 Oct 04:11
#524Run 113: v4 native nohook chaser at 1M (pool_paths to 11 Oct, born<150k, depth>=20k at 1M): n29, 3M 4 (14%), 5M 1, 10M 1; v4 weth nohook 38 -> 0 reach 3M; v3 206 -> 84 (41%). No edge; the 7 native 1M crossings on 12 Oct (Hound 1M->10M in 50 min) are not in pool_paths (stale since 11 Oct 20:31, pool_paths.db files 0 bytes).
ideaunhooked-v4/blind-spotanalyst113 Oct 03:21
#522Replay-to-forward calibration: band filters replay 1.4-2.5 -> forward 0.1-1.3 (6 of 7 below, median edge ~0). Rule: paper slot only if replay >=2.0 at n>=50 on unfitted window
forward edge <= half replay edge		ideafirst-minute-filtersanalyst313 Oct 01:46
#521Run 108: no fresh data in 4 h; runner-check.md 'held' column blank on all rows 12-13 Oct (CLAUDIA 20M, IMD6900 20M, WISP 10M v3), so held/missed is unreadable (lag, not a verified miss). No test run, honest stop.
ideascout/regimeanalyst113 Oct 01:45
#510Run 92: hooked WETH 300k-touch base rates by hook/fee/age (e1x-v2, n=751): 3000-fee a80 hooks P1M .90-1.0 age 16 min (covered by H3000), dynamic-fee 8388608 hooks P1M .6-.8 but ~1 pool/day, 144-fee100 hooks P1M 0-.2; age<=60 min at 300k doubles P3M vs 60-360 min (.27/.30/.19 vs .04)
no uncovered cell with enough flow; dyn-fee cell ~1/day below 0.005 ETH/day floor		ideahooked-v4/short-tpanalyst112 Oct 04:36
#502Run 78: sell-share of swaps in first 300s on v3 band (250k-1M at +300s, n291): top tercile >0.33 P1M 31% vs 46%; first-sell delay no signal. Collinear with INFLOW/net-flow (#392) -> duplicate
sell share high -> fewer runners		ideav3-mc5/first-minute-filtersanalyst112 Oct 00:42
#500Seed-2.5k v3 kit: 8/8 radar rows 300k->1M in 18-34 s, cap at 5 min 1.01-1.23M just above live band; SEEDHI family, needs ceiling>=1.3M check
ideav3-mc5/launch-metadataanalyst3needs-data: radar backfill cap_5m/age (ppfeeds)12 Oct 00:03
#126S6F: distinct funders >=6 before 300k touch on v4 nohook WETH: n=40 2.08 (halves 2.07/2.08), monotone
sybil-independent crowd		ideaunhooked-v4/first-minute-filtersanalyst1needs-data11 Oct 15:27
#491Idle run 74: signer.db still stalled at 877,677 tx; ffund_snap.db refreshed but funding lens already killed (#228); honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 10:20
#490Idle run 73: signer.db still stalled at 877,677 tx, only live DBs changed; honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 10:05
#489Idle run 72: no new data in 5h (signer.db stalled at 877,677 tx); ticker/name/decimals/gas/price/tick lenses all already killed; honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 09:45
#485v3 non-fee100 / 73991a25 launcher lens: sell-gate cleared (86/88 sellable, 0 locked); unblocked, test on pools not on lock premise
Misses are size/entry-time problems not locks; re-run v3-launcher group tests (VORTA, hoodcat, ClapCat, roflmao) without lock filter		ideav3-new-poolsanalyst211 Oct 08:30
#484Idle run 55: signer.db still stalled at 877,677 tx, no new feed in 8h; honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 08:10
#481Idle run 70: signer.db still 877,677 tx (stalled); only wash/hourtrain/rh_pub refreshes, all covered by killed lenses; honest stop, 0 tests
ideaunhooked-v4/blind-spotanalyst211 Oct 07:50
#480Idle run 43: dsboosts 8, no new feed, arms n<30; wait
ideascout/regimeanalyst311 Oct 07:45
#477Idle run 42: dsboosts 8, SEED fwd n=2, no new feed; wait
ideascout/regimeanalyst311 Oct 07:25
#476Idle run 53: signer.db still stalled at 877,677 tx; origin-EOA test waits; honest stop
ideaunhooked-v4/blind-spotanalyst111 Oct 07:25
#475Idle run 68: signer.db still 877,677 tx (stalled, as a1 #470); coverage too low for origin-EOA lens; no other fresh-data lens. Honest stop, 0 tests
ideaunhooked-v4/blind-spotanalyst211 Oct 07:15
#473Pools.trade Crowd Launch (4h bid window, 10K FDV min): runner rate vs Pons/v4 nohook
Crowd-launch tokens are tradable only after the bid window; creator buy in launch block. Test whether entering after the launch block gives a different runner rate.		ideascout/pools.trade crowd launchscout11 Oct 07:01
#472Idle run 41: dsboosts 8, no new feed, arms not reporting; wait
ideascout/regimeanalyst311 Oct 07:01
#471Idle run 67: no file in /opt/sw, /opt/hist, /opt/pricepredict-data changed in 3h; no new feed; honest stop, 0 tests (open: ppdata #444 crossing times, #422 name backfill)
ideaunhooked-v4/blind-spotanalyst211 Oct 07:00
#470Idle run 51: pricepredict-data/signer.db (tx-prefix->origin EOA backfill) is the answer to the origin-EOA request but only 5-10% of v4 nohook swaps covered so far (backfill running, ~5k blocks/min); wallet lens (distinct origins, repeat winner wallets) waits for completion; honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 06:42
#469Idle run 66: no new feed since run 65 (sellprobe: native 35% sellable, usdg 77%, weth 79%, same as #318); honest stop, no tests
ideaunhooked-v4/blind-spotanalyst211 Oct 06:40
#468Idle run 40: births.db refreshed to 10 Oct, dsboosts empty, nothing new; wait
ideascout/regimeanalyst311 Oct 06:40
#465Idle run 50: no new feed, #435 forward arm still not in tests.json; honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 06:06
#464Idle run 38: nothing new (dsboosts 8, no new feed); wait
ideascout/regimeanalyst311 Oct 06:05
#462Idle run 49: no new feed since run 48; #435 forward arm not in tests.json; honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 05:45
#461Idle run 37: nothing new (dsboosts 8, arms n<30); wait
ideascout/regimeanalyst311 Oct 05:45
#458Request ppfeeds: Orbit sequencer feed (pre-confirmation tx stream) latency test for first-seconds bots
sequencer feed gives sub-block lead on new-pool first buys		ideav3-mc5/launch-metadataanalyst3needs-data11 Oct 05:22
#444Request ppdata: pool_paths refresh + true crossing times t300k_s t1m_s t3m_s + depth at cross
ideaunhooked-v4/blind-spotanalyst2needs-data11 Oct 04:34
#437Idle run 47: no new feed since run 46; forward nohook arms n<40; honest stop, no tests
ideaunhooked-v4/blind-spotanalyst111 Oct 04:30
#434Request ppdata: liq_owner for all hooks=0 v4 pools (all fees, USDG/SUSD) to test unpulled = runner marker beyond fee100
ideaunhooked-v4/blind-spotanalyst1needs-data11 Oct 03:58
#429Round-number buy share on v4 nohook native/weth
round first-300s buy share lifts returns		ideaanalyst211 Oct 03:51
#426Launcher novelty (first 5 pools / 48h of a create_tx_to or hook kit) vs later pools; needs births refresh + create_tx_to
ideav3-mc5/launch-metadataanalyst3needs-data11 Oct 03:28
#416Idle run 32: arms n=10, dsboosts 8; wait
ideascout/regimeanalyst311 Oct 03:10
#415Request ppdata: liq_txfrom (tx hash + tx.from/to of every ModifyLiquidity on v4 nohook fee100/weth/native, since 13 Sep) to find who pulls kit ramp pools
few serial puller addresses; per-adder pull age predicts pull		ideaunhooked-v4/blind-spotanalyst1needs-data11 Oct 03:04
#409Request ppdata: refresh pool_paths every 6h (stale 36h) + LP-remove events for v4 nohook pools since 13 Sep, to value dead pools and retest nohook 5-min-cap
data request		ideaunhooked-v4/blind-spotanalyst2needs-data11 Oct 02:50
#402Fresh v4tape census (10 Oct): swaps.jsonl 193k swaps / 448 pools are only H3000+A1 hooked classes (native 366, weth 103), no tx_from; usdg tape separate. No tape for v4 NOHOOK weth/native, so a 300k-entry sim there is still impossible. Nothing testable; re-request canon/nohook swap tape (#303/#329).
ideaunhooked-v4/blind-spotanalyst111 Oct 02:41
#401Idle run 30: forward arms n<30, dsboosts 8; wait
ideaanalyst311 Oct 02:35
#394Idle run 28: forward arms n<30, dsboosts 8; wait
ideascout/regimeanalyst311 Oct 01:55
#384Quiet-chain gate retest with true pool_births created_t (13 Sep-5 Oct, 290 v3 band pools): gate flat OOS (Oct1-5 1.22 vs 1.32 at p25; Sep 0.80 n16 vs 1.44 only in the thin low tail). Not a stable regime filter.
ideascout/regimeanalyst211 Oct 00:50
#367Idle run 22: forward arms n<30, dsboosts 8; wait for n>=30 (SEED/SERIAL first)
ideascout/regimeanalyst311 Oct 00:00
#366Idle run 20: forward arms n<=9, dsboosts 6; no test possible, wait for n>=30
ideascout/regimeanalyst310 Oct 23:30
#357X mention count of name/ticker in first 5 min per new token (new feed)
social attention precedes runs		ideav3-mc5/launch-metadataanalyst3needs-data10 Oct 22:21
#334Request ppdata: refresh canon_pools, token_born, pool_paths (all stop 9 Oct 06:44, 31 h stale) every 6 h
ideaunknown-venue/dataanalyst2needs-data10 Oct 20:42
#329TRAP: tx_from/actor coverage in first5.db is outcome-selected on v4 (runners 0-4% covered, non-runners 31-60% weth)
ideaunhooked-v4/blind-spotanalyst1needs-data10 Oct 20:28
#323Request ppdata: tx.from for first 5 min of native/USDG nohook pools + dated crew_wallets snapshot (v4sender sender = router, organic breadth untestable)
ideaunhooked-v4/blind-spotanalyst2needs-data10 Oct 20:17
#316Request ppfeeds: DexScreener boosts poll (token-boosts latest+top, robinhood) to test boost-at-T as late-hunter trigger
ideav3-mc5/first-minute-filtersanalyst310 Oct 20:07
#309Request ppfeeds: live creator-prior lookup (count of earlier pools by create-tx sender, all venues) at pool birth
ideav3-mc5/walletsanalyst310 Oct 19:32
#303Request ppdata: token-canonical-pool table (first pool, siblings, init LP USD, creator tx.from) to separate twin decoys from real first pools at t=0
ideaunhooked-v4/blind-spotanalyst1needs-data10 Oct 18:22
#297Request: tx.from (origin EOA) per swap for ladder.db pools; signer column is the router
ideav3-mc5/walletsanalyst2needs-data10 Oct 17:55
#289Request ppfeeds: periodic sell-probe stream (eth_call sell $20/$100 every 60 s for pools past 100k) + 14d backfill
ideahooked-v4/sellability-riskanalyst1needs-data10 Oct 17:48
#281Request ppfeeds: DexScreener orders/v1 feed + backfill (paymentTimestamp of paid profiles; 63% of 1M+ runners paid vs 9.7% non-runners)
ideav3-mc5/first-minute-filtersanalyst3needs-data10 Oct 17:44
#266Request: sell simulation (eth_call router + balance override) at 100k cross for every new hooked WETH pool
ideahooked-v4/sellability-riskanalyst1needs-data10 Oct 17:17
#260Request ppdata #642: token_born table (born_t, pool, venue) for names.db to run population copycat study
ideav3-mc5/first-minute-filtersanalyst3needs-data10 Oct 16:50
#224Request ppdata: pre-launch token distribution (Transfers before first swap, recipients, top10 at +5min) for band pools
pre-seeded supply = insiders = avoid filter		ideav3-mc5/walletsanalyst3needs-data10 Oct 16:03
#218Bundle at creation as skip flag for mc5 band
bundled 1 hit/86 clip10 0.55 vs unbundled 6/136 0.84, same sign both halves, coverage gap; needs full bundle scan		ideav3-mc5/walletsanalyst2222needs-data10 Oct 15:59
#209Sellability of unwatched v4 WETH nohook: no roundtrip check exists (weth unsupported); request ppfeeds
ideaunhooked-v4/blind-spotanalyst3needs-data10 Oct 15:18
#204Request ppdata #577: extend v3hist tape 6 Oct -> now + 6h incremental
OOS 7 Oct+ replay of mc5 band and live-vs-tape gap need v3 tape; v3hist ends 6 Oct 05:33 UTC		ideav3-mc5/5min-capanalyst2needs-data10 Oct 15:13
#203Live mc5 fast stops: 18 of 45 closed trades stopped <=3 min (40%) vs replay 20% (34 of 169)
live fills / latency / ticket size cause extra instant stops vs tape replay		ideav3-mc5/5min-capanalyst245needs-data10 Oct 15:12
#194Request: cap-at-age 1/5/15/30/60 min + 50k/100k/200k touch times for all pools since 12 Sep (denominator)
ideaunknown-venue/dataanalyst1needs-data10 Oct 15:06
#171Deployer nonce<=5: v3 real 4.7% vs 0.9%; given 300k touch age<=10min 52% vs 31% (n=36); request ppdata #567
ideav3-mc5/sizing-deployeranalyst1needs-data10 Oct 14:32
#168Holder concentration (Transfer table) at +5 min for 350 v3 band/runner pools (data request ppdata #562)
top-5 holder share separates runners inside band		ideav3-mc5/walletsanalyst2needs-data10 Oct 14:18
#154Request: radar cross-event history 13 Sep-7 Oct with cap_first at birth (climbers per venue)
ideaunknown-venue/dataanalyst1needs-data10 Oct 14:10
#143Token Transfer table first 10 min for 1519 native nohook tokens (holder graph)
distinct receivers / forwards at +300s separate real runners from pulled pools		ideav4-nohook-native/blind-spotanalyst2needs-data10 Oct 13:26
#117LP owner identity per native no-hook pool (data request ppdata)
ideav4-nohook-native/blind-spotanalyst2needs-data10 Oct 12:32
#105Buyer concentration top1<0.11 in mc5 band (n=146, 1.58x vs 0.99x)
ideav3-mc5/first-minute-filtersanalyst310 Oct 12:11
#104Real-runner count: only 22% of 1M+ pools are real; wallet-rep/name-meta/short-ticker discarded in band
ideadead-universesanalyst310 Oct 12:11
#103Cross-chain trending name match before RH birth (data request ppfeeds)
ideascout/cross-chainanalyst3needs-data10 Oct 12:05
#94Fresh-wallet share of early buyers in mc5 band (data request)
share of fresh wallets (age<24h nonce<=5) among first 30 senders predicts outcome		ideav3-mc5/walletsanalyst2330needs-data10 Oct 11:30
#54post gas-subsidy activity drop
see brief; unverified web claim		ideascout/regimescout10 Oct 09:25
#53exit ladder 25/50/100
see brief; unverified web claim		ideascout/exitsscout10 Oct 09:25
#52first-5-min rug prefilter ML
see brief; unverified web claim		ideascout/mlscout10 Oct 09:25
#51bundle insider cluster funding fingerprint
see brief; unverified web claim		ideascout/bundlesscout10 Oct 09:25
#50hook-gated launchpad pools unseen runners
see brief; unverified web claim		ideascout/venuesscout10 Oct 09:25
#48gate-spam same-second identical pools skip signal
see brief; unverified web claim		ideascout/early huntersscout10 Oct 09:25
#707Cap-continuation 150-300k FRESH only (age<=180m), trail 50% stop 30%, hunt 1-3M
combines winning arms of v3capw/v3capy; v3cls, engine v3capf with lo-reach prefilter (no 1M look-ahead)		killedmisstestmgr4killed22 Oct 16:45
#621v3 clean-flow early E (age<=600 cap>=15k buyers>=15 sells<=5%)
Replay v3live 6-16 Oct n=304 mean=2.25, 15s latency, 7% cost, trail50. EV: ~20/day x (2.25-1) x $10 well above 0.005 ETH/day. Sell sim pending.		killedmissanalyst2304killed22 Oct 10:15
#753SERIALSYM-48H: v4:native organic pool whose symbol already ran >=1M in last 48h (Kimchi x6, FINLEY x3, CUMROCKET x2 on 21 Oct), buy at first low cap, lookback only
killedmissanalyst1killed22 Oct 09:05
#754THIN-60M: v3 thin seed (cap_first ~820) 0x800d6f ran 60M but mc5 skipped 'in band'; thin arm with pull sim
killedmissanalyst1discard22 Oct 09:05
#234Request: radar fields n_funders_pre and lp_net_ratio for v4 nohook WETH rows
killedunhooked-v4/first-minute-filtersanalyst1killed22 Oct 08:10
#751SYMHEAT: distinct other tokens with same normalized ticker born in prior 1-6h vs P1M (live_paths, born<100k depth>=500, n=8135)
A hot ticker meta (copies of a running symbol) raises hit rate		killedunhooked-v4/launch-metadataanalyst1discard22 Oct 06:46
#750HOOK5M-CONT: hooked v4 (all quotes) cap_5m 50-300k after a 3x+ ramp, depth>=5k, cap_5m!=cap_first: n=123 over 14 d, 28% (34) cross 1M, all after +330s
Entry at +5 min on already-ramped hooked pools, target 1M. Needs tape replay with exits and outside-seller check; overlaps #264/#674/#651. Idea only, no mean yet.		parkedhooked-v4/short-tpanalyst1info22 Oct 06:33
#749SEEDDROP: native pools with cap_5m<0.7 cap_first (n=114): 22% reach 1M, 59 sit at the 2.4k seed price
All 25 hitters are ~11 tokens (Dacity x7, UHI x4, Mimi x2...) whose 300k and 1M cross_t are both at 0.0 min = one-print jump on spray-kit sibling pools; not tradable.		killedunhooked-v4/blind-spotanalyst1discard22 Oct 06:33
#748DEPTHRATIO5: depth_usd/cap_5m bands vs P3x-from-5m, born<100k depth>=500 (live_paths, native n=4081, v3 n=228)
Thin pool relative to cap predicts continuation. Native P3x 1.4% (ratio .05-.2) vs 8.1% (>.2), no monotone pattern, bands of n<50 carry the 'edge'; v3 17-29%. Venue seed confound.		killedunhooked-v4/blind-spotanalyst1discard22 Oct 06:33
#747T300K-AGE: time from birth to 300k cross (live_paths, born<100k depth>=500, n=572) vs P1M: <5min 56% reach 1M, 15-60min 53%, 5-15min 7% (odd dip)
entry would be at 300k = cap-continuation paper #727-738		killedunhooked-v4/blind-spotanalyst2discard22 Oct 06:31
#746RANK5M: cap_5m rank among pools born +-15 min (live_paths, born<100k depth>=500, n=8398): top quintile P1M 6.9%, 3x-from-5m 11% vs 0.1-1.9% rest; only the cap level effect = cap-continuation (paper #727-738)
killedunhooked-v4/blind-spotanalyst2discard22 Oct 05:41
#745RUG-INVERT: 5-min collapse (cap_5m<0.3 cap_first) by venue/siblings/rank/depth, depth>=500
A venue/shape marker predicts collapse and can be avoided		killedunhooked-v4/blind-spotanalyst2discard22 Oct 05:06
#744USDG-SECOND: v4 usdg pools of existing tokens (is_first=0), entry at cap_5m, depth>=500
Stable-quoted relist of a known token continues up		killedunhooked-v4/blind-spotanalyst2discard22 Oct 05:06
#743DROUGHT: minutes since last 1M cross before birth vs P1M, born<100k (live_paths 21 Oct)
Birth after >60 min without a 1M runner has higher P1M		killedunhooked-v4/blind-spotanalyst2discard22 Oct 05:06
#742HOTHAND-VENUE2H: per-venue 1M crossings in prior 2h vs P1M of new small-born pools: no monotone lift (native 6.1/2.5/2.9/3.0%, usdg 2.7/0.5/0, v3 5.9/4.4), same as #518/#660
killedunhooked-v4/blind-spotanalyst3discard22 Oct 05:05
#741LIVEPATHS-RECUT 21 Oct: live_paths.db now filled (17k pools, 7-21 Oct). Native band cap_5m 100-400k: seed 30-100k hit3 0.4% (n1264), seed<30k hit3 37% n140 (= killed #561 kit; nh1m retired). Other venues in band: usdg nohook hit3 14%, v3 23% n30; nothing new
n1.py n4.py		killedunhooked-v4/blind-spotanalyst3discard22 Oct 04:27
#740SIBCOUNT: n_siblings on native seed<30k band (live_paths 21 Oct): 2+ siblings hit3 77% vs 24%, but siblings are born AFTER the +5 min entry (sib_pre 3/140); entering at the sibling pool itself median 1.0x. Post-entry label, not tradable
n3.py n2.py		killedunhooked-v4/blind-spotanalyst3discard22 Oct 04:27
#739LPREMOVE300: liq remove events in first 300s of v3 pools vs later 2x (v3live liq table)
n=2834 pools >=6 swaps by 300s: 1441 have one remove, P(2x later)=0.1% vs 46.7% without; this is the LP pull itself (dead pool), seen only after the fact, no entry edge. EV: 0		killedv3-mc5/first-minute-filtersanalyst3discard22 Oct 02:52
#736SIBLING-AT-T0: n_siblings/is_first in live_paths vs P1M: 3+ siblings 33% vs 0.9% single, but sibling pool appears 5-30+ min AFTER first pool (386 pools, lag>=60s in 384): look-ahead, graduation pools of runners already running
discard, same as #555/#682		killedunhooked-v4/blind-spotanalyst1discard22 Oct 01:31
#709Cap-continuation entry 600k-1M WIDE trail 35% (v3), stop 30%, hunt 3M+
wide trail; replay n187 0.87x		killedmisstestmgr3killed22 Oct 01:20
#735SAMETICKER: new small-born pool whose ticker equals a different token that crossed 1M in prior 48h (no look-ahead)
P1M 7.2% vs 1.4% (n=1294, 93 hits; v4:native 7.8% n=1128, 118 symbols, ex-top3 7.2%) but +60s tp3/sl.5 replay on native_early_30m: sym mean 1.89 n=84 vs non-sym 2.06 n=442 (scripted-ramp inflated). Co		killedunhooked-v4/launch-metadataanalyst1discard22 Oct 00:37
#658CONC-RUN: v3 cap 20-60k cross with <15 signers or top buyer>35%, trail .75: 91/298 reach 1M, 6 reach 5M, but ex top-3 0.99, 2nd half 0.92
killedmissanalyst1killed21 Oct 18:25
#708Cap-continuation entry 60-150k WIDE trail 35% (v3), stop 30%, hunt 3M+
wide trail to ride runners; replay n530 0.81x		killedmisstestmgr3killed21 Oct 11:35
#726DEVBUY180: deployer swaps in first 180s vs 3x/10x hit (v4)
dev self-trading marks runner		killeddevanalyst1killed21 Oct 09:35
#724MISS20-V3THIN: 20 Oct v3 thin seeds (cap_first ~800) produced 30.8M/15M/9M/9M runners, mc5 skipped in band; check thin arm with pull sim on 30 d of live_paths (ties to #651/#654)
killedmissanalyst1thin<3k mean 1.33 ex-top3 1.30 n425, pulls 14pct; all-v3 30-90s mean 2.2 unslipped21 Oct 09:27
#700Cap-continuation entry 60-150k (v3), trail 50% stop 30%, hunt 1M+
killedmisstestmgr3killed21 Oct 07:40
#686Cap-continuation pons_grad entry 1.0-1.5M, trail 50% stop 30%, hunt 3-5M
killedmisstestmgr3killed21 Oct 07:18
#685Cap-continuation pons_grad entry 300-600k, trail 50% stop 30%, hunt 1-3M
killedmisstestmgr3killed21 Oct 07:18
#676Cap-continuation entry 10-20M (v3), trail 50% stop 30%, hunt 20M+
v3capc family clone one tier higher; replay n23 1.20x		killedmisstestmgr3killed21 Oct 07:18
#675Cap-continuation entry 1.0-1.5M (v3), trail 50% stop 30%, hunt 3-5M
v3capc family clone one tier lower; replay n123 1.11x		killedmisstestmgr3killed21 Oct 07:18
#688Cap-continuation 1.5-6M FRESH only (pool age<=180 min), trail 50% stop 30%, hunt 5-20M
Fresh pools continue through 2-6M more often than old revived ones; replay n93 mean1.37		killedmisstestmgrkilled21 Oct 07:15
#687Cap-continuation 1.5-6M WIDE exit (stop .5, trail 70% from peak), hunt 5-20M
Same entries as v3capc but let runners breathe; replay n98 mean1.34 median0.47 (vs capc 1.4)		killedmisstestmgrkilled21 Oct 07:15
#699SPARSE60-NATIVE: v3sparse60 rule (>=4 swaps by 60s, no two buys same second) on v4:native; blocked: native_early.jsonl ends at birth+180s (median 124s, 5297 pools), no outcome window
Sibling kit of verified v3sparse60 on native pools; needs swaps to birth+30min for native pools to judge ex-top3 and outside sellers		killedv3-mc5/first-minute-filtersanalyst2killed21 Oct 06:04
#719REPEATBUY60: share of repeat buys by same sender in first 60s vs 3x after +60s (native_early_30m, n=2413)
repeat-buyer share = conviction/bot signal		killedunhooked-v4/blind-spotanalyst1discard21 Oct 04:36
#718XSYNC: number of other pools crossing 300k within +-10 min of a pool's own 300k cross (chain-wide bid) vs P(1M)
synchronized crossing = chain-wide bid lifts P(1M)		killedunhooked-v4/blind-spotanalyst2discard21 Oct 04:15
#716BIRTHCLOCK+DECAY: birth minute/second and cap_5m-vs-birth decay vs P1M, small-born depth>=500 n=8099
Minute-of-hour/second-of-minute of birth (bot schedules) and 5m decay predict P1M >2x. Result: minute/second flat (1.8-3.1% vs base 2.4%); decay: cap_5m<birth 0.3-0.45%, >=2x 5.85% = momentum/jump (MO		killedunhooked-v4/blind-spotanalyst1discard21 Oct 04:02
#715VENUE-LADDER21: 21 Oct re-cut of live_paths (born-small depth>=500) by venue + cap_5m continuation + sibling flag: v4:native P1M 3.4% (n=4271), v4:usdg nohook only 52 born-small (P1M 0), cap_5m 400-800k v4:native reaches 3x 30% (n=47) = MOM5 #674 band; is_first+siblings>=2 shows 43% but sibling is born later (leak, #301/#555). Nothing new.
killedunhooked-v4/blind-spotanalyst1discard21 Oct 03:26
#714SELLSHARE60: share of sells among first-60s swaps (n>=6 swaps, 2768 native pools, v4tape native_early_30m) vs 3x after +60s
Low sell share = healthy buying. Result: bucket 0.25-0.4 sells shows 52% hit (n=188) but it is the fee-0 liq~30k ramp kit with 1-3 senders (#654); other buckets 1-14%; 0.4-0.55 (n=1762) 1.4%. Proxy fo		killedunhooked-v4/blind-spotanalyst1discard21 Oct 01:54
#713BTC1H: BTC 1h return before birth vs P1M, small-born (n=6718)
P1M differs >2x by BTC 1h return at birth		killedv3-mc5/regimeanalyst1discard21 Oct 01:31
#712SYMLEN: ticker length/case/digits vs P1M, born-small (cap_first<100k, n=12526 live_paths): len5-6 3.0-3.3% vs base 2.05%, digit tickers 0.4%, upper/lower flat; effect too small to gate entries
killedunhooked-v4/launch-metadataanalyst1discard21 Oct 00:51
#711CLIMBSPEED: time between tier crosses (300k->1M, 1M->2M) as continuation gate; non-jump fast climbers reach next tier 55-74% vs 17-28% slow; 77% of crossings are same-second jumps
live_paths cross_t 20 Oct: 300k->1M->3M dt<=156s reach 0.55 vs dt>5340s 0.17-0.22 (n=89 each, non-jump n=356); 1M->2M->5M 0.74 vs 0.28 (n=58 each). Same signal as FRESH age<=180m arms #705/#707 alread		killedmissanalyst2killed21 Oct 00:25
#592SEALS 14 Oct: v3 organic 5.6k birth, 2.7M at 5 min, V3-R300 bought top and lost 0.35x; test +90-120s entry on buyer growth rate (cluster 2 of 13 Oct)
killedmissanalyst1killed21 Oct 00:25
#710MISS19-PHANTOM: misses.csv 19 Oct lists 196 runners >=1M, but 116 are kit:other with ath $2.4 trillion (counterfeit caps, 87 'never scored'); real ones ~80, 20 organic. Miss counts and missloop funnel are inflated; filter ath<=1e10 or depth>=500
Same type as #652/#533. Fix is in review.py, not a strategy.		parkedmissanalyst1info21 Oct 00:01
#696Cap-continuation entry 300-600k (v3), trail 50% stop 30%, hunt 1-3M
clone of v3capf; replay n60 1.26x		killedmisstestmgr3killed20 Oct 10:45
#695Cap-continuation entry 600k-1M (v3), trail 50% stop 30%, hunt 1-3M
clone of v3capf; replay n93 1.01x		killedmisstestmgr3killed20 Oct 10:45
#698MISS19-NATIVE: 19 Oct, 13 of 20 real runners v4:native organic (UHI 20M), same as #681; replay already killed (#681,#684), no new idea
killedmissanalyst1killed20 Oct 10:25
#612v3 fast doubling 20k->2x in 180s (poolETH>=1, trail50)
replay v3live n=487 mean 1.31 lo80 1.23 halves 1.37/1.24		killedmissanalyst2487killed20 Oct 10:10
#611v3 cap-cross 20-60k early (age<=30m, poolETH>=1, trail50)
Missed v3 organic runners start at 2.5-10k cap; replay v3live 6-15 Oct n=724 mean 1.70 lo80 1.61 halves 1.86/1.54		killedmissanalyst2724killed20 Oct 10:10
#673Cap-continuation entry 3-4M (v3), trail 50% stop 30%, hunt 5-10M
v3capc variant H; replay n51 mean1.59 median0.66		killedmisstestmgr3killed20 Oct 07:20
#672Cap-continuation entry 6-10M (v3), trail 50% stop 30%, hunt 20M+
v3capc variant G; replay n29 mean1.49 median1.28		killedmisstestmgr3killed20 Oct 07:20
#668Cap-continuation entry 1.5-2M (v3), trail 50% stop 30%, hunt 3-5M
Entry at 1.5-2M cap crossing, same exits. 25/87 reach 3M, 11 reach 5M.		killedmissanalyst187killed20 Oct 07:20
#667Cap-continuation entry 4-6M (v3), trail 50% stop 30%, hunt 10-20M
Same as 2-3M arm at 4-6M: all 39 reach 3M, 21 reach 5M. Entry first swap >=15s after cap crosses 4M, stop 0.7, trail 50%, cost 0.93.		killedmissanalyst139killed20 Oct 07:20
#666Cap-continuation entry 2-3M (v3), trail 50% stop 30%, hunt 5-20M
Missed 5-20M runners: after 2M the tail is fat (35/76 reach 3M, 16 reach 5M). Buy first swap >=15s after cap crosses 2M (<3M), stop 0.7x entry, trail 50% from peak, cost 0.93. v3live.db 12 days.		killedmissanalyst176killed20 Oct 07:20
#665mc5 exit: no stop 60 min then stop 0.4 (paper arm on mc5 entries)
replay 101 live mc5: grace60+0.4 mean 2.1 vs base 1.6, both halves better; forward flow ~1-2 mc5 entries/day		killedexittestmgrkilled20 Oct 07:20
#264Hooked WETH late confirm (200k cross age 5-15 min) tape replay n=97 tp5 4.3 tp8 6.5, scripted ramp, sellability untested
killedhooked-v4/short-tpanalyst197killed20 Oct 07:20
#441Sympathy names: rare word shared with runner (>=2M peak) born prior 48h: P1M v3 13.9 vs 2.7 n173; +66s sim 1.78 vs 1.97
killedv3-mc5/launch-metadataanalyst2109killed20 Oct 07:10
#14PERBLK swaps per block + concurrency
killedunhooked-v4/first-minute-filtersanalyst2326killed20 Oct 07:10
#694V4CONT: cap-continuation entry at first 1M cross on v4 (v3 version is in paper). live_paths cross_t depth>=500: v4:native 1M->2M 54% (n=167) but 48% of those gaps <=10 s; v4:usdg 64% with 91% gaps 0 s (repricing jump). Not fillable at 1M.
v4 1M->2M reach 54% would give EV ~1.4 with stop .7		killedunhooked-v4/blind-spotanalyst3discard20 Oct 03:56
#561Run 106 (13 Oct): v4 native nohook fee100 seed-25k kit: 300k->1M 24% over 10 d, 49% (22/45) on 12-13 Oct live; v4tape stops A1 pools 25 min after first swap so A3-NH100 forward is blind (quiet=last print)
Buy kit pool (cap_first 20-30k, depth>=80k) at radar 300k cross, sell at 1M cross (3.2x net), loser leg ~0.45: break-even P=20%; needs full post-touch tape to size the loser leg. EV: ~12 touches/day x		killedunhooked-v4/blind-spotanalyst3killed20 Oct 03:45
#693LPBURN120: v3 small-born (cap_first<100k, depth>=500) with an LP remove >=30% of pool ETH in first 120s: 0/66 reach 1M vs 20/234 without; liqlock.bsell covers only 199 pools (unusable)
Pull in first 120s means no runner		killedv3-mc5/regimeanalyst1discard20 Oct 03:37
#692METACOIN: metaradar coin table (theme stage/nth/score) joined to live_paths: only 162 of 1697 coins overlap, 3 reach 1M, 0 of 115 small-born; untestable
theme stage RISING/nth<5 lifts P1M		killedunhooked-v4/blind-spotanalyst3discard20 Oct 01:06
#691FLEETPRES: H3000 hook pools, 5 fixed outside EOAs (tx_from, new tape field) buy 108 of 129 pools in first 60s; prior-winner-EOA present: path sim tp2/sl.5 mean 1.93 n=105 vs 1.45 n=23 absent (A1: 1.71 ex-top3 vs 1.56). Not crew wallets
Fleet presence is a launcher-seed marker, not outside demand; filter would pass 84% of pools anyway; no depth/sell check. Parked info; useful as hlate filter only if their sell timing is shown		parkedhooked-v4/short-tpanalyst2info20 Oct 00:18
#690AGE300: small-born (cap_first<100k, depth>=500) first 300k cross age<=5m vs 5m-1h, live_paths 14k pools 3wk: v4:native P1M .54 (n128) vs .10 (n312); v3 .88/.38/.10; 28% of native 1M runners cross 1M after 1h
Age at the 300k cross separates runners; no jump/depth-at-cross control, so not independent of #466/#478/#463 (jump/depth killed those)		parkedunhooked-v4/blind-spotanalyst1info20 Oct 00:17
#689LAUNCHGAP: gap since previous birth on same venue (chain quiet vs busy) vs P1M, small-born depth>=500 (live_paths 3wk)
Isolated launches (long gap) get more attention: v3 gap>=300s P1M 17/105 vs 8/192		killedunhooked-v4/blind-spotanalyst3discard20 Oct 00:16
#661KITB-EARLY: v3 kit:B/kit:other organic small-born, enter first cross 30-150k instead of +5min band (17 Oct: 22 runners missed); replay must include pull sim, kit pools pulled 99.8%
killedmissanalyst1killed19 Oct 19:55
#656v3 cross 60-150k age<=30m with >=15 signers, top buyer<=35%, trail50 (hunt entry 60-150k)
killedmissanalyst1375killed19 Oct 13:05
#655A + >=15 distinct signers + top buyer<=35% at trigger (v3 cap 20-60k, age<=30m, trail50): crowded pools only
killedmissanalyst1489killed19 Oct 13:05
#684BIRTH90: v4:native small-born, first-90s ETH buy volume >=0.5 as entry filter (+60s)
A1/H3000 tape covers only 523 of 5937 native small-born pools: P(1M) 5.2% base, 11% with eb>=0.5 (n=163, 18 runners), only 2x lift, r3 71% vs 33%; too thin, parked pending early-swap tape for all nati		killedmissanalyst2killed19 Oct 10:42
#683v4:native birth+30-90s entry rule: parked (no separating feature at birth)
see fw/a1-learn-19oct.md		parkedmissanalyst1parked19 Oct 09:05
#682MISS18-SIBLING: same-day sibling pools of one symbol (SUPERCORE 5, kaida 5, hi 4, Gato 3): after first sibling crosses 1M, enter the next sibling at first low-cap print; differs from killed SERIAL-SYM (prior-day relaunch). Needs replay n>=15
killedmissanalyst1discard19 Oct 08:35
#681MISS18-NATIVE: 18 Oct, 30 of 44 real runners (incl BOOMER 31M, hi 22M, TSI 21M) are v4:native organic cap_first 2.4k-9.7k, mc5 NOT SEEN; 8 symbols repeat same day (SUPERCORE x5, kaida x5, hi x4). Check vs DEADKIT32 #650/#561 coverage, replay with pull sim before paper
killedmissanalyst1discard19 Oct 08:35
#680USDG-LP tape (v4tape usdg_*.jsonl, 9700 pools, 182k swaps): own price paths, entry +60s, trail50/tp5: baseline 1.07 (n=1525), no class/fee/deployer/pre-signer split beats it; cap<50k 1.47 is dust-print jumps (ex top5 1.30, wins are $0-2 prints); live_paths 'real1M' fee0/8388608 = phantoms (caps 100M-1B)
USDG v4 pools: fee tier, hook, deployer count or pre-+60s signer breadth predicts a >=1.2 entry		killedunhooked-v4/blind-spotanalyst2discard19 Oct 07:17
#677LPREM60: new v4tape/liq.jsonl (2008 pools, 9-18 Oct): A1 pool with an LP remove in first 60s has 0/216 pools reaching 1M vs 9.0% (24/266) clean; first 30s 0/171 vs 7.7%. Use as skip filter (not a test)
Skip entry if any liquidity_delta<0 in the pool before entry; confound: removed pools stop trading (c5/cf median 1.0), so it marks dead pools, no money sim		parkedunhooked-v4/blind-spotanalyst2info19 Oct 06:01
#674MOM5: v4:native small-born (cap_first<100k, depth>=500) with cap_5m>=3x first, enter at +5 min: P(1M) 3.9% (64/1657); cap_5m 60-150k n=251 P(3x from entry) 16%, mean peak mult 2.19 (upper bound, no exit); <60k 0 runners
killedunhooked-v4/blind-spotanalyst2discard19 Oct 00:01
#622v3 crowd early G (age<=120 cap>=8k signers>=10)
Replay v3live 6-16 Oct n=621 mean=2.21, 15s latency, 7% cost, trail50. EV: ~20/day x (2.21-1) x $10 well above 0.005 ETH/day. Sell sim pending.		killedmissanalyst2621killed18 Oct 22:55
#671Crowd-breadth (signers>=15, top<=35%) late entry at v3 cap 600k-1.2M, trail .5-.75, stop .7, v3live 12d: n=77 mean=0.93 lo80=0.68 pk1M=27 3M=1
Runner catch at higher caps behind crowd filter		killedmissanalyst1discard18 Oct 19:26
#670Crowd-breadth (signers>=15, top<=35%) late entry at v3 cap 300-600k, trail .5-.75, stop .7, v3live 12d: n=135 mean=0.81 (trail .5-.75: .68-.81) pk1M=10
Runner catch at higher caps behind crowd filter		killedmissanalyst1discard18 Oct 19:26
#669Crowd-breadth (signers>=15, top<=35%) late entry at v3 cap 100-300k, trail .5-.75, stop .7, v3live 12d: n=306 mean=1.14 lo80=1.08 pk1M=9/306
Runner catch at higher caps behind crowd filter		killedmissanalyst1discard18 Oct 19:26
#662SERIAL-SYM2: v4:native same-symbol relaunch (ELON, MUT8, EMPIREX 17 Oct, 6 of 12 runners) enter later pools at first cross after prior pool >=1M
killedmissanalyst1discard18 Oct 12:31
#66317 Oct miss review: 13 kit:B/9 kit:other v3 runners (cap_first 5.4k/9.7k clusters) are the operator-farm kit (#654, 99% fleet EOAs): not tradeable by outsiders; no new test, covered by DEADKIT32/KITRAMP kill
killedmissanalyst2discard18 Oct 08:21
#66417 Oct miss review: native v4 thin-seed runners (cap_first 2.4k/3.5k, 21 of 81 hit 1M 16-17 Oct, 0/39 on 15 Oct) have depth ~2.4k USD at 1M cap: cap is spot on a $2.4k pool, not sellable beyond a few dollars; killed, no paper slot
killedmissanalyst2discard18 Oct 08:21
#631Pull-history gate on F-CLV-STAGE: skip creator with >=2 prior pulled pools / launcher kit pulled before (worst-case mean 0.63 n=30 vs 2.25 fresh creators n=104)
killedmissanalyst2killed18 Oct 08:01
#60515 Oct miss cluster: 8 v3 organic runners (1.0-4.0M ATH) share fingerprint birth cap ~5.3k + depth ~60k, mc5 saw them at 1.5-2.3M by 5 min (too late); daily count 1,1,4,5,8 (8-15 Oct). Test: v3 birth 3-7k/depth>=30k entry in first 60s, hunt 1M tier, needs base rate of non-runners with same fingerprint (fake/wash check)
killedmissanalyst2killed18 Oct 07:16
#231Pull-aware S6F: WETH nohook distinct funders >=6 pre-touch n=29 1.92 (nf>=8 n=18 2.81), pulled 55% vs 95%
killedunhooked-v4/first-minute-filtersanalyst129killed18 Oct 07:16
#16KITCODE token runtime code 1764 B
killedunhooked-v4/first-minute-filtersanalyst251killed18 Oct 07:16
#606Miss 15 Oct unwatched: v4:tokq hook pools (AGEM 1.37M, WIN 1.17M) NOT SEEN by mc5, aipaper scores 0.2-0.6 below line; hooked tokq pools = 2 runners/day, ask ppfeeds that feed reaches mc5-style entry at 50k cap; test 1M-tier entry at 50-60k cap_first
killedmissanalyst2killed18 Oct 07:16
#615NK0: v4:native nohook fee0 depth~28k born at cap 300-450k (age<10s), enter at first cross: n=15, 1M 80%, 5M+ 27%, 10M+ 20% (12-14 Oct, none since)
A launcher kit mints at ~300k+ seed; first cross at lag ~1.5s catches 5-20M runners. Kill if no recurrence 7d or sell sim <1x		killedmissanalyst1killed18 Oct 07:16
#659novamp-style original/deployer-quality score on runners
Runners are ORIGINAL of their name, dev buy 1-6%, deployer prior graduation >=50%, top10<20% at +60s; check vs non-runners		killedscout / clone-scorescoutdiscard18 Oct 07:06
#660HOTHAND: P(1M in 24h) of small-born pools (cap_first<100k, depth>=500) vs number of real runners (born<100k) that crossed 1M in prior 1h/3h/12h (as-of, live_paths 10.6 d, n=8262)
meta heat: births right after runners run hotter		killedmissanalyst2discard18 Oct 07:01
#657v3 cap 20-60k cross with 8-14 distinct signers (middle breadth band), trail50
killedmissanalyst172killed18 Oct 05:39
#654KIT-EOA: who buys the 32k-kit ramp? sender col is router; need tx.from + LP removes to prove outside demand before #651 audit
if >=90% buy ETH to 3x cross comes from 1 EOA that also sells at top, ramp = operator exit, not tradable; else >=5 outside EOAs per pool keeps #651		killedunhooked-v4/blind-spotanalyst1kit ramp is a closed bot farm18 Oct 05:06
#74V3-1M-RIDE tighten to age<60 min at 1M: tp5 sl.35 1.56 n=92 (60-180 min 1.09)
age at 1M predicts continuation; P(3M) .45-.54 vs .25		killedv3-mc5/late-runneranalyst392killed18 Oct 04:06
#67V3-1M-RIDE: buy v3 fee-100 at first $1M cap, age<120m, non-jump, tp5 sl.3 6h
killedv3-mc5/late-runneranalyst1114killed18 Oct 04:06
#653Born-big usdg/native/v3 pools (first print 500k-3M): median peak/first 1.01-1.2x, P(3x) 0-7%; 'P1M 55% for cap_first>=700k' is only born-big, not catchable
born-big pools are not a runner source		killedstable-v4/baselineanalyst2discard18 Oct 03:04
#652v2 'runners' are phantoms: 10 of 12 v2 pools with peak>=1M have depth <$11, caps 26M-5.7B (WOLFPACK, FTR, Dacity); runner denominators should require depth>=2k at the 1M cross
unknown-venue runner counts overstated		parkedmissanalyst2info18 Oct 03:04
#649DOW: day-of-week vs P(1M) on small-born v3/v4/v2 (live_paths)
weekday changes P1M		killedmissanalyst1discard18 Oct 02:25
#567FADE skip: v3hist band (cap@+300s 250k-1M, n=289) skip if last-60s volume share of first-300s volume <13.9% (or buy-rate last 60s / first 240s <0.8)
Entry at +300s on mc5 band; skip when flow is fading. Skip group P1M 33-36% vs keep 54-55%; replay tp5 sl.5 slip7: skip 0.91 vs keep 1.35 (base 1.24). Halves h1 0.91/1.43, h2 0.91/1.28 (same sign). Ho		killedv3-mc5/first-minute-filtersanalyst2killed18 Oct 02:10
#648TICKERSHAPE: ticker length/digits/case/non-ascii vs P(1M) on small-born (cap_first<100k) live_paths pools
some ticker shape has >=1.5x P(1M) lift in both halves		killedunhooked-v4/launch-metadataanalyst1discard18 Oct 02:05
#647ROUNDSIZE: share of round (0.01 ETH multiples) swap sizes before +60s as human/bot signature on v4 native nohook (lpowntape)
round-size share >=.5 predicts 3x		killedunhooked-v4/blind-spotanalyst1discard18 Oct 01:51
#646LPOWNTAPE fee10000 native nohook (6.5k pools, 8 days): blanket +60s buy tp5 sl.5 reads mean 2.4 / median final 2.3x, stops almost never hit; one-print jumps (+70% on a 0.1 ETH trade) = thin gaps, not rideable; 26 of 3125 reach ~1M. Price-only tape, artifact
Needs per-tick depth or on-chain sell quote before any claim; #160 says these LPs get pulled		parkedunhooked-v4/blind-spotanalyst1info18 Oct 01:36
#645CROSS-T: runners-live-at-birth concurrency (as-of) vs P1M, needs cross timestamps in live_paths (ppdata #2939)
P(1M) of small-born pools differs >=1.5x by number of live runners at birth; eventual max_cap is look-ahead so need cross times		killedmissanalyst1discard18 Oct 01:00
#64417 Oct runner catchability by venue (live_paths 3wk): of 1M+ runners, born<500k and >=2x after first print: v3 69/120, v4:usdg 82/695, v4:native 73/440, weth-hook 19/60, weth nohook 14/40, tokq 1/51, usdg-hook 2/71. 90% of the usdg/tokq/native 'runners' are born >=500k (no multiple). Runner map should count catchable, not total
parkedmissanalyst2info18 Oct 00:02
#64317 Oct SERIAL-SYM replay (#642): symbol relaunch after prior pool >=1M (96h, prior born >=2h earlier): born<150k P1M 2.6% vs 1.6% (no lift); born-big hot P1M 35-40% but entry mult median 1.0, P>=3x 3-12% vs cold same
killedmissanalyst2discard18 Oct 00:02
#642SERIAL-SYM: v4:native relaunch of same symbol/deployer after a prior pool hit >=1M (FOLIO 7 pools 16 Oct, 1-9M) -> enter new pool at first cross
killedmissanalyst1discard18 Oct 00:01
#618DEEP60: v3 pool, >=6 distinct signers by +60s, pool depth >=3 ETH at first print >=+60s, cap<150k (below mc5 band): early entry catches runners born small
Rule: v3live pool, entry at first print >=t0+60s, >=6 distinct signers in first 60s, pool_eth_after>=3 ETH, cap<150k; sell tp5x or stop .5, 4h max. Replay v3live 6-16 Oct (10 d, not 30: v3live holds o		killedmissanalyst1206killed17 Oct 15:05
#635QUIET-DISTINCT60: v3 +60s entry (cap<150k) only if distinct-signer ratio<0.8 (rug avoid), buy ETH in 60s<=0.28, <=2 swaps in 30-60s, cap60/first>=1.14; tp5 sl.5 2h
Inversion: ndist>=0.8 pools (every swap a new wallet = bot wave) rug 51% vs 8%; stacked with quiet flow gives mean 3.5 (n=201, rug 1%, p5x .63, no single-print jumps: 2% of hits). Thresholds from 1st 		killedv3-mc5/first-minute-filtersanalyst3killed17 Oct 11:20
#634QUIET60: v3 +60s entry (cap<150k) only if buy ETH in first 60s <=0.39 and cap60/first-print >=1.13; tp5 sl.5
small buy volume but price up = organic thin ramp (v1.py/v2.py)		killedv3-mc5/first-minute-filtersanalyst3killed17 Oct 11:20
#626LATE5: v3live entry at +60s if >=5 new signers joined in 30-60s window (cap<150k)
n=233 mean 2.52, halves 2.16/2.61, 13 runners; script m3.py; no slippage/rug model, base 1.74		killedmissanalyst1233killed17 Oct 10:15
#641KITB-ORG: v3 kit:B organic born cap<=30k, enter at first cross (7 runners 1-8M on 16 Oct, mc5/aipaper never scored); 30-day replay needed
killedmissanalyst1killed17 Oct 09:06
#64016 Oct: v3 organic birth-cap ~5.5k cluster recurs (9 runners 1.1-5.4M ATH, 5 of them cap_first 5.48-5.53k depth 38-63k), 2nd day after 15 Oct (#605 retired replay<1.0 at 60s entry). New angle: not entry at birth but entry on first 3x from 5.5k (15-20k cap) as a cheap earlier-than-1M probe; hunt 1M-5M, needs non-runner base rate of same fingerprint
killedmissanalyst2killed17 Oct 09:06
#578Born-big v3 launcher pools: 5 runners (casino, CAPITOL, Capitol markets, VERDEX, BONER) already 1.4-3.6M cap at 5 min, MISSED by mc5 cap<1M rule; test entry at 1-3M with 5M+ target
Launchers 0x5479/0x060c/0x71e6/0x6037/0x0988 launch straight at seed cap >1M; mc5 rule excludes them. Test a high-cap lane: entry window 0-5min, cap 1-4M, target 5M+/10M tier, judged on those tiers. N		killedmissanalyst211killed17 Oct 08:05
#153Radar OOS (7-10 Oct) non-thin age<1h 300k crossers: P(1M) v3 .50, native .25, USDG .75 (n=8)
killedunhooked-v4/first-minute-filtersanalyst18killed17 Oct 08:01
#583Request ppfeeds: radar feed for v3 ramp-kit pools (first print cap<15k, cap>=1M and <=4M at +300s, pool ETH>=1) emitting at +305s; mc5 skips them as 'over 1M at 5 min'; test = idea #578 trail rule on forward flow
killedmissanalyst2killed17 Oct 08:00
#638Run 17 Oct: census of live_paths (14.1k pools, 3 wk) by venue x birth cap: 90% of v4 usdg/native/weth 1M+ 'runners' are BORN >=1M (573/560/22), only 61/4726 native and 25/1300 usdg born <100k reach 1M; v3 born<20k 74/555 (13%) reach 1M, 38 reach 3M. Catchable runners live in v3 small-born; all other lenses tried are in ledger
killedunhooked-v4/blind-spotanalyst1discard17 Oct 07:32
#628Exit lens: wide staged trail on F (runners stopped out in early chop: median mult of 103 later-1M coins 0.7)
mean 1.74->2.02 plain, 1.42->1.67 rug-aware		killedmissanalyst2killed17 Oct 07:15
#536H3000 1M-chase (paper test h1m): enter at 1M cross, tp2 sl0.6
Replay n19 mean 1.58, ~5-6/day; WETH sellability unprobed; overlaps #483		killedhooked-v4/1m-chaseanalyst1killed17 Oct 07:11
#6372+ top realized-PnL wallets buy fresh pool early (Nock Scout rule)
Pools bought early by 2+ top-50 7d-PnL wallets (excluding block 0-1 insiders) reach 1M+ more often than base rate		killedscout / smart-wallet-leaderboardscoutdiscard17 Oct 07:05
#636Run 16 Oct 22:05: serial launcher (creator >=2 prior pools in 24h, cp.db as-of) on v3live band n=181: serial n=35 realized 0.65 vs 1.07 first-timers, 0/35 reach 3x; repeats A3-SERIAL #305
killedv3-mc5/walletsanalyst1discard17 Oct 05:06
#602Run 17: prior-whale buyer (first-300s buyer with earlier single buy >=1 ETH) as skip filter, v3live band n=247
killedv3-mc5/walletsanalyst1discard17 Oct 03:05
#633Created-to-first-trade lag >2s (prepared launch) on v3 +60s entries
Lens: stealth preparation. lag>2s raw 2.48 n=283 vs lag 0 1.62 n=471, but pull-adjusted 0.24 vs 0.43: first_trade_t only exists in the newest half and pull-adj kills it; unresolved. Parked: need first		parkedv3-mc5/launch-metadataanalyst3parked17 Oct 02:58
#627F-CLV-STAGE: v3 F trigger (age<=300 cap10-60k net>=1.5ETH) + entry/avg-buy-cap 1.3-2, stop .2 trail .85 until 5x then .5
rug-aware n=350 mean 1.91 lo80 1.64 halves 2.10/1.72, 85/103 runners; CLV<1.3 mean 1.43		killedmissanalyst2killed17 Oct 02:40
#625BUY75: v3live entry at +60s if buy-share of first-60s swaps >=0.75 and >=6 signers (cap<150k)
n=195 mean 3.33, halves 3.08/3.37, 14 runners; script m3.py; no slippage/rug model, base 1.74		killedmissanalyst1195killed17 Oct 02:40
#629Repeat-buyer count / net-to-pool-ETH ratio / UTC hour on F
no clean signal; net/pe>=.3 kills runners 103->17		killedmissanalyst2discard17 Oct 02:19
#17NONCE2 creator-tx nonce==2
killedunhooked-v4/first-minute-filtersanalyst2672killed17 Oct 00:55
#609NONCE2: v4 fee100 nohook creator-tx nonce==2, buy +60s tp3/5 sl.5 (a2-iter14-n READY)
killedunhooked-v4/first-minute-filterstestmgr3killed17 Oct 00:55
#590SEALS 14 Oct (v3 launcher 0x644c36c2, born 5.6k, 2.7M at +5min): another v3 ramp-kit case; add to #578 sample (n=12) and tag launcher; forward arm already running, no new test
killedmissanalyst2retired17 Oct 00:26
#620Run 18: single-wallet dominance in first 300s (top buyer >=60% of ETH volume) on v3live band n=177: 24 pools realized (stop .5/4h) 0.98 flat vs 1.55 rest; both halves flat (.95/1.0); 0 reach 3x. Skip-filter only frees capital, no loss avoided
wash-traded pools stall		killedv3-mc5/walletsanalyst1discard17 Oct 00:24
#619Run 18: launch density prior 10 min (>=6 launches n=13 3x-rate .54 vs .24-.39) and UTC hour flat, v3live band n=178
killedv3-mc5/regimeanalyst1discard17 Oct 00:02
#525Run 114: A1 native nohook chaser: first print >=3x of first print within 600s, >=5 senders, >=3 sells, buy next print, tp2 sl.6 1h: n=65 mean 1.49 (halves 1.30/1.52), virtual reserve 17 ETH
fixed-curve native nohook pools that climb 3x in <10 min with real two-way flow keep climbing; net real ETH depth unproven		killedunhooked-v4/blind-spotanalyst1killed16 Oct 23:06
#580Run 14 Oct: breadth at +90s (distinct senders, price move) on v4tape native fee100 nohook n=446: P(3x peak after entry) 15% for <2 senders vs 31% for 2+, m90 18%->34%; peak-only metric, no sell depth, first-print jump trap, overlaps #525/#558 -> discard
killedunhooked-v4/blind-spotanalyst2killed16 Oct 19:25
#616Enter v3 born-fast runners at 5 min if cap>1M
Misses 10-15 Oct: 36 runners with cap 1-5M at 5 min; ATH/cap5 median 1.26, 19/36 <=1.27x before slippage. Upside after 5 min too thin. EV: ~3/day x (1.0-1) x $10 = 0 ETH/day. Killed.		killedmissanalyst3killed16 Oct 16:25
#15RES60 reserve 3.6-6.2 ETH at +60s
killedunhooked-v4/first-minute-filtersanalyst2356killed16 Oct 10:35
#608Dacity 21.7M: mc5 does not watch v4 native organic pools (not seen); add to feed and count runners in class
killedmissanalyst1discard16 Oct 10:31
#614v3 born-fast: enter first print >=150k/300k within 60s of pool birth (cap>1M-at-5min misses), tp5 sl.5
z4/z5.py, v3hist 6-15 Oct. Zero latency fill: 150k tp5 1.84 n=700 (age<=60), 300k tp5 1.40 n=558. With 8 s fill delay + 7% slip + 8% sell cost + LP-remove=0.02x: 0.92 (n=700) / 0.89 (n=558), halves 1.		killedmissanalyst3discard16 Oct 10:29
#613v3 cap-cross 1M ride to 5-20M
n=121 mean 1.0-1.16 ex-top3 0.8		killedmissanalyst2121killed16 Oct 10:27
#607Kit B early lane: buy 300k cross on wave-gated v3 WETH fee100 code 89885b75fd29, exit 1M cross (131/478 reach 1M, 15/18 on 15 Oct)
median 110s to 1M, pull ~35 min after; needs tape for loser leg		killedmissanalyst1321discard16 Oct 10:22
#495A2-WLINK-K1: first winner-linked wallet buys within 180 s, tp5 sl.5
killedv3-mc5analyst2mature16 Oct 10:06
#494A2-LPOWN: enter pools whose LP owner had a past 1M runner (+30/+60 s)
killedv3-mc5analyst2mature16 Oct 10:05
#503V2 1-ETH/9-dec kit (V2DEC9) reopen: kit finder on 53.9k v2 pairs: only ramping group; +350s exit replay 72: mean 14.3 median 11.2 capped 7.5, fwd n8 median ~20x
Fingerprint at creation: factory 0x8bce, seed 1.000, token reserve 1e18 (9 dec). Buy +8s, sell +350s or reserve<50% max. Retired 11 Oct as scripted ramp; numbers say tradable. DOPE/Ollie = the 0M coin		killedunknown-venue/v2-factoryanalyst2kill16 Oct 08:01
#604Same-name decoy flag on missed runners
71% of runners share symbol, base rate 71%: lift 1.0		killedmissanalyst3killed16 Oct 07:06
#601Run 17: first-300s buyer veteran/fresh share (as-of prior pools) on v3live band n=247
killedv3-mc5/walletsanalyst1discard16 Oct 06:16
#600Run 17: LP add/remove before +300s entry on v3live band (n=247)
killedv3-mc5/walletsanalyst1discard16 Oct 06:16
#599Run 17: v3live born<100k pools, buy at first print >=1M, stop .5, 4h, by time-to-1M tercile (n=133)
killedv3-mc5/regimeanalyst1discard16 Oct 06:16
#598Run 16 Oct: creator sequence lens (births.db v3 creator x v3hist, cap@300s>=100k): first launch 3x-rate 35% (n=667) vs 2nd 3% (n=63) vs 3rd+ 7% (n=153). Confirms live SERIAL flag (#305/#349); no new arm. Discard as duplicate.
killedv3-mc5/walletsanalyst2discard16 Oct 05:38
#597Run 16 Oct: name-clone lens (token name equal to an earlier pool's name in tokmeta, v3hist cap@300s>=100k, n=276 named): clone of a prior 1M runner n=67 reaches 1M 82% vs unique 58% but 3x-after-300s identical 25.4% vs 25.4%; name coverage only 209/8954 pools. No entry edge, discard.
killedv3-mc5/launch-metadataanalyst2discard16 Oct 05:38
#596Run 15 Oct 20:00 UTC: buy-size lens on v3live band (cap@+300s 150k-1M, n=151): round-ETH-amount share and buy-size CV in first 300s vs P(3x within 1h). Round share low/mid/high P3x 36/24/20%, CV terciles 34/28/18% vs base 26.5%; adjacent CIs (+-7pt) overlap, n too small, fin mean 1.2-1.7 all below fee-adjusted need. Dup of #436/#432 (done on v4 only). Discard.
killedv3-mc5/first-minute-filtersanalyst2discard16 Oct 03:03
#572Run 114 (13 Oct): order-flow concentration lens on native nohook seed-25k kit (v4tape pre-touch, n=72 pools with >=8 pre-300k swaps): distinct senders before the 300k touch <=5 -> 22/44 (50%) reach 1M vs >5 senders 5/28 (18%); holds in both time halves (9/19 vs 1/17, 13/25 vs 4/11), Fisher p~0.005. Gate candidate for #561 kit-300k entry; survivor-biased (pools with pre-touch tape only), no stop paths yet
skip kit-300k entries with >5 distinct swap senders before touch		killedunhooked-v4/blind-spotanalyst3discard16 Oct 02:48
#595Run 16 Oct: first-60s price impact per ETH (v3live, n=1105) re-check of #396/#530: lambda quintile 1 P1M 39% vs 0-1% rest, but it is pure volume (median 60s volume 21 vs 0.25-3.9 ETH); max-mult after +60s flat 1.55-2.2 in all quintiles. Duplicate, no edge beyond volume.
killedv3-mc5/first-minute-filtersanalyst2discard16 Oct 00:21
#593Keep mc5 cap_out at 1M/5min: SEALS case shows loosening buys spike tops (V3-R300 0.35x)
killedmissanalyst1killed15 Oct 08:20
#501V2 factory 0x8bce (1-ETH 1:1 seed kit): 3 real runners in 34h (INCOGNITO 1.2M, Ollie 1.05M, DOPE 2.7M), radar 5M-20M crossings = LP-drain artifacts
Watch PairCreated of v2 factory 0x8bceaa40b9acdfaedf85adf4ff01f5ad6517937f (WETH token0, ~1500 pairs/day). Entry: first Sync with price mult m>=8 vs first Sync AND quote reserve q>=2 ETH (honest buyin		killedunknown-venue/v2-factoryanalyst332killed15 Oct 07:10
#587Run 169: v4tape HL3000 class (hookless fee3000, 32 pools since 12 Oct, native+weth) vs live_paths: native ones cap_5m<=55k (PIVOT 61k max), weth ones glitch caps (8e9, 0.03 USD depth) -> no runner, no exit depth. Too rare/thin
killedunhooked-v4/blind-spotanalyst1discard15 Oct 03:51
#585Run 15 Oct: sell-count lens on v4tape native fee100 nohook (entry 12x first print): only 21 qualifying pools, all ns>=20, mean 1.125 -> too thin, wait for tape extension
killedunhooked-v4/blind-spotanalyst2discard15 Oct 02:55
#584Run 15 Oct: pre-entry sell count + entry-print size on v3hist (cap>=250k within 600s, nb>=3, n=511, tp5 sl.5 slip7): ns<5 1.81 vs 1.64 base (halves 1.96/1.69); ns+size>=1ETH combo 1.88 (2.04/1.77) but skipped trades still mean 1.18 -> filter removes +EV trades
parkedv3-mc5/first-minute-filtersanalyst2info15 Oct 02:55
#582Run 14 Oct: ETH 6h/24h return before launch vs P(5x/3x after +300s) on v3hist band n=276 (12 Sep-6 Oct): terciles 24-% P5x 0.22/0.24/0.15 and 6h 0.24/0.20/0.17, within sampling noise (+-4pt), peak-only; BTC/ETH-dump hypothesis not supported
killedv3-mc5/regimeanalyst2276discard15 Oct 00:59
#581Run 156: swap inter-arrival clock (gap CV, same-block share, sender/swap ratio, duration, n swaps) before a 3x trigger on v4tape A1 (native nohook) pools, n=217
Bot-like regular flow before the 3x print predicts a continuation to 2x+		killedunhooked-v4/blind-spotanalyst1discard15 Oct 00:01
#577Wave detector: 10 native fee100 nohook runners on 13 Oct all launched with cap_first ~25k/depth ~87k (one kit, one wave); buy the 2nd-3rd pool of an identical-signature wave
10 of 12 mc5 NOT SEEN 1-2M runners on 13 Oct share venue+cap_first+depth, so they are one operator wave; earlier native fee100 lenses (#138,#140,#141) tested pools singly and sellability was 35% (#469		killedmissanalyst2discard15 Oct 00:01
#545STOPLIQ: mc5 stop 0.5 -> hold with stop 0.1 only if pool WETH >=0.35x entry WETH at the stop print (rug has ~0, shake-out 0.6-0.8): v3hist n=94 mean 2.11 vs 1.34 (halves 3.19/1.03 vs 2.22/0.46), lo10 1.41 vs 0.79; 8/59 live stops later ran >=5x
killedv3-mc5/exitsanalyst294killed14 Oct 08:30
#576v3 first-block bundle (bb>=3) entry <=45s, cap<=100k
3+ buys in first block: 30% reach 1M vs 4%. Replay TP5/SL.5 n=225 mean 1.32 but carried by launcher 0x7399 whose replay is artifact (bb<=2 shows 71% wins); outside it mean 1.01. Needs pool depth + eth		killedmissanalyst3discard14 Oct 07:55
#575Run 14 Oct: hot-chain clock (pools born prior 2h with cap_5m>=300k, known at entry) vs P1M on live_paths band n115: terciles 37/26/33%, halves flip: discard
killedscout/regimeanalyst2discard14 Oct 07:31
#304A3-NH100: nohook fee-100 pools, enter 300k touch, tp2.5 stop.6
replay tp2.5 1.35x pulls 1% n=421; forward arm asked #734		killedanalyst3killed14 Oct 07:10
#212WETH-hook 30m confirm entry cap 250-600k, hunt 3-10M (pool_paths #586)
killedanalyst1discard14 Oct 07:10
#339A3-DIP: v3 band resting buy at -20% below +305s price, 30 min window, tp5 sl.5 4h: 1.34 vs 1.03 same pools (n239, halves 1.50/1.20)
buy the dip after the signal beats buying the signal		killedv3-mc5/5min-capanalyst3discard14 Oct 04:55
#569Run 14 Oct: funder-sibling reputation (ffund_snap + deployer.db x live_paths, cap_first<150k, n=1094 pools, funder known for 234): walk-forward, deployer's funder already funded a prior deployer: 130 pools 0 reached 1M; fresh funder 21 pools 0; prior non-runner 83 pools 1. Serial-farm funders produce no runners, but base rate too low (1/234) to carry a skip rule; coverage 21%
fund-cohort skip/pick has no power at 21% coverage		killedwalletsanalyst2discard14 Oct 04:45
#568Run 14 Oct: half@5x + trail 50% from peak (v3hist band n=291, glitch prints filtered) 1.40 (1.42/1.38) vs tp5 1.21 (1.38/1.04)
Replay-only gain +0.19 from keeping a runner remainder; but live real-path ratchets (#169) and FAST-LADDER (#296) showed worse than fixed tp5 and thin-pool sells at 5x+ fail. Discard: no new evidence 		killedv3-mc5/exitsanalyst2discard14 Oct 04:34
#566Run 140: scheduled-launcher cadence (minute-of-hour, on-5-min-mark, UTC 2h block; pool_paths born<150k n30814): P1M 1.8-2.7% every bin, on-mark 3.3% (n490) vs 2.3%. Flat, discard.
killedscout/regimeanalyst1discard14 Oct 04:11
#565Run 140: round-number cap magnetism (live_paths max_cap n14656): 0.95-1.0M 22 pools vs 1.0-1.05M 23; 2.85-3.0M 19 vs 3.0-3.15M 19. No stall below round caps, no limit-sell edge (2.0-2.85M spike n1714 = born-big kit, known).
killedscout/regimeanalyst1discard14 Oct 04:11
#564Run 139: symbol-case/name-shape lens (names.db x live_paths, v4 native+usdg nohook is_first cap_5m 100k-1.5M, n1633): mixed-case symbol P1M 0.2% (1/614 native) vs 1.6% (16/990); digits, non-ascii, 3+ words, inu/coin words all ~flat. Skip-filter only drops 1 of 17 runners, no price edge in this cell (killed #525). discard
killedunhooked-v4/blind-spotanalyst1discard14 Oct 03:56
#562Run 106: block-0 kit features (b0_creator, b0 recipients, struct) vs P(300k/1M) on kitlive: creator block-0 buy lower 300k rate but weak; recip=0 lift = one-day batch
b0_creator skip		killedv3-mc5/launch-metadataanalyst3discard14 Oct 03:42
#560Run 138: v4sender.db (9.9 GB, swap_sender, blocks 45M-87M) is the Swap-event sender = router, not tx.from: 428 distinct senders in 20k blocks, top 8 are aggregators (8.5k of ~20k swaps). No wallet identity, so no early-buyer/insider lens possible; #323 (tx.from) still needs data
killedwalletsanalyst1discard14 Oct 03:39
#558Run 16 Oct L4: native/usdg nohook cap_5m/cap_first jump groups vs real runner
3x jump by 5 min picks real v4 nohook runners		killedunhooked-v4/blind-spotanalyst2discard14 Oct 03:20
#557Run 16 Oct L3: walk-forward smart early buyers (wallet in >=2 prior runners' first 120s, hit share>=30%) as entry flag on v3live
Pools with >=2 smart early buyers in first 120s run		killedv3-mc5/walletsanalyst2discard14 Oct 03:20
#556Run 16 Oct L2: launch crowding (other first-pools within +-300s) and time of day/weekday vs real 1M+ (depth>=20k)
Crowded launch windows dilute runners		killedv3-mc5/regimeanalyst2discard14 Oct 03:20
#555Run 16 Oct L1: hook-WETH graduation pool (sibling v4:weth born ~1.9M, lag 200-670s) as entry
Buy the no-hook WETH pool a hook pool graduates into		killedunhooked-v4/blind-spotanalyst2discard14 Oct 03:20
#544Run 130: sellcross.db (never used in ledger) says hooked H3000 WETH pools are NOT sellable at the 100k/300k cap cross (0/43 ok, revert 0x4713c18b); native ok but returns ~1e-6 ETH per 0.02 in. Contradicts birth probes (roundtrip WETH 5/5 at 5 s, n tiny). Threatens #535/h1m fills.
If hooked WETH sells revert at the cross, h1m paper fills are fiction; sim artifact (settle to WETH) still possible		parkedhooked-v4/sellability-riskanalyst1info14 Oct 02:15
#554Run 15 Oct L3: v4 nohook native/weth/usdg pools with cap_first>=150k and depth>=20k at 300k reach 1M 94% (n=572)
killedunhooked-v4/blind-spotanalyst2discard14 Oct 02:03
#553Run 15 Oct L2: routed-flow share (recipient != signer) in first 120s, v3live 1411 pools: top quartile P1M 21% vs <2%
killedv3-mc5/walletsanalyst2discard14 Oct 02:03
#552Run 15 Oct L1: round-trip share of sellers, dip absorption (price 3 swaps after a sell), pool-create->first-trade delay, LP adds in first 5 min vs P1M on v3hist band (n=289)
killedv3-mc5/first-minute-filtersanalyst2discard14 Oct 02:03
#550Run 14 Oct: live mc5 rug forensic q_lower/sell_gap separate rugs perfectly but are sell-time fields (leakage), not buy-time
killedv3-mc5/walletsanalyst2discard14 Oct 01:32
#549Run 14 Oct: UTC 20-04 vs day on v3hist band replay: night 2.07 vs 1.17 (3 of 3 thirds same sign) but LIVE night 0/15 mean 0.36 vs day 1.9; contradicts, same as #457
killedv3-mc5/regimeanalyst2discard14 Oct 01:32
#548Run 14 Oct: dip-reclaim re-entry (price <=0.5x then >=0.8x, v3hist band n=97): mean 1.33-1.72 vs base 1.2-1.4, no edge over baseline
killedv3-mc5/exitsanalyst2discard14 Oct 01:32
#547Run 14 Oct: v4 nohook first-mint tick range (liqlock ml, 27.9k pools): range width 1071488 = single-sided seed kit holds all real-depth runners (70/3481); inside: cap_5m>=3x first and depth5m>=50k 61/834 reach 1M but median max 1.2M, 0 reach 3M; same as #529 ceiling
killedunhooked-v4/blind-spotanalyst2discard14 Oct 01:32
#546Run 14 Oct: GeckoTerminal 60s metadata re-cut n=305 v3 (580 rows): socials only on 2 pools (2/2 ran), gt_score>=40 4/4 but n tiny; holders null 297/305; too sparse
killedv3-mc5/launch-metadataanalyst2discard14 Oct 01:32
#543Run 128: DS paid boosts (dsboosts.db) vs runners: 18 tokens, 12 in live_paths, boost lands 13 min+ after launch, 2/12 hit 1M (base 18%); no lead, tiny n
killedscout/regimeanalyst1discard14 Oct 00:35
#542Run 126: DS paid tokenAd / communityTakeover orders (dsorders.db) vs runners: takeover 30 tokens (7 in live_paths, 5 reached 1M, all old coins already ran; CTO is filed after the run), tokenAd 13 tokens (4 in paths, 0 reached 1M). No pre-run lead, no entry signal
killeddead-universesanalyst1discard14 Oct 00:01
#493A3-SEEDHI: v3 fee100 seed<12k, cap5 1-3M, tp5 +300s
seed-small pools above mc5 band 1M cap run: crude hit5 41% n27 vs 0/20 seed-high		killedseedanalyst3killed14 Oct 00:00
#537Run 124 SEEDHI-WIDE: v3 seed<12k, cap5 0.5-3M, wide stop 0.2 + tp20 + 4h: replay 7-12 Oct n=39 mean 2.9-3.3 (ex top3 1.85-2.2; halves 3.5/2.3) vs stop 0.5 1.7 (ex top3 .67); runners dip <0.5x before 21-31x
A3-SEEDHI stops (12/14) cut the runners that dip first; stop 0.2 tp20 hold 4h		killedseedanalyst1killed14 Oct 00:00
#235Seed-cap lens on mc5 band: first-print cap <7.8k (fresh bonding seed) tp5 sl.5 1.37 n=147 (halves 1.47/1.25) vs 0.82 n=143; born >=390k 0 of 73 hit 5x
band entries whose first print was a tiny seed climb for real; born-mid/big are arrival-price jumps		killedv3-mc5/first-minute-filtersanalyst3147killed13 Oct 08:05
#540Run 124: v3 depth at first cross separates nothing in seed kit (kit-fixed 32-67k); band cf>=12k cap5 250k-3M pools are dust (rug rule), replay 0.28-0.36
killedv3-mc5/first-minute-filtersanalyst1discard13 Oct 07:43
#539Run 124: hot-name copycat (same ticker as a 1M+ organic runner in prior 24h, live_paths 11.3k pools): 2/208 runners vs 51/11069 base; n too small
killeddead-universesanalyst1discard13 Oct 07:43
#538Run 124: v4 native nohook seed-25k kit cap_5m 100-150k bucket (live_paths 7-13 Oct): 6/46 reach 1M, EV 1.32 but n small, overlaps #529 sample (6%); plateau 1.0-1.3M
killedunhooked-v4/blind-spotanalyst1discard13 Oct 07:43
#157mc5 band: price momentum c5/c60>=1.1 AND pullback c5/peak5 in (0.7,0.9]: tp5 sl.5 slip7 2.35 n=47 hit5 43% (halves 2.56/1.99) vs 0.94-1.12 for the rest of 226 dust-guarded pools
killedv3-mc5/first-minute-filtersanalyst347killed13 Oct 07:31
#39V3-005 first buy 0.05
killedv3-fee100/first-buyanalyst1killed13 Oct 07:31
#483H3000 late entry at first 300k crossing (age>=10 min, depth>=80k): P1M 92% n107, 300k->1M median 42 min, no jump
killedhooked-v4/short-tpanalyst2killed13 Oct 07:20
#291FAST-LADDER-2M: v3 enter 45s after 2M cross if 1M->2M gap 5-60min, sl.5 tp5: 1.78 n=126 (Sep 1.68, Oct 1.86) bs5 1.49
killedv3-mc5/late-runneranalyst2126killed13 Oct 07:20
#13BUSY60 unhooked fee-100 one-sided early flow (up>=41, down<=6)
killedunhooked-v4/first-minute-filtersanalyst287killed13 Oct 07:10
#534DeFade rug features: dev transfer to fresh wallets, funder chain to bridge, volume/mcap vs unique wallets
These deployer/funder features predict pulls beyond LP-owner check		parkedscout / rug-featuresscoutinfo13 Oct 07:05
#533Counterfeit stock-token quote pools = fake caps; gate on quote mint/burn
Pools quoted in fake/low-issuance stock tokens inflate caps and rug; mint/burn count check removes them		killedscout / stock-quotescoutdiscard13 Oct 07:05
#529Run 117: native fee100 nohook seed-27k kit in pool_paths: 874 pools cap_first 26-27.5k, depth ~90k at 300k; 130 touch 300k, 21 reach 1M (2.4%), 0 reach 3M; 1M odds by cap_5m 0/1.7/6/7/36% (<60k..>250k). 300k entry to 1M 16% x3.3 vs ~0.5 loss = ~0.95 before fees
kit is a 1M-ceiling source, no 3M+; discard		killedunhooked-v4/blind-spotanalyst1discard13 Oct 04:56
#523Run 77: A3 band flags overstate: SEED tm-harness replay 0.90 vs my sim 1.43, forward 0.42. Rule: a band filter matures only if tm harness replay >=1.3 on an unfitted window.
killedv3-mc5/first-minute-filtersanalyst3discard13 Oct 02:21
#240A3-SEED: band + first print <7.8k + 5-min vol>=49k: 1.43 n=136 (1.59/1.26) vs 0.82 rest; testmgr #621
killedv3-mc5/first-minute-filtersanalyst3136killed13 Oct 02:20
#520Run 107: climb-speed dwell (300k->1M, radar 1M->2M) vs next tier: v4 dwell<120s P3M 76-92% is first-print jump (t300k=t1m at birth); v3 <120s 61% vs 17% slow = seed kit #500/#492; radar 1M->2M <15min 5/5 reach 5M vs 3/18 slower but n=23, no entry path. Discard
killedscout/regimeanalyst1discard13 Oct 01:31
#519Run 106: pool_paths 31k pools, cap5m 150-600k born<150k: v4 usdg nohook 0/144 real (depth t1m>=10k), native 11%, weth 13% (senders/depth ratio/cap_first non-monotone); v3 depth/cap ratio low tercile 44% vs 11% = seed-cap lens #235; discard
killedunhooked-v4/blind-spotanalyst1discard13 Oct 01:12
#518Run 106: hot-hand regime (runner 1M in prior 1/3/6h vs P1M of 300k touchers, seen_peakcap n=500): 33-44% flat, no gating signal; discard
killedscout/regimeanalyst1discard13 Oct 01:12
#517Run 106: bundled-at-creation hooked WETH (seen_peakcap, n=1046 ok): bundled P1M 43% vs 14%, but born at 300k (age 0): entry after birth flat cap_5m->15m/30m/60m med 1.0 (n8 hooked, 60 v3), nothing buyable; discard
killedhooked-v4/first-minute-filtersanalyst1discard13 Oct 01:12
#516Run 104: radar 1M->3M continuation v4 24/55, v3 4/14 (survivorship, no EV; covered by #74/#291/#479/#483)
killedhooked-v4/short-tpanalyst1discard13 Oct 00:41
#492v3 1M-crossing chaser: depth>=20k at 1M, born<150k: P3M 55% P5M 40% P10M 30% n53 vs baseline 39/29/18 n218 (peak-based)
killedv3-mc5/first-minute-filtersanalyst2killed13 Oct 00:15
#515Run 102: ticker length vs P1M (seen_peakcap + names.db, n=1046 ok-cap): 2-char tickers 36% vs 18% base, but it is the SI/AI/HI serial-ticker family artifact (#100); not new
killeddead-universesanalyst1discard13 Oct 00:01
#18FAST5 enter 5 s after first swap (ramp bot pools)
killedunhooked-v4/first-minute-filtersanalyst2killed12 Oct 20:45
#279Unhooked WETH fee3000 runners (24 of 131): all depth 0 decoy twins
killedunhooked-v4/blind-spotanalyst2discard12 Oct 20:35
#2625-Min Cap band replay on v4 nohook WETH/native/USDG (pool_paths 4 wk): P2x .39/.43/.26 vs v3 .58, supply 2/day
killedunhooked-v4/blind-spotanalyst157discard12 Oct 20:35
#257Hazard: 300k entry with no 1M after 2h has P1M<=8% (v3 .05, nohook .01-.07): 2h time stop
killedunhooked-v4/blind-spotanalyst3discard12 Oct 20:35
#514gridfwd forward arms LIVE/E3M/B2M (+trail T5x/T10x) on tm grid replay
killedv3-mc5/replay-gridtestmgrkilled12 Oct 16:20
#513wlink copy-exit: sell when signal winner wallets sell
copy winner exit beats tp5/stop.5		killedv3-fee100/wallet-signalresearchkilled12 Oct 14:48
#512wlink: follow only top-20% hit-rate winner wallets / K2
killedwalletanalyst2killed12 Oct 14:45
#454Capacity and cost robustness of #435 late-sell arm (position size, entry lag, cost)
Arm survives realistic size, lag, cost		killedunhooked-v4/first-minute-filtersanalyst1748killed12 Oct 13:56
#435Late-sell inversion: skip nohook pools whose sells >44.8% of swaps in 200-300s; entry +315s tp3 stop.5, pull-aware sim
active pools with net selling at 200-300s rarely run; skipping them lifts late-entry mean		killedunhooked-v4/first-minute-filtersanalyst1killed12 Oct 13:56
#177Wide exits on band entries: tp10 sl.3 4h 1.67 (halves 1.78/1.57, 30 hits), tp20 2.15 (16 hits); ladders worse; sell depth unproven
killedv3-mc5/exitsanalyst3killed12 Oct 09:15
#265v3 late 1M entry (fair 1M milestone, age 12-90 min, stop .5, tp2): harness 88 trades mean 1.46/1.61/2.32 (OOS 6-10 Oct n=19), ex top2 1.31-1.40
killedv3-mc5/late-runneranalyst188killed12 Oct 07:15
#511Off-chain catalyst (Robinhood listing/CEO mention) runs old large caps
Large old tokens jump 16-400% within 1-3 days of Robinhood app listing/announcement; single-token evidence (CASHCAT), survivor bias		parkedscout / catalyst eventsscoutparked12 Oct 07:05
#505Run 80: new /opt/pricepredict-data/chain/b0NN.db (ev + liqfrom: tx.from of liquidity txs, per 1M-block batch, still building, w1.log crashed at k==13 parse). Plan when complete over blocks of nohook weth/native: liquidity-adder identity (tx.from) with earlier runner vs P1M, pre-registered 2x lift, >=10 runners per date half. Not run: batch incomplete.
liq adder with earlier 1M runner lifts P1M 2x		parkedunhooked-v4/blind-spotanalyst1info12 Oct 06:51
#504Launch-density (pools born chain-wide in prior 10 min) and birth minute-of-hour vs P1M, radar born<150k, 30.7k pools: flat on v3 (2.4-4.2%), nohook weth/usdg/native (no trend); only hooked weth falls with density (7.7%->1.1%, batch spam, H3000 known)
killedscout/regimeanalyst2discard12 Oct 03:35
#509Run 84 census non-v3 nohook 1M+ pools (3491): latest 40 are multi-M first-print jumps with depth<500 USD; no tradable unwatched runner
killedunhooked-v4/blind-spotanalyst2discard12 Oct 02:12
#508Run 84 kit-fingerprint scan (venue,quote,fee,cap_first) 23.9k pools: only high cells are hooked weth3000 seed 22-36k (90-100%, gated, retired) and v3 fee100 seed 2-5k (known); no new unwatched kit
killedunhooked-v4/blind-spotanalyst2discard12 Oct 02:12
#507Idle run 47 (12 Oct): organic 1M census 11 Oct = hooked WETH 6 (gated), v2 4, v3 3, native 2; live mc5 exits re-check (4 of 56 trades = 79% of return, ladders lose); hook registry = fixed insider list
Census and exit re-check only; no new rule. Hooked WETH stays gated (registry owner EOA lists wallets in batches), ratchet exits already killed in #169.		killedunknown-venue/dataanalyst3discard12 Oct 01:41
#499Radar+misses join 7-10 Oct: 25 of 28 organic missed runners are venue v3, not unwatched v4; unwatched-pool misses are jumps/H3000/phantoms
unwatched pool types hold the organic misses		killedunhooked-v4/blind-spotanalyst3discard12 Oct 00:03
#487skip_sellquote/skip_quote are gmgn quote API failures (401 x16, 429 ban x4, 400), not sell judgements; 16/25 skipped picks reached 1M+ vs 0/21 closed buys
parkedfml2/skip-sellquoteanalyst2info12 Oct 00:02
#498Momentum acceleration cap_15m/cap_5m>=2 (cap_15m<700k, born<150k), 23.7k pools pool_paths: P1M weth nohook 10% n184, native 12% n98, usdg 4% n199; v3 10%; hooked weth 39.5% n263 (H3000 cell, known) vs bases 1.3-4.3%
Late-momentum entry on unwatched nohook v4 beats base		killedunhooked-v4/first-minute-filtersanalyst2discard12 Oct 00:01
#3H3000 tp5 first replay (pull-gated)
killedhooked-v4/tp-ladderanalyst2112killed11 Oct 16:30
#2H3000 hooked pools, tp ladder / time exit
killedhooked-v4/tp-ladderanalyst2110killed11 Oct 16:30
#142Unique-hook fee100 WETH (0x..144 per-coin hooks) re-check: +90s tp2/3, pull=0
H100 family still pays on 399 pools		killedhooked-v4/short-tpanalyst2399killed11 Oct 16:30
#11H100 LOWFDV filter (FDV<4.63 ETH)
killedhooked-v4/short-tpanalyst253killed11 Oct 16:30
#8H100 hooked fee-100 (0x144 hook bits)
killedhooked-v4/short-tpanalyst2433killed11 Oct 16:30
#38V2DEC9 9-decimals kit
killedv2-ramp/kitanalyst1killed11 Oct 16:30
#37V2RAMP seed 1.000, buy +8s sell +300s
killedv2-ramp/kitanalyst1killed11 Oct 16:30
#245Hook e5e7..e044 native fee0 fast climber (age<=10m, 100k->300k<=4m): P1M 58% n=115 vs 35%, halves .59/.57, supply dying
killedhooked-v4/short-tpanalyst1115killed11 Oct 16:07
#348A3-BREADTH: >=30 distinct tx.from buyers in first 300 s on v3 band: 1.57 n111 vs .83; with seed&vol49 1.78 n79
breadth (IPO oversubscription) separates real demand from one-wallet volume		killedv3-mc5/walletsanalyst3111killed11 Oct 15:40
#389Aged-coin lane on v3 via capcross: pools crossing 300k after age >=6 h: n=6 of 284, untestable; aged runners live in v4 nohook native, not in capcross
killedv3-mc5/late-runneranalyst2killed11 Oct 15:40
#305A3-SERIAL: skip mc5 band entries whose create-tx sender launched an earlier pool: P1M 0/81 vs 20.6% (thirds 0/15,0/18,0/48), sim tp5 .77 n55 vs 1.23 n211, halves agree
killedv3-mc5/walletsanalyst3266discard11 Oct 15:28
#373Idle run 23: forward arms fwd n<=8, dsboosts 8; wait
killedscout/regimeanalyst3discard11 Oct 15:28
#342Loss streak math v3 band: expected longest streak median 22 (95% 35), 15 in a row is not a regime signal; Kelly 8%, half 4%
killedscout/regimeanalyst3discard11 Oct 15:28
#184Wallet age of first 20 buyers (data request ppdata #573)
killedv3-mc5/walletsanalyst3discard11 Oct 15:28
#149Later-pool lineage: nh_USDG pool of a token with earlier real runner 6.1% vs 1.5%
killedunhooked-v4/blind-spotanalyst1413discard11 Oct 15:28
#173Venue lifecycle: real runners/day since 3 Oct v3 8.9, USDG nohook 2.1 dying, hooked 0.3, native ~0; native nohook runners are 4-12 buyer script ladders <=2.1M
killedunknown-venue/dataanalyst1discard11 Oct 15:28
#92mc5 band concurrency (in-band pools +-30/120 min)
busy hours dilute attention, skip when many pools born		killedv3-mc5/5min-capanalyst2330discard11 Oct 15:28
#135v4 real climbers by launcher/tx_to: no unknown pocket (109 H3000, 55 0x58da nohook, 33 USDG)
killedv4-nohook-native/blind-spotanalyst3226discard11 Oct 15:28
#256A3-COPY: band symbol already had a 1M token: tp5 1.39 n=106 (1.47/1.22) vs 0.96; with seed+vol 1.79 n=55; freshness flat
killedv3-mc5/first-minute-filtersanalyst3discard11 Oct 15:28
#271DexScreener paid token profile (orders API, paymentTimestamp) on new births 6-9 Oct: P1M 28% vs 3.5% base; profile median 17 min after birth; entering on profile (30 s delay, v3 live tape) tp2 sl.5 1.13 vs control 0.96, peak P2 55% vs 29%, P3 28% vs 10%; n=84 tokens
parkedv3-mc5/first-minute-filtersanalyst3info11 Oct 15:27
#128Order-flow toxicity/router/metronome in mc5 band: flat; swaps<10 in 5 min 0.96 vs 1.37 (confirms #91)
parkedv3-mc5/first-minute-filtersanalyst3330info11 Oct 15:27
#146USDG nohook 300k-touch ride, glitch-masked: 1.3-1.6 (best half@3x+trail.5 1.64, ex-top3 1.42), run23 2.93 was a 1e4x cap glitch
killedstable-v4/first-minute-filtersanalyst159discard11 Oct 15:27
#497A1-V3R300: v3 first mcap<200k, reserve>=14 ETH and >=150 buys at +300 s
killedv3-mc5analyst1discard11 Oct 15:26
#185Tier ceiling: 0 of 29 real no-hook v4 (WETH 10, native 19) climbers reach 3M; v3 28%, hooked WETH 46%
unhooked v4 targets must be 1M tier; sell at 1M		parkedunhooked-v4/first-minute-filtersanalyst129info: unhooked v4 target 1M tier, sell at 1M; nh100 covers v4 nohook 300k flip11 Oct 15:26
#150H3000 hooked pools: birth hour 08-20 BKK real-runner 38.6% vs 25.5% night; 17/19 days
killedhooked-v4/short-tpanalyst1395no forward flow: hooked h3k retired (not tradable)11 Oct 15:26
#12H100 partial exits / silence exit / net-flow exit / trail
killedhooked-v4/dead-endsanalyst2retired 11 Oct Ben verify-or-kill: hooked pools not tradable (hook registry gate, no sellability)11 Oct 15:26
#488Day-of-week and fee-value lens (pool_paths 21 Sep-11 Oct, born<150k, nohook v3/v4 weth/native): DOW flat (v3 2.7-3.6%, weth 0.6-3.4% noisy, n<=600/cell); fee: weth fee3000 18/41 P1M but depth 0, 1-3 senders, one-print jumps (known decoys #279)
weekday or fee value picks runners		killedunhooked-v4/blind-spotanalyst1discard11 Oct 09:21
#486Time-of-day and prior-24h regime (radar pool_paths, born<150k): P1M rate flat by UTC third (v3 3.6/2.6/3.2%, nohook weth 2.1/1.8/2.2%, usdg 1.3/1.5/1.1%); v3 prior-24h known-outcome hit rate quartiles -> next P1M 5.1/3.1/4.5/2.3% (no persistence, n869 each). No pause/size switch possible.
regime persistence		killedscout/regimeanalyst2discard11 Oct 09:01
#482Causal launch burst (pools of any venue born in prior 60 s) vs P1M, born-small nohook weth/native/usdg + v3 (31k pools, pool_paths 11 Oct)
burst>=3 prior pools lowers/raises P1M		killedunhooked-v4/blind-spotanalyst1discard11 Oct 07:56
#479Hooked v4 born-small, 300k cross at age 5-60 min, depth>=20k: P1M 62% P3M 31% n209 (days stable); entirely fee3000 weth = H3000 family (83% P1M n149); fee100/native/usdg hooked 8-20%. depth@cross>=100k: 89% P1M. Confirms #192 (already running); no new arm
H3000 300k-cross entry		parkedhooked-v4/short-tpanalyst2info11 Oct 07:36
#478True crossing times (pool_paths t300k_s, depth_t300k, 11 Oct refresh): nohook t300k<30s depth>=20k looks 94% P1M/82% P3M n506 but 470 have cap_first>=1M (born-big artifact). Born-small (cap_first<150k) gradual nohook: P1M 14% P3M 0% n474 (halves 12/16); no t300k bucket >27%. DISCARD
nohook real gradual 300k crossers with depth reach 1M >40%		killedunhooked-v4/blind-spotanalyst2discard11 Oct 07:36
#474Pretag crew label (launch table 26.5k, 1988 joined to canon) vs P1M on v4 nohook
crew-labelled launches (pretag) run more often than unlabelled		killedv4-nohook-native/blind-spotanalyst1discard11 Oct 07:01
#466capcross refresh (10 Oct 22:55): 300k crossers, jump vs gradual P1M. v4 nohook gradual (cap<=450k at cross) 0/28 reach 1M, jump 52/68 (one print, no depth); v4h gradual swaps>5 1/9; v3 gradual 31% n128. Confirms #424, no new edge
killedunhooked-v4/blind-spotanalyst2discard11 Oct 06:27
#467Fresh radar 7-10 Oct born-low 300k->1M by venue: native nohook 3/87, usdg 3/13, weth nohook 0/2, H3000 5/10: no new pocket
killedunhooked-v4/blind-spotanalyst3discard11 Oct 06:26
#463Exit lens: depth at 1M crossing per venue (radar 8-10 Oct, 1492 crossings): v4 native/usdg/weth nohook 1M crossers are 94-97% first-print jumps (cap_first>=300k), only 17/242 native, 24/84 usdg, 3/11 weth have depth>=5k; hooked/stock all have depth
killedunhooked-v4/blind-spotanalyst2discard11 Oct 05:50
#46015-min persistence (cap_15m vs cap_5m) as entry-at-15min filter, v3 and v4 nohook, cap_15m 150-600k, P1M by held/faded
Coins that held or grew from 5m to 15m run more		killedunhooked-v4/blind-spotanalyst2discard11 Oct 05:36
#459Seller identity from token Transfer logs (free getLogs) for late sellers 200-300s, 237 pools
Late sells concentrated in one wallet / early buyers dumping predicts losers		killedunhooked-v4/blind-spotanalyst1237discard11 Oct 05:25
#457Band sub-cap (300-350/350-420/420-550/550-1000k at +300s) and UTC hour block vs tp5
sub-band or hour gives edge		killedv3-mc5/first-minute-filtersanalyst3350discard11 Oct 05:22
#456Crowding: >=4 other band entries in prior 60 min (causal) vs tp5 result
crowded hours lower P(5x)		parkedv3-mc5/regimeanalyst347info11 Oct 05:22
#455Regime: ETH/BTC 60m/4h return before entry vs mc5-band tp5 result (n=326)
ETH or BTC drawdown pre-entry lowers band tp5 mean		killedscout/regimeanalyst3326discard11 Oct 05:22
#453Regime overlay on #435 arm: births in prior 60 min as skip rule (canon vs births.db)
Skip the busiest-quartile chain hours		killedunhooked-v4/blind-spotanalyst1748discard11 Oct 05:17
#452Early (0-100s) and mid sell share and late-minus-early trend vs sim result, nohook weth/native
Sell-share trend over the first 300s adds to the late-sell filter		killedunhooked-v4/blind-spotanalyst1748discard11 Oct 05:17
#451Order-flow toxicity (VPIN analog): USD-weighted sell share 200-300s and last-60s share vs count share
USD-weighted sell share separates losers better than the count share used in #435		killedunhooked-v4/blind-spotanalyst11117discard11 Oct 05:17
#450Exit variants on late-sell arm (#435 set): trailing, half@2 ladder, tp2/5/10, stop .3/.7
A trailing or ladder exit beats flat tp3 stop.5 on nohook weth/native entry +315s		killedunhooked-v4/blind-spotanalyst1748discard11 Oct 05:17
#449Early path shape (late-growth share, cap_1m/cap_first) at +5min, cap5m 150k-1M, radar pool_paths: v3/v4hook/v4nohook terciles vs P1M: halves flip (v3 late-share lo 23->48%, c1/cf hi 58->49% while lo stays ~20%; nohook 11-15% flat)
killedunhooked-v4/blind-spotanalyst2discard11 Oct 05:05
#448As-of prior-pool count of creator (serial vs first-time) on v4 nohook weth/native/usdg by cap5m band: raw 'serial runs 56%' in weth 100-300k is dust (cap5m->max_cap 1e7-1e10, depth 0, one-print); with depth>=3000 no signal in <300k bands (0-9%)
serial kit creators run on weth nohook		killedunhooked-v4/blind-spotanalyst2discard11 Oct 04:50
#447Buyer kind mix (addr_kind proxy/infra) in first 300s: tx_from is EOA in 99.9% swaps, lens empty
bot contracts as buyers		killedv3-mc5/walletsanalyst2discard11 Oct 04:49
#438Early holder distribution (Transfer logs, free getLogs 4k-block range works): holders/top1/top10 at +300s vs hit on v3 band n197
A3-DISP: holder breadth/concentration at +300s predicts band winners		killedv3-mc5/walletsanalyst3197discard11 Oct 04:42
#445Holders at +300s (Transfer logs) on v4 nohook born-small: nh>=61 P1M 3.4-6.9% vs 1.4% but P(>=5x from entry) 8%
killedunhooked-v4/blind-spotanalyst3340discard11 Oct 04:42
#443Hot-hand spillover: organic 3M runner peaked in prior 60 min vs new-pool organic P1M
killedscout/regimeanalyst2discard11 Oct 04:34
#442Supply/decimals fingerprint (non-1e27 = kit marker) vs P1M; fee3000 supply1e29 hooks0 weth pools = full-range dust, 16/21 fake runners
killedunhooked-v4/blind-spotanalyst2discard11 Oct 04:34
#440Sparse-launch (<=3 same-cell pools in prior 60 min) vs organic P1M: v3 5.5 vs 2.1, hook weth 11.3 vs 1.5, within-day matched; no lift at +66s/+300s entry sim
parkedhooked-v4/first-minute-filtersanalyst2info11 Oct 04:34
#439ETH/BTC return before birth vs P1M per cell (23k born-small pools)
killedscout/regimeanalyst2discard11 Oct 04:34
#436Arrival-rhythm and buy-size dispersion (bot cadence) in first 300s on nohook weth/native
metronome bots vs human flow		killedunhooked-v4/blind-spotanalyst1discard11 Oct 04:15
#433USDG/SUSD nohook custom-fee-tier pools early activity and fee tier vs P1M (7491 pools)
killedunhooked-v4/blind-spotanalyst1discard11 Oct 03:58
#432Wash index / repeated buy sizes in first 300s on nohook weth/native
killedunhooked-v4/blind-spotanalyst1discard11 Oct 03:58
#431LP pull survival (liq_owner): pulled<15min vs never pulled, fee100 nohook
killedunhooked-v4/blind-spotanalyst1discard11 Oct 03:58
#430First-300s order flow (swapamt) vs P1M on v4 nohook weth/native born-small
killedunhooked-v4/blind-spotanalyst1discard11 Oct 03:58
#428Creator reputation (earlier pools >=24h old, same tx.from, any reached 1M) vs P1M on born-small (cap_first<300k), canon 29.9k pools
winner-creators repeat		killedunhooked-v4/blind-spotanalyst1discard11 Oct 03:36
#427Radar 8-10 Oct 1M crossings by venue: organic filter (depth>=5k, cap_first<300k) leaves 23 of 175; nohook v4 only 3 (JACK, TORCH, 1 usdg)
parkedunhooked-v4/blind-spotanalyst2175info11 Oct 03:31
#423Unknown-venue runners: all 23 born 7 Oct, 0 on 8-9 Oct = one-day batch, not a pool type
parkedunknown-venue/dataanalyst3175info11 Oct 03:28
#425e1-runners.csv cannot give base rates (controls P5M=0 by construction); aged>=6h 45% to 5M is sampling artefact
killedunhooked-v4/blind-spotanalyst3info11 Oct 03:28
#424capcross 414 pools at 300k: non-jump v4 nohook 0/16 reach 1M, hooked 1/9, v3 33%; big mass is first-print jumps
killedunhooked-v4/blind-spotanalyst3discard11 Oct 03:28
#421Pull predictors at entry (swaps in 300s, swaps in 60s, cap0, reserve) on momentum entries
killedunhooked-v4/blind-spotanalyst2392discard11 Oct 03:28
#420Momentum 1m->5m (cap5m/cap1m>=1.5) entry at +5min on v4 weth/native nohook born-small, cap5m 100-300k
parkedunhooked-v4/blind-spotanalyst2392info11 Oct 03:28
#419Inversion: n_senders_5m and depth_5m as skip filters on nohook pools with cap5m 50-500k
killedunhooked-v4/blind-spotanalyst24174discard11 Oct 03:28
#418Weekend / time-of-day birth of v4 nohook born-small pools vs P1M (pool_paths 14 Sep-9 Oct, n7348)
parkedunhooked-v4/blind-spotanalyst27348info11 Oct 03:28
#417Token bytecode (eth_getCode: template hash, PUSH4 selectors, size) vs P1M, 118 v4 nohook runners vs 272 non-runners
killedunhooked-v4/blind-spotanalyst2390discard11 Oct 03:28
#414Kit-pool concurrency (other template pools born in prior 10/30/60 min) as farm-saturation regime for ramp-pool entries
more concurrent kit pools = more pulls		killedunhooked-v4/blind-spotanalyst1347discard11 Oct 03:04
#413Aged smooth climbers on v4 nohook (r6 tape): runners that reach 1M after >=1h age
slow climbers can be entered at 300k		killedunhooked-v4/blind-spotanalyst1171discard11 Oct 03:04
#412Gradual cap-cross entry (150-300k, prior print >=80% of th, age<15m) on v4 nohook weth/native from r6 swap tape, 19 days, with LP-pull-aware exits
ramp pools climbing into 150k keep going: tp3 stop.5 mean 1.5 price-only		killedunhooked-v4/blind-spotanalyst1677discard11 Oct 03:04
#411Token mint-to-first-swap gap (creators.db ts vs canon t_first) vs P1M on v4 born-small
stealth pre-minted tokens (long gap) run more		killedunhooked-v4/blind-spotanalyst110683discard11 Oct 03:04
#410Token address vanity (leading/trailing zeros, repeated-char runs) vs P1M, born-small first pools v3/v4 weth/native/usdg (canon, 17k pools)
vanity/mined token address marks sophisticated launcher		killedunhooked-v4/blind-spotanalyst117442discard11 Oct 03:04
#408Cell calendar 5-day blocks: native hooked P1M 47.6% n307 on 22-26 Sep then 0-6%; hooked WETH peaks 12% 27 Sep-1 Oct; runners come in ~5-day kit campaigns
campaign decay		parkedunhooked-v4/blind-spotanalyst2info11 Oct 02:50
#406Gas price/priority fee of early buyers (lpowntape 10 Oct, 157 pools): flat 20-30M wei, priority null; no signal
gas urgency		killedv3-mc5/walletsanalyst2discard11 Oct 02:50
#405Tier velocity 300k->1M on nohook cells (non-jump): P3M native 18% usdg 14% weth 9% vs v3 36%, hooked WETH 46%; no gap bucket helps
velocity on nohook		killedunhooked-v4/blind-spotanalyst2discard11 Oct 02:50
#404Closing-line value 5m->60m on v4 weth/native nohook looks positive (+0.17 log) but 80% of pools silent 55m, stale caps; dead at 0.5 gives 1.03. Needs LP-remove data
positive CLV		killedunhooked-v4/blind-spotanalyst2discard11 Oct 02:50
#403Sibling pools of same token predict runners: look-ahead trap (siblings born after success; 1 in 2000 visible at +5m); sibling visible at entry no lift (v3 8/39 vs 77/317). Resolves #325 as look-ahead
siblings lift P3x		killedunhooked-v4/blind-spotanalyst2discard11 Oct 02:50
#400Radar_hist born-small (cap_first<150k) 300k crossers reaching 1M: v4 weth nohook 64/411 (15.6%, median 21 min), v4 native nohook 37/179, v3 weth 199/596; coarse thresholds, no 600k level so TP/stop sim impossible
weth nohook 300k entry; 15.6% reach 3.3x; need tape with sell depth		killedunhooked-v4/blind-spotanalyst1discard11 Oct 02:21
#399Launcher persistence (prior pool of same launcher hit 1M before this 300k cross) on radar.jsonl 8-10 Oct, 114 born-small pools: 0 qualify; launchers are kits/factories, 3-day window too short
launcher with recent runner raises next pool P1M		killedv3-mc5/walletsanalyst1discard11 Oct 02:21
#398First-sell latency and buy-size drift (median buy last vs first 100s) on v3 band n262: halves disagree
killedv3-mc5/first-minute-filtersanalyst2discard11 Oct 02:19
#396Kyle lambda (cap move per USD, last 120s of first 300s) on v3hist band n244: terciles 1.64/1.30/1.29 mean, halves 1.91/1.73/1.17 vs 1.06/1.47/1.14, low-lambda = low-volume proxy, not stable
killedv3-mc5/first-minute-filtersanalyst2discard11 Oct 01:57
#395Creator buys in launch block (a1feed early.jsonl, v3 fee100, 161 pools mcap60>=20k): P1M 40% (22/55) vs 4.7% (5/106); inside mcap60>=200k 21/48 vs 4/15; 16 of the 22 hits cross 1M at age <1 min (jump trap); confounded by born-high cap. Known bundle family.
killedv3-mc5/walletsanalyst1discard11 Oct 01:55
#393INFLOW OOS on v3hist reserve field: pool_eth_after only exists from Oct 6 (0 rows before), so only 68 of 348 band pools testable, all already in-sample
killedv3-mc5/first-minute-filtersanalyst2discard11 Oct 01:44
#391Bundle dump speed (sold60/sold300 from bundle_fund pool_early, 168 bundled v3 fee100 pools) vs P1M: sold60 flat (39/46/43%), sold300 lo 71% vs hi 21% but is outcome-contaminated (not known at 60s entry); bundle_usd tercile 18/41/70% = known (tm-bundle35k)
killedv3-mc5/walletsanalyst1discard11 Oct 01:41
#390capcross v4 USDG nohook: 63 crossings of 300k, 55+ have swaps_so_far<=2 and cross 1M/5M with 0 min gap: born-high jump artifacts, no climb to enter
killedunhooked-v4/blind-spotanalyst2discard11 Oct 01:23
#388Swap-count acceleration last 60s vs prior 60s on v3 band: 1.29 n37 vs 0.92 (p .04), weaker proxy of INFLOW
killedv3-mc5/first-minute-filtersanalyst2discard11 Oct 01:23
#386Init LP size (canon init_liq_quote quintiles) vs P1M on v4 native/weth-hook first pools: runners n<=5 per cell, no signal; weth nohook/v3 rows lack the field. Open lead #1 from run 31 stays untestable until canon refresh (stale 9 Oct, births.db ends 7 Oct).
killedunhooked-v4/blind-spotanalyst1discard11 Oct 01:01
#385Deep funder walk (8 hops) on u100 creators: common origin?
u100 creators converge to few roots		killedwalletsanalyst2discard11 Oct 01:00
#381Quiet-chain regime: <=14 new v3 pools in prior 60 min -> band entries 0.67 n45 0/45 5x vs 1.22 (Oct 6-10, halves .70/.65) but OOS Sep13-Oct5 flips (Oct1-5 lo 1.48 vs hi 1.02; Sep lo 1.26 vs 1.64)
parkedscout/regimeanalyst2mature11 Oct 00:30
#370VWAP premium: cap at +300s above first-300s usd-weighted avg cap, v3live band (n=129 pools, 64 usable): tp5 sl.5 mean 1.17 vs .89, both time halves agree, bootstrap diff CI -0.05..+0.64
closing-line analogy: price above volume-weighted fair value = buyers still winning		parkedv3-mc5/first-minute-filtersanalyst2info11 Oct 00:22
#382Sniper distribution into entry: share of sells in 185-305s from wallets that bought in first 60s; and sniper buy share, n=129 band: mean .84-1.17 no monotone, hit5 counts 0-6
killedv3-mc5/walletsanalyst2discard11 Oct 00:22
#380GeckoTerminal 60s metadata (gt_score, holders, socials, honeypot flag) vs outcome, n=65: socials/holders empty at 60s, gt_score is pool-size proxy
killedv3-mc5/launch-metadataanalyst2discard11 Oct 00:22
#379LP pull timing: first big v3 LP remove age is monotone decaying (574 in <1m of 938), no fixed-age cliff; post-entry pull hits 16/129 band entries (0.58 vs 1.09), none by creator
killedv3-mc5/regimeanalyst2discard11 Oct 00:22
#378Time stop / flat-stop on v3 band (T 15-120 min, mid-check 15-60 min at 0.8-1.5x), 130 pools: all <=1.05 vs base 1.02
killedv3-mc5/exitsanalyst2discard11 Oct 00:22
#377Radar 10 Oct census of 1M crossings (150): real organic runners (cap_first<300k, depth>=5k) = v4 weth hook 5, stock hook 5 (token/token artifact), v2 2 (ramp kit), v3 5, native nohook 1 (JACK), usdg nohook 1; Mouse 20M weth nohook born 24.8M age 0 = jump
parkedunhooked-v4/blind-spotanalyst2info11 Oct 00:22
#376pp3 nohook v4 post-500k slot features (sell>buy, holders, lp removal): 'sell>buy' 3x 25% vs 6% is entry at collapsed cap1 (7k-200k) rebound, no depth
killedunhooked-v4/blind-spotanalyst2discard11 Oct 00:22
#375Speed 1M->2M predicts 2M->3M: <1024s 77% vs 60% (n105), in-sample
parkedv3-mc5/late-runneranalyst1105info11 Oct 00:16
#372Token name shape (length<=5, space, digit) on v3 band via tokmeta: n 2/12/0, no power
killedv3-mc5/launch-metadataanalyst2discard11 Oct 00:02
#371LP add/remove in first 300s on v3 band: add n=2, remove n=3, too few (v3live has first_trade_t on only 1948 of 8996 pools)
killedv3-mc5/walletsanalyst2discard11 Oct 00:02
#369Custom initial price on nohook v4 (weth/native/usdg): 1.3-2.2% vs base 1.2-1.8%, halves flip. Fee-tier map re-check: only v4 weth hook fee3000 (84%, H3000) converts. DISCARD
killedunhooked-v4/blind-spotanalyst1discard11 Oct 00:02
#368Custom (non-kit) initial price on v4 weth hooked: unique-px pools 12.2% P1M n=98 (12 runners, 74 creators) vs 4.3% base; modal kit price 0/637. Halves 16%/9%. Confounded by low cap_first; thin
A human-set start price (not a kit default) marks a real launch		parkedhooked-v4/first-minute-filtersanalyst1info11 Oct 00:02
#365Forward BREADTH check, 186 kitenrich rows: 9 band pools, 7 with buyers300>=30 all lose; n too small
parkedv3-mc5/walletsanalyst3info10 Oct 23:11
#364Tape shape on v3 band (top3 print share, sell fraction, signer round-trip, last-5 direction, last-60s change): non-monotone
killedv3-mc5/first-minute-filtersanalyst3discard10 Oct 23:11
#363Last-print age at +300s: >10s stale 0/90 reach 5x vs 13.8% fresh, but is a VOL5 proxy (101/106 stale have vol5<49k)
killedv3-mc5/first-minute-filtersanalyst3discard10 Oct 23:11
#362v2 radar crossings SI and CHOP (born 10k, 1M in ~3000 s, depth 58-105k, launchers 0x9a8c, 0xdb73): same scripted-ramp v2 kit as V2RAMP (21 fixed wallets, t3 #716); v2 CRAWL/MEME thin; v4:susd AAPL = stock artifact
killedv2-ramp/kitanalyst3discard10 Oct 22:47
#361BTC/ETH 1h/6h/24h return at entry vs band result (252 pools, terciles): non-monotone, hit5 7-18% no pattern
killedscout/regimeanalyst3discard10 Oct 22:47
#360Block concentration lens (swaps per block, max swaps in one block, top-block buy share, supply) on v3 band 277: swaps/block>=1.4 1.34 n105 vs .98 (1.44/1.17 halves); inside vol49 1.42 vs 1.13; corr with vol5 .41; supply constant 1B
parkedv3-mc5/first-minute-filtersanalyst3info10 Oct 22:47
#359DexScreener boosts: only 6 boosts in 304 polls since 20:20 BKK; no test possible, boosts are rare
killedv3-mc5/first-minute-filtersanalyst3discard10 Oct 22:47
#358Forward check of BREADTH (kitenrich buyers300>=30) on 5 forward band pools 9-10 Oct: max multiples 1.7/1.4/3.0/2.0/1.1x, 0 hit 5x; n too small
parkedv3-mc5/walletsanalyst3info10 Oct 22:47
#290A3-DISP: skip mc5 entries with top10 holders <30% at +5min or >=5 pre-launch transfers
dispersed supply = no runner; skip lifts band mean 1.12->1.22, n=280, halves agree		parkedwalletsanalyst3needs-data10 Oct 22:21
#356USDG e044 fee0 hook kit: canon proxy like v3 but tape = wash, sim 0.86 hit5 0
stable-quoted 044 kit is a v3 clone pocket		killedhooked-v4/short-tpanalyst3discard10 Oct 22:21
#355Runner autopsy v3 band: live flag union catches 100/120, missed 20 are born-high or dead-air bloomers
flags miss enterable runners		killedv3-mc5/late-runneranalyst3info10 Oct 22:21
#354Hot-coin regime (earlier pools alive in prior 60 min): P1M 3% -> 11% monotone on v3 band but entry return not (hot<2 0.85, hot>=5 0.90)
many live coins raise runner odds		parkedv3-mc5/5min-capanalyst3info10 Oct 22:21
#353Buy-size homogeneity + weekday/hour-of-week on v3 band: non-monotone, halves disagree
uniform bot buys or weekday change band return		killedv3-mc5/first-minute-filtersanalyst3discard10 Oct 22:21
#352Hook-bytecode family map of hooked v4: only H3000 campaign converts; rest born-big or batch jumps
unwatched hook families hold enterable runners		killedunhooked-v4/blind-spotanalyst3discard10 Oct 22:21
#351Born-high v4 nohook (cap_first>=500k): median max/first 1.00, 3-6% do 5x: pinned jump set; born-low unwatched P1M 1-2%, no pocket
killedunhooked-v4/blind-spotanalyst3discard10 Oct 21:40
#350Router mix / proxy-wallet share first 5 min on v3 band: flat
killedv3-mc5/walletsanalyst3discard10 Oct 21:40
#347Request ppfeeds: sell-probe launcher 0x017c60 USDG fee-1 pools
killedunhooked-v4/blind-spotanalyst3confirmed10 Oct 21:15
#341USDG fee-1 launcher 0x017c60 (AAPLx3L SI Muse): born 500k, pinned exactly 2.0M, fake-runner kit; radar false misses
killedunhooked-v4/blind-spotanalyst3discard10 Oct 21:05
#346Band depth at 5 min $36k same for 5x hitters and rest; stake <=1k impact <3%
parkedv3-mc5/exitsanalyst3info10 Oct 21:05
#345Late bloomers all-venue: radar has no true-age history, untestable
parkedv3-mc5/late-runneranalyst3info10 Oct 21:05
#344Winner-linked early buyers at matched cap: lift is cap level / H3000 family, no entry rule
killedv3-mc5/walletsanalyst3discard10 Oct 21:05
#343H3000 climbers P5M by seed tercile 9/29/33% (n35 each, in-sample)
parkedhooked-v4/short-tpanalyst3info10 Oct 21:05
#340CJK/non-ASCII name script: n=40 10% vs 5.7% P1M, not significant
killeddead-universesanalyst3discard10 Oct 21:05
#338LP range width at creation (Mint log): 92k vs 1.09M ticks hit5 40% vs 25% n=92, 2 presets
parkedv3-mc5/launch-metadataanalyst3info10 Oct 20:47
#337Unwatched nohook v4 census 7-10 Oct (215 1M crossings): 1 of 38 non-jump, stock artifacts again
killedunhooked-v4/blind-spotanalyst3discard10 Oct 20:47
#336Repeat-buyer share first 5 min on low-seed v3 band: flips with activity
killedv3-mc5/walletsanalyst3discard10 Oct 20:47
#335Launcher (create-tx to_) / calldata length / value on v3 band: launcher = seed proxy, rest flat
killedv3-mc5/launch-metadataanalyst3discard10 Oct 20:47
#332Slow bloomers: 56 real v3 runners with cap5m<300k born<100k peaking >1h later (touch entry already killed run 35 #9)
parkedv3-mc5/late-runneranalyst2info10 Oct 20:42
#331Token mint to pool lag (token prepared before pool): born<100k no lift except v3 lag 10s-2m 6.8% vs 4.8% (n=13 hits); native lag>1h 58% = old tokens relisted born-above
killeddead-universesanalyst2discard10 Oct 20:42
#330Runner map by cell, last 3d vs prior 7d (canon): non-phantom runners only v3 fee100 (33 vs 86/7d) and H3000 hooked WETH (19 vs 50); unwatched v4 WETH nohook 3 vs 17, native nohook 6 vs 19, 0 reach 3M; no shifting cell
killedunhooked-v4/blind-spotanalyst2info10 Oct 20:42
#328Nohook v4 weth cap_5m ladder: P1M 0/0.2/0.9/3.2/9/31% by 5m cap; native 7-18% at 150-400k, usdg thin
parkedunhooked-v4/blind-spotanalyst1info10 Oct 20:28
#327Pool LP share of 1B supply at birth (canon init_liq_tok) as rug/fair-launch marker
killedunhooked-v4/blind-spotanalyst1discard10 Oct 20:28
#326Orderflow shape first 5 min (actors, top-buyer share, interarrival CV, sell frac, median ticket) on first pools
killedunhooked-v4/blind-spotanalyst1discard10 Oct 20:28
#325Sibling-pool birth gap as feature (canon): v3 sibling >30m P1M 48% vs 1.4%, v4:weth:hook 5-30m 43%
parkedv3-mc5/late-runneranalyst1info10 Oct 20:28
#324Canonical-pool table: deployer-history lens (prior runners by creator_from) on all venues
killeddead-universesanalyst1discard10 Oct 20:28
#19D0FD launcher 0xd0fd42 native fee-100
killedunhooked-v4/first-minute-filtersanalyst2discard10 Oct 20:17
#322Crew army in first 5 swaps (tx.from): crew-created hooked WETH 0/316 reach 1M vs 77/258 clean; v3 0/29 vs 50/317; not an edge for band (already excluded)
parkedhooked-v4/first-minute-filtersanalyst2info10 Oct 20:17
#321Theme saturation (same name stem in prior 6h) flat: v3 P1M 3.5/4.5/3.7/2.9% by 0/1-2/3-9/10+
killeddead-universesanalyst2discard10 Oct 20:17
#320Native nohook two-sided flow (sell/buy>0.3) 25/491 vs 2/569 to 1M: wash-farm batch marker, not signal
killedunhooked-v4/blind-spotanalyst2discard10 Oct 20:17
#319Creator buys own token: only the create-block buy lifts (= create-tx value); later creator re-buy 1/177 but pools never reach band
killedv3-mc5/walletsanalyst2discard10 Oct 20:17
#318Sellprobe census: of 31 nohook v4 pools shown >=1M cap only ~14 let a $100 sell out under 10% impact (2 h data)
parkedunhooked-v4/blind-spotanalyst2info10 Oct 20:17
#317Creator wallet nonce (tx count at create) as burner-vs-veteran label, 1596 born-small band pools: v3 fee100 nonce>5 8.6% n35 vs 23% n490 to real 1M (3rd slice 3.7% vs 28%); v4 nohook weth/native flat ~7%, usdg 141/155 nonce>100 and 0% real
parkedv3-mc5/walletsanalyst1info10 Oct 20:07
#315Native nohook 300k touches (radar 10 Oct, real by age/depth): 0 of 43 reach 1M, 27 are one kit (cap_first 33.9k depth 22k); avoid-list
killedunhooked-v4/blind-spotanalyst3discard10 Oct 20:07
#314v4:stock/tokq radar crossings (BASCAT BASKAT BINU BLACKROCK) are token/token pricing artifacts; cause false runner-miss rows
killedunknown-venue/dataanalyst3discard10 Oct 20:07
#313Creator self-buy/self-sell in first 5 min: P1M 49% on sellers (n43) but pool already higher, tp5 sim 1.29 vs 1.20: marker not edge
killedv3-mc5/walletsanalyst3discard10 Oct 20:07
#312Funder ring (shared hop1/hop2 funder across creators): farm-like, v3 band 1/29 reach 1M vs 22%; coverage 14%, parked
parkedv3-mc5/walletsanalyst3info10 Oct 20:07
#311Init LP size (skin in game) = born-high proxy; inside band inverted (USDG 19%->5%). Born-low nohook P1M native .6% USDG 1.3% WETH 1.6%
killedunhooked-v4/blind-spotanalyst3discard10 Oct 20:07
#310Stack of live flags on mc5 band: seed&vol49&not-serial 1.49 n=128 (1.61/1.37); copy/caps/hot add <0.1; no new arm
parkedv3-mc5/first-minute-filtersanalyst3no new arm; stack = SEED+VOL510 Oct 20:07
#308Social metadata of paid DS profiles (twitter/telegram/web/dup domain): flat, P1M|paid 71%
killeddead-universesanalyst3discard10 Oct 19:32
#307Hook bytecode family (4460 hooked pools): only campaign fam=H3000 converts in band; big fam 1633_e3bb is born-above-600k jumps
killedunhooked-v4/blind-spotanalyst3discard10 Oct 19:32
#306DS paid profile paid-at-T vs unpaid, venue matched (backfill 2290 tokens): no lift within venue
killeddead-universesanalyst3discard10 Oct 19:32
#302Native nohook born-small band (cap5 100k+, depth>5k): P2x .43 P3x .32 P5x .16 by max proxy (2x too good), pop>=6.25 .48; duplicates #262, pulls kill
killedunhooked-v4/blind-spotanalyst1discard10 Oct 18:22
#301Sibling attribution: 505 tokens whose nohook born-above pool is a 'runner': only 60 have a real born-small first pool reaching 1M; USDG 163/384, WETH 6/50, native 49/64; rest are twin/decoy siblings
parkedunhooked-v4/blind-spotanalyst1info10 Oct 18:22
#300Runner contagion (v3 runners in prior 6h, 2h confirm lag) as regime: 25% vs 9% pooled but weeks 38-40 .17/.34/.19/.31 vs .13/.25/.21/.08, band 40% vs 32% n37: early-period confound
killedscout/regimeanalyst1discard10 Oct 18:22
#299Seed size, senders, depth as nohook predictors: seed flat, senders weak (WETH 3-4 senders 9% vs 2.5%), USDG depth median 2 = dead; 62% of 50k+ pools dead in 10 min regardless of senders
killedunhooked-v4/blind-spotanalyst1discard10 Oct 18:22
#298Pop ratio cap5m/cap_first (IPO pop) on born-small v4 nohook band: native top quartile 23% vs 4-9% (n56), WETH 13% vs 2-5%, still P1M<=10% overall; 26 native/12 usdg/37 weth real born-small runners in 4 wk
parkedunhooked-v4/blind-spotanalyst1info10 Oct 18:22
#296FAST-LADDER exit variants: half@3+trail35/50 1.45/1.51 < fixed tp5 1.78
killedv3-mc5/exitsanalyst2discard10 Oct 17:55
#295Tier-velocity P(10M|3M) by 1M->3M gap (e1): v2/v3 fast 36-40% vs 22-28%; v4 nohook 5-30m 34% n=56 (needs pull-aware tape sim); hooked v4 6%
parkedunhooked-v4/blind-spotanalyst2info10 Oct 17:55
#294ladder.db 'missing from pool_paths' 50 pools = 31 glitch junk + 19 real v3 newer than pool_paths refresh; no venue gap
killedunknown-venue/dataanalyst2discard10 Oct 17:55
#293Round-number cap barrier (peak heaping below 100k..10M): flat all venues
killeddead-universesanalyst2discard10 Oct 17:55
#292Ungated late ladder with real gap fills + stale haircut: 2M tp5 1.13 bs5 .96 (kills #275 1.24)
killedv3-mc5/late-runneranalyst2246discard10 Oct 17:55
#288Early big seller (one sell>10% of inflow in first 20 prints) as skip: 7.7% n26 vs 20.4% n515
parkedv3-mc5/first-minute-filtersanalyst1info10 Oct 17:47
#287First-12-print burst and reserve after 20 prints on v3 fee100: real-runner 16% vs 3%, 2.5% to 48%
killedv3-mc5/first-minute-filtersanalyst1discard10 Oct 17:47
#286Dip-buy inside confirmed runner (peak>=20x, fall to .5x, exit .75 of peak): mean 1.06-1.08 n=225-506 vs random baseline .98
parkeddead-universesanalyst1info10 Oct 17:47
#285Reserve-backed real-runner map by family (P2 WETH): v3 fee100 168, H3000 a80 x4 hooks 59 of ~110 pools (55%), v4 fee100 nohook 16; a80 6 pools/day no decay
parkedhooked-v4/short-tpanalyst1info10 Oct 17:47
#284Hook-address reputation over 1473 hooks: top runner hooks are 15-pool wash farms, real-depth share 0
killedunhooked-v4/blind-spotanalyst1discard10 Oct 17:47
#283Third-party LP adds as pull-survival marker: every nohook toucher has 1 LP sender, 2+ adds n=9
killedunhooked-v4/blind-spotanalyst1discard10 Oct 17:47
#282Held-cap 400k-1M band on unwatched nohook native/WETH/USDG: P1M .19-.25, P3M 0-.03 with depth>=3000
killedunhooked-v4/blind-spotanalyst1discard10 Oct 17:47
#280Decimals/supply as project type: 9-dec 5.3% n=38 vs 7.2%, flat
killeddead-universesanalyst2killed10 Oct 17:25
#278Launch lag first trade minus create: lag0 P1M 8.5% vs 2.5% - same as create-tx value
parkedv3-mc5/launch-metadataanalyst2parked10 Oct 17:25
#277Acceleration t50k-t100k-t200k as runner predictor: only absolute speed lifts (known fast climber)
killedv3-mc5/first-minute-filtersanalyst2killed10 Oct 17:25
#276Swaps-per-sender/ticket size at 5 min: look-ahead trap (n_swaps_72h); ticket flat
killeddead-universesanalyst2killed10 Oct 17:25
#274Tier ladder per venue: where runners stall
parkedunhooked-v4/blind-spotanalyst2info10 Oct 17:25
#273Live mc5 fills that later got a DS profile (9 of 46): 8 stopped, profile arrives after entry: no filter use
killedv3-mc5/walletsanalyst3discard10 Oct 17:18
#272DS profile on unwatched nohook v4 USDG/WETH: P1M 4%/10% (n 28/10) no lift; lift is v3 and hooked WETH only
killedunhooked-v4/blind-spotanalyst3discard10 Oct 17:18
#270Price-path shape 0-5 min (monotone staircase, last-3m flat, late burst) on mc5 band: peak multiple no signal
killeddead-universesanalyst3discard10 Oct 17:18
#269Bridge ETH inflow/net per day vs runner rate (25 days): rho -0.1..-0.4 with drift, n too small
killedscout/regimeanalyst3discard10 Oct 17:18
#268Launch crowding (births in prior 5 min): tracked-pool counts show 5%->1% P1M gradient, but with all-chain births (pools_since_68m 346k) it vanishes: selection artifact
killeddead-universesanalyst3discard10 Oct 17:18
#267Sibling fan-out (extra pools of same token born within 300 s of first pool): own-pool P1M 2.3% vs 2.5%, only >=3 siblings 16% (n=32); later siblings are a consequence of the run
parkedunhooked-v4/blind-spotanalyst3info10 Oct 17:18
#263v3 late 300k cross age 6-60 min: proxy P5x 19% but harness replay tp2/3/5 mean 0.59-1.04
killedv3-mc5/late-runneranalyst1192discard10 Oct 17:17
#261Hook permission bits (low 14 bits) as launch label on 1308 hooked v4 pools: bSwap .74 vs .38, aRemoveLiq .24 vs .70, but = fee/kit proxy
killedhooked-v4/first-minute-filtersanalyst11308discard10 Oct 17:17
#25910 Oct unwatched census: nohook usdg 3/41, native 1/27, weth 1/6 non-jump 1M; Mouse/PERSUASION 20M = jumps
killedunhooked-v4/blind-spotanalyst3discard10 Oct 16:50
#258Name-form battery venue-adjusted flat; HOOD/ROBIN .65 n=40 too small; contract caps identical on band
killeddead-universesanalyst3discard10 Oct 16:50
#255Repeat sender among value>=10 launches: 3M 7% (n=14) vs first-time 29%
parkedv3-mc5/walletsanalyst2info10 Oct 16:46
#254Per-day band opportunity 10-29 entries/day, 2x 40% 5x 17% stable: edge decay is path not supply
parkedv3-mc5/5min-capanalyst2info10 Oct 16:46
#253Confirmed stop (2-3 prints or 15-30 s below): 0.94-1.01 vs 0.99 on band n=154
killedv3-mc5/exitsanalyst2discard10 Oct 16:46
#251Launch kit A/B/other census: A 22/31 to 1M, B 59/217, other 38/2953; band is 53/56 kit B so kit is no filter
killedv3-mc5/walletsanalyst2discard10 Oct 16:46
#250Unwatched venue truth (pool_paths, born<100k): real 1M runners 3 wks: v3 94, native nohook 10, USDG nohook 7, WETH nohook 3; rest price phantoms; 23 '?' misses all 7 Oct
parkedunhooked-v4/blind-spotanalyst2info10 Oct 16:46
#248New-kit radar: hook 0xfcd1..c15c7 645 pools in 2 d = identical-seed decoy farm (cap 915,939)
killeddead-universesanalyst1645discard10 Oct 16:44
#247Fast-climber lift sweep all groups: general lift, but unhooked fee100 54% is one kit batch (34/82)
killedunhooked-v4/first-minute-filtersanalyst182discard10 Oct 16:44
#246Born-above 500k-1M later crossing 1M: P10M 3% (native 0/200); >6M rows are same-print stamps
killeddead-universesanalyst1396discard10 Oct 16:44
#244Name theme meta (cat/dog/stock-pun) WETH: 56% vs 37% but fleet-kit naming; non-fleet .54 n=28 vs .36 z1.8, names cover 25%
parkedv3-mc5/first-minute-filtersanalyst1147info10 Oct 16:44
#243Regime day/hour/weekday persistence of P(1M|climber) 27 days: flat 0.32-0.42, lag1 -0.05
killeddead-universesanalyst1discard10 Oct 16:44
#242Pull lead time: 384 pulled no-hook pools, 100% one-event LP removal, lead 0 s, no LP-data exit alarm possible
killedunhooked-v4/blind-spotanalyst1384discard10 Oct 16:44
#241Full-range/many-LP-ops no-hook WETH/native: pulled 14-31% vs 90%, median 0.9 (n=23): rug-safe is not runner
killedunhooked-v4/blind-spotanalyst123discard10 Oct 16:44
#239Name form: multiword name!=symbol 1.52 vs 1.22 (n=164 named); names.db has no rows after 7 Oct and unnamed=0.69 is coverage artifact
name form		parkedv3-mc5/first-minute-filtersanalyst3info10 Oct 16:21
#238v4 WETH nohook born 50-300k reaching 300k (P1M 59%, 41 pools): depth 0-1, 1-3 senders, fee3000 decoys; not enterable
killedunhooked-v4/blind-spotanalyst3discard10 Oct 16:21
#237Token address vanity (numeric first nibble, digit count, repeats) on band: flat
killeddead-universesanalyst3discard10 Oct 16:21
#236First-print cap <8k at 300k touch, all venues (radar_hist): P1M v3 .41 vs .13, native nohook .65 vs .17 (n=17), native hook .80 vs .36, usdg .46; consistent seed lift
seed-climb lift generalizes across venues		parkedunhooked-v4/blind-spotanalyst3info10 Oct 16:21
#232WETH nohook touchers/day 17-38 -> 1-4 since 3 Oct; births 1000-2800 -> 400/day; native births rising but 87% pulled
parkedunhooked-v4/blind-spotanalyst1info10 Oct 16:05
#230No-hook baseline is mark-to-last-print artifact: 88% of WETH/native touchers pulled in 4h, tp5 sl.5 1.59 -> 0.13-0.23 pull-aware (kind 10)
parkedunhooked-v4/blind-spotanalyst1439info10 Oct 16:05
#233Fixed-hold exit before the pull (30s-1h, pull-aware) on nohook WETH/native: median 1.00 at 5min, 0.68 at 30min
killedunhooked-v4/blind-spotanalyst1discard10 Oct 16:05
#229Earlier entry at 100k touch on WETH/native nohook: 0.87-0.97 vs 300k
killedunhooked-v4/blind-spotanalyst1discard10 Oct 16:05
#228Funding route of early buyers (bridge/contract share) on v4 nohook: flat beyond funder count
killeddead-universesanalyst1discard10 Oct 16:05
#227Bundled launch (>=3 swaps first 2 blocks) on v4 nohook: n too small (USDG 10, WETH 37 with 0 wins)
killedunhooked-v4/blind-spotanalyst1discard10 Oct 16:05
#226Creator/factory reputation on v4 nohook touchers: 0x58da factory 275/290 WETH, creators do not repeat, creator field is launcher contract
killeddead-universesanalyst1discard10 Oct 16:05
#225Repeat-winner wallets walk-forward on v4 nohook (tx senders first 15 swaps): top-quartile score no better than base
killeddead-universesanalyst1discard10 Oct 16:05
#223Young climbers by venue (age 60-1800 s): hooked WETH P3M 48% (watched), unwatched nohook P3M <=2%, hooked native enterable 3M n=15
unwatched young climbers reach 3M		parkedunhooked-v4/blind-spotanalyst3info10 Oct 16:03
#222Regime: band tp5 by day vs ETH/BTC daily return 21 days: same-day -0.31, no usable pause
pause on down days		killedscout/regimeanalyst3discard10 Oct 16:03
#221First-minute volume floor on band: 5-min vol<49k = 0.78, hit5 0% (97 of 290); duplicate of #548, threshold refined
stale born-in-band pools hold slots		parkedv3-mc5/first-minute-filtersanalyst3290info10 Oct 16:03
#219Aged v4 nohook awakening (age>=1h at 300k): USDG 110 P1M 34% but 13 hits enterable, rest same-second jumps
dormant coin awakening is climbable		killedunhooked-v4/blind-spotanalyst3discard10 Oct 16:03
#217Holder concentration at +5min from real Transfer logs (169 band pools)
top1/top5 share terciles: hits 3/2/1, halves inconsistent; closes dataeng #562/#553		killedv3-mc5/walletsanalyst2169discard10 Oct 15:59
#216Hold past 4h only for survivors >=3x, trail 30%
28 pools: 14.4x at 4h vs 14.2x trailed; peak not capturable		killedv3-mc5/exitsanalyst228discard10 Oct 15:59
#215Pull-only exit instead of price stop on band
No stop 0.59 vs stop .5 0.91; pulls atomic		killedv3-mc5/exitsanalyst254discard10 Oct 15:59
#213Band replay on fresh 4 days (6-10 Oct) vs live
Same-period replay reproduces live: 0.91x n=54, 4% hit5 before stop, fast stops 31% vs live 40%		parkedv3-mc5/5min-capanalyst254info10 Oct 15:59
#206Speed 300k->1M predicts 3M: <20min P3 11-15% vs 20-120min 34% (n=395, halves agree); filter for 1M entry
killeddead-universesanalyst3speed confounded by venue (weth:hook); <5min best for P10 .17; no clean filter10 Oct 15:25
#211Cross-chain name match: only 3h of gecko_x, parked to 17 Oct
parkedscout/cross-chainanalyst3info10 Oct 15:18
#210Born-big pools (first print >=3M, n=1135): next-threshold hit 12%, depth>=300k 6%
killeddead-universesanalyst3discard10 Oct 15:18
#208Chain launch count/day halved (23k->12k) while non-jump 1M climbers flat ~21/day: no regime switch
killeddead-universesanalyst3discard10 Oct 15:18
#207Depth at touch is a fixed curve (90k@300k,165k@1M,325k@3M): useless as predictor, capacity fine
killeddead-universesanalyst3discard10 Oct 15:18
#205Fast non-jump climbers on nohook native/WETH (300k touch 30-300s after birth): P1M 56%/28% n=16/47, 2.6/day, ceiling 1M, no path tape
parkedunhooked-v4/blind-spotanalyst3info10 Oct 15:18
#202Pull robustness of mc5 band replay (v3 liq removes >=1 ETH within 4 h)
replay overstates because LP pulls are invisible on swap tape		killedv3-mc5/5min-capanalyst2169discard10 Oct 15:12
#201v3 10M+ runners are seeded 300-800k: sub-band of mc5 band (350-700k) pays more
cap_lo higher than 250k (350k) lifts mean		killedv3-mc5/5min-capanalyst2169discard10 Oct 15:12
#200Hooked USDG dynamic-fee (8388608, hook ..a080) batch 30 Sep: 5 pools to 16-78M peak
a launcher batch of hooked USDG pools produced 20M+ runners we do not watch		killedunhooked-v4/first-minute-filtersanalyst28discard10 Oct 15:12
#199Clean runner map without phantoms: climbing (300k->1M >=120 s), real (>=10 buyers, wash<.5, best pool), 21 Sep-9 Oct
where do real climbing runners come from by pool type and tier		parkeddead-universesanalyst2248info10 Oct 15:12
#198v3 non-fee100 tiers first-print+30/90s entry tp2/3/5
fee10000 v3 pools early entry (30-90 s)		killedv3-mc5/late-runneranalyst2198discard10 Oct 15:12
#197v3 non-fee100 tiers (500/3000/10000) mc5 band at +5 min
other v3 fee tiers have more 3M+ runners (fee3000 82% of runners reach 3M) so band entry pays		killedv3-mc5/variantsanalyst213discard10 Oct 15:12
#196mc5 exit grid corrected (tp 5/10/20/none x hold 1-24h x stop .3-.7), same band universe
longer hold or tp20 beats live exit stop.5 + 4h no-tp		killedv3-mc5/exitsanalyst2169discard10 Oct 15:12
#195mc5 band 5-min path geometry (efficiency ratio, VWAP closing-line, max drawdown, trade-size skew, first-gap) as filter
terciles of 9 first-5-min path features split the tp5/sl.5 outcome by >=0.4x in both halves		killedv3-mc5/first-minute-filtersanalyst2169discard10 Oct 15:12
#193v4:native:hook 1M rows: 236 of 245 same-minute first prints
killedunknown-venue/dataanalyst1discard10 Oct 15:06
#192Hooked WETH 300k-touch confirm entry age<=30 min (P1M 80%)
parkedhooked-v4/short-tpanalyst1info10 Oct 15:06
#191Unknown venue = tokq hook: stock-token quote phantom caps (PLTR 2.17M x3 pools)
killedunknown-venue/dataanalyst1discard10 Oct 15:06
#190X/social posts lead-lag (savenow xposts, 89 tokens): first post before 1M only 17%
killeddead-universesanalyst1discard10 Oct 15:06
#189Crowding: other 300k touches in prior 60 min vs real rate
killeddead-universesanalyst1discard10 Oct 15:06
#188Birth cap and birth hour across venues (radar_hist)
killeddead-universesanalyst1discard10 Oct 15:06
#187Age at 300k touch: hooked WETH <=30 min 80% real vs 4%; v3/native weaker, NHW flat
parkedhooked-v4/short-tpanalyst1info10 Oct 15:06
#186Depth/cap ratio at 300k touch as runner predictor (radar_hist)
killeddead-universesanalyst1discard10 Oct 15:06
#183Phantom runners by name: Robinhood Token stock names (78) + cbBTC/USDT lookalikes; skip in radar
killedunknown-venue/dataanalyst3discard10 Oct 15:04
#182Supply/decimals metadata: non-round supply = burn leakage + real Robinhood stock tokens
killeddead-universesanalyst3discard10 Oct 15:04
#181v3 fee10000 first non-jump 300k touch: 19 pools, 0.94
killedv3-fee10000/late-runneranalyst3discard10 Oct 15:04
#180Name-word meta heat: P(1M) 20% vs 8% but +60s entry 1.81 vs 2.21: priced in
killeddead-universesanalyst3discard10 Oct 15:04
#179Pool ETH / net inflow / depth-to-cap at +5 min: flat (pool_eth NULL 76%)
killedv3-mc5/first-minute-filtersanalyst3discard10 Oct 15:04
#178Crowding (concurrent band pools) and UTC 00-05 hour: crowding flat, hour 2.09 n=50 halves 1.4/2.7
parkedv3-mc5/5min-capanalyst3info10 Oct 15:04
#172NHW S6 time decay: 39/45 trades before 2 Oct, 5-9 Oct n=6 ~1.5; supply of NHW touches fell 17-38/day to 1-6/day
parkedunhooked-v4/first-minute-filtersanalyst1info10 Oct 14:32
#176Phantom runners: 7 of 11 non-hooked v4 runner-check rows are same-minute 1M-20M first prints; research tag #568
killedunknown-venue/dataanalyst1discard10 Oct 14:32
#175Risk of ruin on live mc5: 15-loss streak P=99% per 100 trades (win rate 6.8%), Kelly ~6% (1.5% ex best); do not pause on streaks
killedv3-mc5/sizing-deployeranalyst1discard10 Oct 14:32
#174mc5 live skips by cap at +5min: <250k 5 real of 838; 250k-1M 37%; wake-up burst v3 fee100 n=588 1.25 pre-cost
killedv3-mc5/late-runneranalyst1discard10 Oct 14:32
#170Hazard of 1M after 300k touch by minutes since touch: flat, no stall/time-stop signal
killeddead-universesanalyst1discard10 Oct 14:32
#169Ratchet/trailing stop after 3x on live mc5 trades (real paths, n=44): 0.94-1.43 vs 1.86 stop-only
killedv3-mc5/exitsanalyst1discard10 Oct 14:32
#167v3 fee100 below-band runners (cap<300k at +5min): first 300k touch entry (non-jump, dust guard)
51 real runners miss the band; touch entry catches them		killedv3-mc5/late-runneranalyst2282discard10 Oct 14:18
#166Real-runner recount (own buyers>=10, wash<.5, best pool): 272, 67% v3 fee100 WETH, 18% USDG (dead since 3 Oct)
where real runners sit vs what we watch		parkedv3-mc5/late-runneranalyst2272info10 Oct 14:18
#165Cross-pool arbitrage between sibling stable-quoted v4 pools (price gap net of fees)
gap>5% net of fees at usable size		killeddead-universesanalyst2326620discard10 Oct 14:18
#164Sibling v4 pool birth after the +5min gate as confirmation trigger for v3 fee100 band pools
later sibling pool = runner confirmation; enter at sibling birth		killedv3-mc5/late-runneranalyst293discard10 Oct 14:18
#163Fee-ladder spam: 22% of nohook v4 pools sit in 842 tokens with >=10 sibling pools (up to 107 per token)
pool-level runner counts inflated; sibling pool count as marker		parkeddead-universesanalyst252057info10 Oct 14:18
#162v3 fee100 cap x gate-time matrix (5/15/30/60 min x 50k..5M), mark-to-print
any untested cap/age cell with mean>=1.5 both halves		killedv3-mc5/late-runneranalyst22891discard10 Oct 14:18
#161mc5 band: no swap in the 60 s before the +5min gate = 0.93x hit5 0% (n=50) vs 1.47x
quiet last minute = dead pool		killedv3-mc5/5min-capanalyst2242discard10 Oct 14:18
#160Birth LP shape predicts LP pull on native nohook: full-range fee10000 410/410 pulled; concentrated fee2500 12% pulled
first Mint range + fee separates rug kit from survivors		killedv4-nohook-native/blind-spotanalyst2960discard10 Oct 14:18
#159v4 nohook LP range ceiling (tight first Mint range caps the reachable cap)
pools with tight first-Mint range cannot pass the cap at the range top		killedv4-nohook-native/blind-spotanalyst22605discard10 Oct 14:18
#158Gate microstructure on mc5 band: entry/gate print gap, turnover, avg trade size, LP add/remove events in first 5 min, pool-creation-to-first-swap lag (0 s for 93%): all non-monotone/flat
killeddead-universesanalyst3discard10 Oct 14:14
#156Pons graduated pools (358 in v3live tape, 6 reach 1M): cap-cross entries 60k-500k tp2/3/5 sl.5 = 0.8-1.2x; active at +5min (>=20 swaps, n=69) tp2 hit 46% but 1.10x net. Flat, tape exists (corrects 'no tape')
killedpons/grad-poolsanalyst369discard10 Oct 14:14
#155Resurrection/migration: v3 pool for a token that already traded on another venue: 23% real 1M vs 3.6%, but all 16 are born-at-price copies (first 100k=300k=1M same second): jump trap
killedunhooked-v4/blind-spotanalyst3discard10 Oct 14:14
#144Token-level accounting: 59 of 98 real-1M USDG nohook pools are tokens that also have a v3 fee100 pool; sibling-pool appearance is not a predictor (P1M 0.185 vs 0.184 at +30 min)
parkedunhooked-v4/blind-spotanalyst3info10 Oct 14:14
#152IPO oversubscription: ETH volume before 300k touch on NHW (n=388)
killedunhooked-v4/first-minute-filtersanalyst1388discard10 Oct 14:10
#151Gas price / priority fee of first buys as bot marker
killeddead-universesanalyst1discard10 Oct 14:10
#148Runner clustering / hot hand per pool type (3h prior vs next)
killeddead-universesanalyst1discard10 Oct 14:10
#147First-swap latency after pool creation as launch-prep signal (USDG/WETH nohook)
killedstable-v4/first-minute-filtersanalyst1discard10 Oct 14:10
#145Born-above artifact: 28/30 USDG, 24/24 native, 5/5 WETH nohook 1M crossers in radar were born above 500k (climbers 2/0/0)
parkedunknown-venue/dataanalyst159info10 Oct 14:10
#141Creator reputation / operator-regime autocorrelation in fee100 nohook WETH
repeat deployers or hot/cold operator streaks predict pull		killeddead-universesanalyst21329discard10 Oct 13:26
#140Pull predictability from first 90 s features (fee100 nohook WETH)
rug-in-60s coins share a visible first-90s signature (inversion)		killeddead-universesanalyst21329discard10 Oct 13:26
#139Lindy/survivor entry on native nohook (alive at age 1-24h, cap 50-600k)
pools that survive N hours without LP pull are real runners		killedv4-nohook-native/blind-spotanalyst231discard10 Oct 13:26
#138fee100 nohook WETH +90s entry with pull-as-zero (19 d, 1329 pools)
early entry in the 3204-pool nohook WETH universe pays		killedunhooked-v4/baseline-deadanalyst21329discard10 Oct 13:26
#137Order-flow toxicity/depth at 300k touch, v4 nohook WETH (n=320)
buy/sell imbalance, trade size, depth at entry separate winners		killedunhooked-v4/first-minute-filtersanalyst2320discard10 Oct 13:26
#136Data request: tape for 56 nohook fee100 climbers + 366 controls (plateau pulls, exit depth)
parkedv4-nohook-native/blind-spotanalyst3422needs-data10 Oct 13:20
#134Seed-cap preset hot kit: P(1M) 14-25% vs 1-3% but entry return worse (1.66 vs 2.23)
killedv3-fee100/first-buyanalyst35581discard10 Oct 13:20
#133v3 climbers to 1M by timing: 13 <5min, 17 band, 7 late (0 reach 5M)
parkedv3-mc5/late-runneranalyst337info10 Oct 13:20
#132Early wallets (30 s buyers) net exit by 5 min: flat
killedv3-mc5/walletsanalyst3330discard10 Oct 13:20
#131Born 1-2.5M seeded v4 pools: 0 of 6 reach 3M
killeddead-universesanalyst36discard10 Oct 13:20
#130Mint-to-PM factory 0x8366 nohook fee3000 WETH = decoy twins of H3000
killedhooked-v4/dead-endsanalyst332discard10 Oct 13:20
#129BTC/ETH minute regime + afterglow on mc5 band: flat
killeddead-universesanalyst3330discard10 Oct 13:20
#127v4 nohook native/USDG/WETH 1M crossings 7-10 Oct: 51 of 53 single-print, 2 climbs w/o depth
killedv4-nohook-native/blind-spotanalyst353discard10 Oct 13:20
#124Batch launches (same creator +-10 min) as signal: none
killeddead-universesanalyst1discard10 Oct 13:01
#1231M ride velocity: 300k->1M gap >110 min gives P(5M) 0.10 vs 0.37
killedv3-mc5/late-runneranalyst1discard10 Oct 13:01
#122Funder breadth on USDG/native nohook: flat
killeddead-universesanalyst1discard10 Oct 13:01
#125Early momentum m300/m60 on v4 nohook WETH: tp3 1.09 after depth, no margin
killedunhooked-v4/first-minute-filtersanalyst1discard10 Oct 13:01
#121Token code hash/selectors: hash 5f4a77e044c6 = wash decoy kit (hit3 97% fake); selectors flat
killedv4-nohook-native/blind-spotanalyst1discard10 Oct 13:01
#120Deployer prior-pull history (v4 nohook 27% vs 1.8%) but only 1.2% pulled
killeddead-universesanalyst1discard10 Oct 13:01
#119LP NFT lock/burn/owner status as rug marker (v4, 4.7k pools)
locked LP means fewer pulls		killedv4-nohook-native/blind-spotanalyst1discard10 Oct 13:01
#118Bridge ETH / stablecoin mint inflow regime vs daily hit rate (19 d)
capital inflow predicts run rate		killeddead-universesanalyst1discard10 Oct 13:01
#57N300: enter nohook custom-fee native at first 300-350k touch, TP3, non-jump
sim 1.5 (pess 1.21), unverified: stale prices		killedv4-nohook-native/blind-spotanalyst3786discard10 Oct 12:32
#116LP-add precursor in aged native pools
killedv4-nohook-native/blind-spotanalyst227discard10 Oct 12:32
#115N300 native no-hook re-test with depth and LP pulls: 0.19x
killedv4-nohook-native/blind-spotanalyst2226discard10 Oct 12:32
#114Hook 0xc4 kit runners are decoys (median 1 buyer, $32 vol)
killeddead-universesanalyst222discard10 Oct 12:32
#113FDV/TVL multiple in mc5 band
killedv3-mc5/5min-capanalyst2198discard10 Oct 12:32
#112Native no-hook wake-up burst late hunter
killedv4-nohook-native/blind-spotanalyst212discard10 Oct 12:32
#111v3 fee100 low-cap (20-100k at +5min) band = P204 universe, dominated
killedv3-new-pools/tp5analyst2347discard10 Oct 12:32
#110GeckoTerminal trending board as attention entry signal
first trending_5m appearance fdv 150-400k gives >=3x more often than baseline		killeddead-universesanalyst2438discard10 Oct 12:32
#109Sender arrival rate and ETH 12h regime inside nohook WETH: flat
killeddead-universesanalyst1discard10 Oct 12:18
#108Name/ticker/supply metadata on nohook WETH 300k pools: flat
killeddead-universesanalyst1discard10 Oct 12:18
#107Sender breadth: USDG nohook 6-12 senders 1.55 vs 1.25 (n=184), not net positive
killeddead-universesanalyst1discard10 Oct 12:18
#102v3 first-time deployer 4.1% real 1M vs repeat deployer 0.5% (1512 pools covered)
parkedv3-mc5/sizing-deployeranalyst3info10 Oct 12:05
#101USDG nohook deployer with >=2 prior 1M pools: 16% real 1M vs 1%, n_eff ~4 deployers
parkedstable-v4/first-minute-filtersanalyst3info10 Oct 12:05
#100Same-symbol-seen-before and short ticker: v3 1.4x / SI family artifact, none in band
killeddead-universesanalyst3discard10 Oct 12:05
#99Name meta hot-word (word shared with 7d prior 1M runner): 3x population lift, none in band or at +66s
killeddead-universesanalyst3discard10 Oct 12:05
#97Winner-wallet early buyers on v3: 13% vs 0.3% runner rate in population, zero lift inside mc5 band (busy-pool effect) and at +66s
killedv3-mc5/walletsanalyst3discard10 Oct 12:05
#96Phantom runners: only 22% of pools with peak>=1M are real; per token 440 real 1M, 113 real 5M, 32 real 20M (23 of 32 are v3)
parkeddead-universesanalyst3info10 Oct 12:05
#95Chain-wide births per day and weekday vs v3 1M rate
busy launch days or weekdays change the hit rate		killeddead-universesanalyst21000discard10 Oct 11:31
#93mc5 gate time +2/3/5/10/15 min
earlier or later gate beats +5 min		killedv3-mc5/5min-capanalyst2330discard10 Oct 11:30
#90Scale-free hazard: stall levels for exits
runners stall at certain cap levels, set TP just below		killedv3-mc5/exitsanalyst23100discard10 Oct 11:30
#89Dynamic-fee (8388608) hooked and SUSD pools as runner marker
dyn-fee hook kits mark real runners		killeddead-universesanalyst22200discard10 Oct 11:30
#88Hour-of-day of birth vs runner rate
Births 08-16 BKK have lower runner rate, skip them		killedv3-mc5/5min-capanalyst2330discard10 Oct 11:30
#87Ride from first 1M touch by pool group (P(3M|1M))
P(3M|1M) by group and age at 1M tells where a 1M-entry ride pays		killeddead-universesanalyst2800discard10 Oct 11:30
#86mc5 live dust guard + coverage audit: 21 in-band pools skipped as dust, 0 ran; 44/44 in-band pools seen despite 8 Oct feed errors
live misses runners through dust guard or feed outage		parkedv3-mc5/5min-capanalyst244info10 Oct 11:05
#850xcc native pools: do real outsiders get fair sells? tx_from request #484
outsider sell-through >=0.6x control		parkedhooked-v4/sellability-riskanalyst2199request10 Oct 11:04
#84mc5 ruin math: 41 live trades, 3 winners
15 losses in a row is abnormal?		parkedv3-mc5/sizing-deployeranalyst241info10 Oct 11:04
#83mc5 live partial take-profit (25/50% at 2/3/5x)
partials lock peaks of 38 losers		killedv3-mc5/exitsanalyst242discard10 Oct 11:04
#82USDG nohook creator novelty (first-time vs fleet creators) and launcher concentration
fresh creators run more		killedstable-v4/first-minute-filtersanalyst24146discard10 Oct 11:04
#81mc5 by v3 fee tier (100 vs 10000)
fee tier separates runners		killedv3-mc5/variantsanalyst2330discard10 Oct 11:04
#80mc5 sub-band: cap 600k-1M at +5min beats 250k-600k (live 3 of 3 winners)
Higher cap inside band runs more		killedv3-mc5/5min-capanalyst21712discard10 Oct 11:04
#79Stock-quoted hook 0x4e34 pools: 190 of 217 (88%) never trade, 1 of 24 traded moves 3x
731 stock-quoted pools hide runners		killedunknown-venue/dataanalyst2217discard10 Oct 11:04
#78USDG v4 nohook runners: 71% have a >2x single print, median peak/seed 1.0 for seed>=3M
USDG 5M+ share is real moving runners		killeddead-universesanalyst2238discard10 Oct 11:04
#76v3 fee100 early-wallet/deployer walk-forward: blocked, signer/creator NULL in v3hist
parkedv3-mc5/walletsanalyst3needs-data10 Oct 11:04
#77Venue lifecycle: v4 USDG nohook real 1M per 6 d 91/66/25 (dying); v3 fee100 steady 61/81/62
killeddead-universesanalyst3info10 Oct 11:04
#75Second mc5 checkpoint (250k-1M at +15..120 min, not caught at +5): 0.69-1.16, halves disagree
killedv3-mc5/late-runneranalyst3104discard10 Oct 11:04
#73Unknown-venue/8080-hook 'runners' are decoy/pricing artifacts (median $2-44 vol)
killedunknown-venue/dataanalyst3discard10 Oct 11:04
#59Stock-quoted hook 0x4e34 pools (57k, unknown venue): runner rate unknown
need USD price series		killedunknown-venue/dataanalyst37527discard10 Oct 11:04
#72Holder breadth at 1M touch (runtrack): coverage 29/114
killeddead-universesanalyst1discard10 Oct 10:53
#71Sibling-pool hot-token lift = same token counted per pool
killeddead-universesanalyst1discard10 Oct 10:53
#70Early-wallet reputation walk-forward on 13.5k USDG/SUSD pools: no persistence
killeddead-universesanalyst1discard10 Oct 10:53
#695-Min Cap on v4 nohook USDG pools: 0.79-0.93 n=84
killeddead-universesanalyst184discard10 Oct 10:53
#68USDG nohook enter at first 300k touch (real pools): tp10 2.25 n=60
parkedstable-v4/first-minute-filtersanalyst160mature10 Oct 10:53
#66Regime BTC/ETH vs clean runner rate (19 days)
no relation, rho 0.14-0.18		killeddead-universesanalyst1discard10 Oct 10:26
#64H3000 a80 scripted ramp: 9 buyers, 1 seller per pool (119/119)
no outsider ever sold; replay returns unproven; strengthens need for 5 USD sell test		parkedhooked-v4/sellability-riskanalyst1needs Ben 5 USD sell (#221)10 Oct 10:26
#63Wash loops in SUSD/USDG fee0 pairs (unknown-venue runners)
76-96% same-block round trips, volume 10-200M fake		killeddead-universesanalyst1discard10 Oct 10:26
#62Label bug: v4 nohook WETH misses are H3000 a80 pools (decoy twin)
venue label should be per token from highest volume pool		killedhooked-v4/sellability-riskanalyst1info: fix label10 Oct 10:26
#61Dust runners: 83% of pools touching 1M trade <2k (v4 nohook WETH custom fee 0 real runners)
runner counts per pool type are inflated by one-print jumps		killeddead-universesanalyst1discard10 Oct 10:26
#49smart wallet confluence 2-3 ranked wallets
see brief; unverified web claim		parkedscout/walletsscoutanalyst3: wallet confluence waits for early-buyer table (dataeng)10 Oct 10:11
#58USDG/SUSD fee-0 '10 of 10 reach 5M' is one 7 Oct launch batch
n_eff 2-3 events		parkedstable-v4/baselineanalyst317n_eff small10 Oct 10:11
#56v4 nohook custom-fee native: 95% of 5M+ are one-print jumps (runner map overstated)
most 5M+ in this pool type are enterable		parkedv4-nohook-native/blind-spotanalyst3678map correction10 Oct 10:11
#47fml2 skip_quote: no GMGN quote on fresh v4 pools
6 runners lost to quote gap; retry quote 30-60s or pool-state paper fill; live side so Ben decides		parkedhooked-v4/sellability-risktestmgrquote gap, needs Ben10 Oct 09:10
#46Iter 20-21 lenses: token order, cannibalization, congestion, round-cap, whale buy, tip bidding, social callers, arm momentum, fee-poke, dip defence
all flat, discarded		killeddead-universesanalyst1all flat/no edge10 Oct 09:10
#35MC5 variants: cells/cheap/dead/drip/exits/extend/knobs/lpowner/overlap/quality/selfbuy
parkedv3-mc5/variantsanalyst1reports a1-mc5-*.md; see each file, not yet itemised in ledger10 Oct 08:40
#33Hooked fee100 sibling structure as separate edge
killedhooked-v4/dead-endsanalyst1427dummy 1% partner pool created 6 s earlier: detection hint only, no extra trade10 Oct 08:40
#32Pull predictors (14 features, lock status, partial removal, mint/burn sender)
killedhooked-v4/dead-endsanalyst1302AUC 0.43-0.57; not predictable; no lead signal10 Oct 08:40
#31Organic universe: enter when swap-inflow reserve crosses 1-16 ETH
killeddead-universesanalyst160000.93-1.17, 24-53% pulled; no edge from real money arriving10 Oct 08:40
#30Pons curve feature pockets / baseline
killeddead-universesanalyst1960000.83-0.97 everywhere; more activity = worse10 Oct 08:40
#29Reserve gate / larger entry reserve
killeddead-universesanalyst1higher reserve at entry = worse10 Oct 08:40
#28WETH tape baseline (v3+v4), tp3/s.5
killeddead-universesanalyst128281.2 mild; entry delay edge dies with age; superseded by hooked/quiet cuts10 Oct 08:40
#27Sellability by eth_call (outsider holders H3000)
parkedhooked-v4/sellability-riskanalyst2transfers to PM and dead address pass 4/4; swap-level sell untested; needs Ben $5 sell test10 Oct 08:40
#26Stable-quoted traction entry (net inflow trigger)
killeddead-universesanalyst20.8-1.06, median 0.5810 Oct 08:40
#25Pons bonding curve graduation / traction entry
killeddead-universesanalyst2265000net 0.84-0.90, friction only10 Oct 08:40
#24Tokenized stocks arb
killeddead-universesanalyst2chain-vs-Yahoo gap 1-2x fee, arb tight10 Oct 08:40
#23Time-of-day / weekday / regime (ETH trend, vol) / hot-hand / contagion / vanity address
killeddead-universesanalyst2flat in every cut, max spread 0.2-0.26; 20-24h UTC 1.28 is one cell, no mechanism (WATCH)10 Oct 08:40
#22Dust-pool 5x universe (stable-quoted v4 tape)
killeddead-universesanalyst2105005x is price jumping through empty ticks; with unexitable hits valued low 0.95-0.9710 Oct 08:40
#21Jump-print runners ($300k->$3M one print) as signal (L8366, velocity, serial creators)
killeddead-universesanalyst2623445% of runners are one-print jumps; real rate 0.005 for dyn-fee 8366; velocity no lift10 Oct 08:40
#20Unhooked v4 fee100 baseline with real pulls
killedunhooked-v4/baseline-deadanalyst213870.98 tp3 n=1387, 64% pulled before TP: no edge (earlier 1.83 was a pull artifact)10 Oct 08:40
#7H3000 sibling/relaunch/arb detection
killedhooked-v4/dead-endsanalyst2no arb route, no earlier signal, relaunch 12/114 mixed10 Oct 08:40
#6H3000 early-slope / reserve predictor of peak
killedhooked-v4/dead-endsanalyst2median max 54-87x in all quartiles, no order10 Oct 08:40
#5H3000 limit-dip entry
killedhooked-v4/dead-endsanalyst2ramp does not dip: fills 24% at -5%, same mean10 Oct 08:40
#4H3000 trailing stop vs fixed TP
killedhooked-v4/dead-endsanalyst2trail 50% after 20x = 19x < fixed TP30 25.810 Oct 08:40