---
name: screen-log-outcomes
description: The daily scoring of every open row in [[screen-log]], runs of 1 and 2 Sep 2026 only. Screen-time facts stay in screen-log; later measurements land here. FULL at the 32KB cap — runs from 3 Sep 2026 are in [[screen-log-outcomes-2]]. Read both.
sources: [cowork]
aliases: [outcome-updates, screen-outcomes, scoring-log]
---

# Screen log outcomes

- [stated] Ben asked on 31 Aug 2026 for a database that keeps updating itself so calls get scored instead of forgotten. This is the scoring half of it.

## Rules carried over from [[screen-log]]
1. Screen-time columns are NEVER edited. Only this file grows.
2. Every open row gets a dated line every run, including "nothing changed".
3. Multiple is always mc_now divided by mc_screen, and max_mc_since is the running high.
4. A falsifier that cannot be measured is written UNMEASURED with the reason, never silently skipped.

## Dead thresholds
liquidity under $5,000 or volume h24 under $500 closes the row as OUTCOME: dead.

---

## 1 Sep 2026, day 1 for all eight open rows

Source: DexScreener tokens endpoint, base-token pairs only, liquidity and volume summed, marketCap from the highest-liquidity pair. DexScreener 403s the container without a browser User-Agent; send one.
robinhoodchain.blockscout.com was behind a Cloudflare interstitial for this entire run, so **no holder counts were obtainable**. Every holder-based falsifier is UNMEASURED today.

| Ticker | mc_screen | mc_now | multiple | max_mc_since | liq | vol24 | alive |
|---|---|---|---|---|---|---|---|
| SCL | $287,730 | $161,240 | 0.56x | $287,730 | $52,375 | $202,275 | yes |
| BOW | $7,747,316 | $6,046,340 | 0.78x | $7,747,316 | $416,826 | $2,979,486 | yes |
| DICE | $778,508 | $538,184 | 0.69x | $778,508 | $97,313 | $455,077 | yes |
| BACKED | $1,053,287 | $1,288,548 | **1.22x** | $1,288,548 | $194,601 | $1,118,608 | yes |
| RAM hub, 2nd pass | $5,945,606 | $8,323,806 | **1.40x** | $8,323,806 | $1,245,301 | $12,145,190 | yes |
| RAM hub, 4th pass | $5,762,199 | $8,323,806 | **1.44x** | $8,323,806 | $1,245,301 | $12,145,190 | yes |
| ARC | $342,605 | $546,291 | **1.59x** | $546,291 | $68,169 | $474,553 | yes |
| RAM satellite, auth row | see note | $6,431,314 | not scored | n/a | $1,067,618 | $31,627,499 | yes |

No row is dead. No 7d, 30d or 90d mark is due: every screen was written 31 Aug 2026.

### SCL `0xBe92b334E045Bbfd292a28e54f8C75aF2FC07bBE`
- h24 -46.77%, h6 -27.99%, h1 buys/sells 18/9. mc_now $161,240 against $215,478 quoted in the ARC comps only 18 hours earlier
- falsifier 1 already FIRED 31 Aug 06:49 (dev bridged remainder, buybacks stopped)
- falsifier 2 UNMEASURED: price is falling, holder count unreadable, Blockscout gated
- falsifier 3 (position manager not deployed) not due until 30 Sep
- **the NO is right so far, 0.56x at day 1**

### BOW `0x451b42A15100C340CA12F7c66DE06fac5EA2D751`
- h24 -19.53%, buys/sells h24 1,902/1,829
- www.longbow.cash/api/stats: totalSupplied $76,473.41 (was $72,067.88), tvlWithCollateral $238,345.52, **totalBorrowed $41,247.82, up 102.8% in one day**, utilisation 58.56% (was 33.55%), depositorCount **27** (was 25), openInterest $196,016.91 across 40 positions
- falsifier 1 NOT fired: rewardRate() on `0xEba502e1177f5fad9432c2deb273569D8381e01A` re-read via RPC at 1,267,997,775,689,893,298,709 raw = 0.001268 USDG/sec, IDENTICAL to screen; periodFinish still 1788581524 = 5 Sep 2026, so the refund test is 4 days away
- falsifier 2 (depositorCount under 100 or borrowed under $50,000 on 21 Sep) not due, and both legs would fire today
- **THE BORROWING GREW ALMOST ENTIRELY INSIDE THE DISABLED MARKET.** PONS/USDG borrow $34,144.53 of $41,247.82 = **82.8% of all borrowing, up from 71.7%**, still enabled:false, supply $37,507.74. The only other debt on the protocol is SPY/USDG $3,988.04 and ETH/USDG $1,410.07
- falsifier 3 is written as PONS borrow FALLING, so it did not fire. The exposure it was written to catch got worse. Consider rewriting it as a concentration test, not a direction test: PONS above 75% of total borrowing while enabled stays false
- 2 net new depositors against a doubling of debt is the day-1 read: the lending side is growing on one memecoin collateral position, not on users

### DICE `0x3F9f0b6073Ee8c495Aed96869AF31850fED40FeB`
- h24 +72.51% but h1 -23.77%, h6 buys/sells 119/216, h24 806/1,110. The four-integration announcement pop is being distributed into
- falsifiers 1 and 2 not due until 21 Sep. Falsifier 3 (new PonsLaunchLocker tranche to `0x4acd2c88a239a924e47fc4995114ca1bb0ca3cad`) not re-measured
- **the NO at $778,508 is right so far, 0.69x at day 1**

### BACKED `0x7168563B0E70124f0C7c0cF2F13a8D1861BAf4A5`
- **NEW HIGH.** h24 +69.5%, buys/sells h24 1,481/1,055, h6 203/145, h1 34/23. Main pool liq $162,837, vol24 $1,037,310
- falsifier 1 (main-pool 7d average under $50,000/day) nowhere near: today alone was $1,037,310, which at the 3% tax adds roughly $31,100 to the vault in one day
- **FALSIFIER 3 COUNTER STARTED, day 1 of 14.** DENAR `0x3786728a2c49C4617Bf4FE5BD82b90B6B0dF5508` mc $1,750,815, liq $152,564, vol24 $2,274,406, versus BACKED $1,288,548. The team's second token is above its first. Fires only on 14 straight days. Reset the counter the first day BACKED is back on top
- falsifier 2 (reserve NAV under $250,000) not due until 30 Sep. Reserve not re-measured today
- **the YES at lottery size is ahead by 1.22x at day 1**

### RAM, satellite authenticity row `0x5173D45A1191eE33cBB7D8c7e65f21B04eD54802`
- satellite reads mc $6,431,314 across 30 base pairs, liq $1,067,618, vol24 $31,627,499
- **do NOT read that as a 4.7x from the ~$1.37M quoted at screen.** It is exactly the trap that row was written to name: satellite market cap is bridged supply times price, and bridged supply itself grew (11,856,596 to 13,358,933 on 31 Aug alone). Score RAM on the hub rows only

### RAM, hub, second pass `0x555570a286F15EbDFE42B66eDE2f724Aa1AB5555`
- hub mc $8,323,806, FDV $39,219,931 (from $28,023,821), price $0.1655, liq $1,245,301, vol24 $12,145,190. h24 +522%, h6 -18.9%, h1 buys/sells 28/43
- **THE BUY TRIGGER ON THIS ROW FIRED, on the second of its two legs.** "Ramses' share of Robinhood Chain DEX fees crosses 1.5% from 0.53%": DefiLlama gives Ramses CL V2 Robinhood fees **$264,404/day** against Robinhood Chain total fees **$13,823,321/day = 1.913%**. Ramses is now rank #7 on the chain, behind Uniswap V4 $4,770,656, Pons V2 $4,426,890, Uniswap V3 $1,831,628, GMGN $1,316,399, Robinhood Chain $1,070,833, Pons V1 $635,406, and ahead of Uniswap V2 $248,802
- Ramses CL V2 all chains today: $708,001, of which Hyperliquid L1 $443,286, Robinhood $264,404, Polygon $311, Arbitrum $0.43
- the price leg did NOT fire: no retrace to $0.04 to $0.06, price went the other way
- falsifier 1 (chain DEX fees under $6M/day) not fired, $13.82M. Falsifier 2 (Ramses RH fees under $30k/day) not fired and moved 4.1x the other way. Falsifier 4 (circulating above 80M) not fired: mc/price implies 50.3M circulating and total supply is still 236,934,876
- **CAVEAT, and it is the one that matters: this firing is not a buy signal.** This pass was SUPERSEDED. The fourth pass said NO on the SUPPLY side (Gate 1, coverage 17.9x, $20.98M of exercisable xRAM against roughly $4.4M of free float), not on the fee side. A fee trigger cannot close a supply gate

### RAM, hub, fourth and final pass
- mc $8,323,806 vs $5,762,199 at screen = **1.44x**, FDV $39,219,931 vs $27,159,358
- **the NO is behind by 1.44x at day 1 and this row is the live test of the coverage gate**
- none of the three flip conditions was re-measured today. Next run must query the xRAM vault `0xae6d5fce541216bda471d311425b5412d9f1deb9` net RAM flow (it was inverted 106:1 OUT on 31 Aug against a 2:1 IN baseline) and the deployer `0x20d630cf1f5628285bfb91dfac8c89eb9087be1a` outflows. That flow number, not the fee number, decides this row

### ARC `0xce845443428867d271929bd739148B85524B536a`
- **NEW HIGH, and the widest gap against a verdict of any open row.** h24 +73.05%, h6 -12.39%, buys/sells h24 1,218/997, h1 19/19
- falsifier 2 (liquidity under $30,000) NOT fired, $68,169. Falsifiers 1 and 4 not due until 14 and 30 Sep. Falsifier 3 (@ArcLiquidity silent 7 days) not measured, no X call was made this run
- **CORRECTION to the ARC screen row, carried in from the operator-graph run of the same day.** The screen listed `0xf70da97812cb96acdf810712aa562db8dfa3dbef` first among ARC's spray-bot top-15 buyers at 270 tokens. Dune degree test on 1 Sep: that address has sent to **145,708 addresses and was the first funder of 105,696 of them**. It is infrastructure, a router or hot wallet, not a spray bot and not a trader. Drop it: **10 of the top 15, not 11.** The other ten are untested and still stand
- the number that misled: "early into 270 tokens" reads as a spray bot only if the address is a trader. Out-degree was never checked on the buy side. **Apply the operator-graph degree filter to BUYER wallets too, not just funders.** That rule was written for funders and the same failure mode was sitting one column over

## Standing notes for the next run
- DexScreener needs a browser User-Agent from this container or it returns 403 on every call. Not documented in [[data-access]] before today
- Blockscout RH v2 was Cloudflare-gated for this whole run, consistent with the intermittent-limiting correction already in [[data-access]]. Holder-count falsifiers need a fallback: read Transfer-event unique addresses off the RPC instead
- BOW, BACKED and DICE all expose public product APIs that answer their own falsifiers without any explorer. Prefer those
---

## 2 Sep 2026, day 2 for the eight 31 Aug rows, day 1 for RIPE

Source: DexScreener tokens endpoint, base-token pairs only, liquidity and volume summed, marketCap from the highest-liquidity pair. **RH chain RPC found and working: `https://robinhood-rpc.publicnode.com`, chainId 0x1237 = 4663. It 403s without a browser User-Agent, same as DexScreener. Add to [[data-access]].** Base RPC `https://mainnet.base.org` needs no UA.

| Ticker | mc_screen | mc_now | multiple | max_mc_since | liq | vol24 | alive |
|---|---|---|---|---|---|---|---|
| SCL | $287,730 | $68,998 | **0.24x** | $287,730 | $31,234 | $94,401 | yes |
| BOW | $7,747,316 | $6,612,519 | 0.85x | $7,747,316 | $441,983 | $2,476,037 | yes |
| DICE | $778,508 | $256,734 | **0.33x** | $778,508 | $61,959 | $270,353 | yes |
| BACKED | $1,053,287 | $1,021,020 | 0.97x | $1,288,548 | $165,358 | $408,888 | yes |
| RAM hub, 2nd pass (FDV basis) | $28,023,821 | $44,507,209 | 1.59x | $44,507,209 | $1,415,168 | $4,402,207 | yes |
| RAM hub, 4th pass (FDV basis) | $27,159,358 | $44,507,209 | **1.64x** | $44,507,209 | $1,415,168 | $4,402,207 | yes |
| ARC | $342,605 | $91,761 | **0.27x** | $546,291 | $27,567 | $265,667 | yes |
| RIPE (cross-chain minted x price) | $48,300,000 | $115,472,589 | **2.39x** | $115,472,589 | $576,946 | $5,068,791 | yes |
| RAM satellite, auth row | see note | $9,598,264 | not scored | n/a | $1,118,311 | $14,383,213 | yes |

No row is dead. No 7d, 30d or 90d mark is due.

### FALSIFIER FIRED: ARC number 2
`0xce845443428867d271929bd739148B85524B536a`. **Liquidity $27,567.17 summed across 5 base pairs, below the $30,000 threshold. FIRED 2 Sep 2026.**
- h24 **-83.38%**, mc $91,761 from a day-1 high of $546,291. mc_screen $342,605, so **0.27x**
- buys/sells h24 841/799, h1 31/9, h6 96/62. Volume $265,667 = 9.6x the remaining liquidity, still churning
- **the NO was right, and the 1.59x day-1 print was a one-day head fake.** Recorded because the 1 Sep line called it "the widest gap against a verdict of any open row" and that reading inverted in 24 hours. The number that misled: a day-1 multiple on a 3-day-old token with $53k of liquidity is noise, not evidence
- falsifiers 1 and 4 not due until 14 and 30 Sep. Falsifier 3 (@ArcLiquidity silent 7 days) not measured, no X call this run

### SCL `0xBe92b334E045Bbfd292a28e54f8C75aF2FC07bBE`
- h24 -57.13%, h6 -12.62%, m5 buys/sells 0/0, h1 2/0. mc $68,998, liq $31,234, vol24 $94,401
- **0.24x at day 2, the best-performing NO in the book.** Falsifier 1 fired 31 Aug and nothing has contradicted it
- falsifier 2 (holders below 686 while price falls) UNMEASURED again, no holder source. Falsifier 3 not due until 30 Sep
- dead thresholds not yet hit but liquidity is 6.2x the $5,000 floor and falling

### BOW `0x451b42A15100C340CA12F7c66DE06fac5EA2D751`
- mc $6,612,519 = 0.85x, h24 +11.34%, h6 +40.21%, buys/sells h24 1,944/1,492
- **FALSIFIER 1 NOT FIRED AND IT MOVED THE OTHER WAY, HARD.** rewardRate() on `0xEba502e1177f5fad9432c2deb273569D8381e01A` raw **3,064,092,381,749,161,300,668**, which is **2.4165x** the screen figure of 1,267,997,775,689,893,298,709. On the screen's own scaling that is 0.003064 USDG/sec = **$264.73/day = $96,626/yr contracted**, up from $39,988/yr. Holder yield rises to about 1.46% of market cap per year from 0.52%
- **periodFinish EXTENDED from 1788581524 (5 Sep 2026 04:12 UTC) to 1788858349 (8 Sep 2026 09:05 UTC). The refund happened, three days early.** The falsifier's second leg is now dead until 8 Sep
- /api/stats: totalSupplied **$353,234.17** (was $76,473), tvlWithCollateral $481,044.66, totalBorrowed **$142,780.54** (was $41,247.82, so 3.46x in one day), utilisation 64.4%, depositorCount **31** (was 27), openInterest $644,935.99 across 186 positions, marketCount 53, bowStaked 144,140,077.68
- **THE CONCENTRATION MOVED, IT DID NOT IMPROVE, AND THE PROPOSED REWRITE WOULD HAVE MISSED IT.** Borrowing by market: wsNET/USDG **$83,375.95, enabled:false, 58.4%** · PONS/USDG **$50,870.56, enabled:false, 35.6%** · SPY/USDG $3,988.09 enabled:true · ETH/USDG $1,410.08 enabled:true · INDEX/USDG $942.52 enabled:false · NOTHING/USDG $605.00 enabled:false. **Disabled markets are 94.0% of all borrowing.** PONS alone fell from 82.8% to 35.6%
- **REWRITE FALSIFIER 3 AS AN AGGREGATE TEST, NOT A PONS TEST: disabled markets above 75% of total borrowing.** The 1 Sep note proposed "PONS above 75% while enabled stays false", which reads clean today at 35.6% while the actual exposure nearly quadrupled and simply moved to a new memecoin collateral, wsNET
- 4 net new depositors against a 3.46x rise in debt. The lending side is still growing on disabled memecoin collateral, not on users

### DICE `0x3F9f0b6073Ee8c495Aed96869AF31850fED40FeB`
- mc $256,734 = **0.33x**, h24 -51.75%, h6 -25.24%, buys/sells h24 639/726, h6 172/161
- the 31 Aug announcement pop has fully round-tripped and then some. The NO at $778,508 is right
- falsifiers 1 and 2 not due until 21 Sep. Falsifier 3 not re-measured

### BACKED `0x7168563B0E70124f0C7c0cF2F13a8D1861BAf4A5`
- mc $1,021,020 = 0.97x, back below the screen price after a 1.22x day. max_mc_since stays $1,288,548
- h24 -24.78% but h1 +9.94% and h6 +21.05%, buys/sells h24 626/505, h1 23/10. Volume $408,888, down from $1,118,608
- falsifier 1 (main-pool 7d average below $50,000/day) not fired: main pool did $408,888 today across all pairs, still 8x the threshold. At the 3% tax that is roughly $12,300 added to the vault today, down from $31,100
- **FALSIFIER 3, DAY 2 OF 14.** DENAR `0x3786728a2c49C4617Bf4FE5BD82b90B6B0dF5508` mc $1,747,294, liq $147,960, vol24 $1,000,174 versus BACKED $1,021,020. **DENAR is still above and the gap widened from 1.36x to 1.71x.** Counter continues
- falsifier 2 (reserve NAV under $250,000 on 30 Sep) not due, reserve not re-measured

### RAM, hub `0x555570a286F15EbDFE42B66eDE2f724Aa1AB5555`
- **BASIS CHANGE, READ THIS BEFORE QUOTING THE NUMBER. DexScreener's marketCap field for the hub flipped from a circulating basis to a full-supply basis between runs.** 1 Sep it printed $8,323,806 at $0.1655 (about 50.3M tokens). Today it prints $44,507,209 at $0.1882 (about 236.5M tokens, i.e. the full 236,934,876 supply), and marketCap now equals fdv. **That is a reporting change, not a 5.3x**
- **Score RAM on FDV from here, on both passes.** FDV $27,159,358 (4th pass screen) to $44,507,209 = **1.64x**. Price $0.1146 to $0.1882 = 1.64x, an exact match, which is the check that confirms the multiple is real and the printed-cap jump is not
- h24 +15.50%, h6 +5.61%, h1 buys/sells 5/30, liq $1,415,168, vol24 $4,402,207
- satellite `0x5173D45A1191eE33cBB7D8c7e65f21B04eD54802` reads mc $9,598,264, liq $1,118,311, vol24 $14,383,213. Still not scored, still the bridged-supply trap
- **THE DECIDING NUMBER IS STILL UNMEASURED.** xRAM vault `0xae6d5fce541216bda471d311425b5412d9f1deb9` net RAM flow was not queried: Dune capacity this run went to the operator graph, which produced the first genuine repeat operator and took priority. Deployer `0x20d630cf1f5628285bfb91dfac8c89eb9087be1a` outflows also unmeasured. **These two numbers decide this row and have now been deferred twice. Run them first next session, before the operator graph**
- the NO is behind by 1.64x at day 2 and the coverage gate has not been re-tested

### RIPE `0x4D3f37a965b21aB4122e92Dd41D2693E742c883b` (RH) + `0x2a0a59d6b975828e781ecac125dba40d7ee5ddc0` (Base)
- **THE WIDEST GAP AGAINST A VERDICT IN THE BOOK: 2.39x against a SELL at day 1.** Price $4.62 to **$10.98**, h24 +1482% on RH and +1376% on Base
- cross-chain minted supply re-measured on chain: RH `totalSupply()` **289,858.30** (was 210,706) + Base **10,226,770.92** (was 10,241,168) = **10,516,629.22 RIPE**. Real cap **$115,472,589**. Aggregators still wrong: DexScreener RH prints $3,178,012, Base prints $112,834,180
- **REVERSAL TRIGGER 1 NOT FIRED. Endaoment `0x9f20f25f037046721a292b19a486932ef390eaf9` holds 0 RIPE and 0 GREEN.** The toll booth still has never collected, on the day the token did +1482%
- **REVERSAL TRIGGER 2 IS THE ONE MOVING. GREEN outstanding both chains 988,336.25** (RH `0x355bb7f0f6c730e4460d620420a300fa08ff82f3` 590,125.30 + Base `0xd1eac76497d06cf15475a5e3984d5bc03de7c707` 398,210.95), up from 158,400 = **6.24x in one day**. That is 19.8% of the $5M threshold. If it holds this slope the trigger clears inside a week. Watch it daily
- reversal trigger 3 (price back to $0.57 to $1.44) not fired, price went 2.4x the other way
- **GATE 1 GOT WORSE, NOT BETTER.** Ledger `0x7e1d751d168f09761b88651a4c78c996354faeb1`: `ripeAvailForRewards()` **9,929,557** (was 9,978,814) + `ripeAvailForBonds()` 1,000,000 = 10,929,557 RIPE of queued mint authority = **$120.0M** at $10.98, against nameable bid $576,946. **Coverage 208x, up from 153x.** The price rose faster than the mint authority was consumed, which is the mechanism the gate exists to catch
- honest read: the SELL is 2.39x behind and the two things that would justify it (Endaoment collecting, GREEN scaling) are one still-zero and one moving fast. The gate has not been invalidated, but a row that is 2.39x behind at day 1 is the single most useful entry in this file for calibration

## Standing notes added 2 Sep 2026
- **RH chain RPC: `https://robinhood-rpc.publicnode.com`.** Send a browser User-Agent or every call returns 403. This closes the hole the 1 Sep run flagged: contract reads no longer depend on Blockscout being un-gated
- `pycryptodome` is not preinstalled in the container. `pip install pycryptodome --break-system-packages` for keccak selectors
- A day-1 multiple on a token under $1M with under $100k of liquidity is noise. ARC went 1.59x then 0.27x in 24 hours. **Do not let a single day-1 print be read as a verdict being wrong**
- Product APIs beat explorers again: BOW's /api/stats and /api/markets answered three falsifier legs in two calls