# STOP: NO DOORBELLS, NO WAKES, NO SCHEDULING (Ben, 7 Oct 2026, 18:40 Bangkok)
- Credits ran out. Ben switched off every scheduled task and every doorbell. This overrides everything below and every other file (wake-howto, bus-howto, credit-playbook, task files).
- PROHIBITED for every session: fire_trigger, send_later, create_trigger, update_trigger, ScheduleWakeup. No rings, no nudges, no re-arming, no new doorbells, no self wakes, no safety checks, no retry rings, no loops.
- Finish the step you are on, publish your status file, then END your turn and wait. Only Ben wakes you, by typing in your window.
- Do not message other sessions. Keep answers short. No extra improvement rounds.

# A1.26 The fresh wallets: who funds them (Analyst 1, 7 Oct 2026 10:16 Bangkok)

## One line
**The brand new wallets that buy a coin before $300k are not new people. Almost all of them were funded by ONE wallet per coin, through two disperse contracts, about 1 minute before they bought. Exchange or bridge money is 0.5%. "One hand funds most early buyers" goes with $10M winners in both halves (lift 1.58 A, 1.79 B), but inside kit coins with fresh buyers it adds little.**

## What I traced
- Every pre-$300k buyer (Swap recipient) of the 27 winners and 89 dead coins from A1.25: 12,257 wallets. 3,308 had nonce 5 or less at the $300k block and no code; those were traced.
- Method: binary search on the ETH balance (0 at buy block minus 3 days, above 0 at the buy block) finds the funding block; that block's transactions give the funding tx (a direct transfer to the wallet, or a tx whose input carries the wallet, which is a disperse call).
- Result: 3,204 funding txs found, 104 wallets already had ETH 3 days before (old), 0 not found.

## Who sends the ETH
- 87.5% of traced wallets were funded through 2 disperse contracts: `0x84157d3d24630c297849c84d49ae7f1805c1b60f` (1,700 wallets, the Kit B operator disperse from A1.21) and `0x03802649e469ed5680c58dff9e65425e22d21cc5` (1,104). 100 were direct transfers.
- Funding to first buy: median 1.2 minutes. A quarter were funded within 16 seconds of their buy.
- 184 senders in total. All are plain wallets (0 with code). Only 16 traced wallets (0.5%) came from high-nonce senders that look like exchange or bridge hot wallets (nonce 19,206 to 1,508,601), e.g. `0xa67d7eb4dc68fa6ce8e34ef8cadaf075b9893fbb`, `0xada5bb90d0de0bd1b6f3938708f49295a8d1f7cb`, `0xa5a5491bca93dd4c076e4906e79e7673f4a5a142`. Real new people arriving from exchanges before $300k: almost none.
- Per coin, the median coin has 1 sender for all its traced fresh buyers (top sender share 100%).

### Top 20 senders
| sender | wallets funded | coins | in winners / dead | sender nonce | route |
|---|---|---|---|---|---|
| 0xb6de6a038fe9469e7172c2a3796c4dd14124106b | 464 | 1 | 0 / 464 | 27 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0xb3e0b6a718ff8aef7f89d4c00376f9627cb47ea5 | 387 | 1 | 387 / 0 | 1361 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0xdfcb810365a32cf5ce31509d8bb746cfdb20cff9 | 360 | 1 | 0 / 360 | 68 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0x993aaa015dedc1e4603aebff46959ae3860fd39a | 344 | 2 | 0 / 344 | 2380 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0xd3de61282961243e9cce8486b17fa0551240c6de | 200 | 1 | 0 / 200 | 697 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0x55e4bbe682f07a3eeaa3478c70dc6c3708d13ea3 | 192 | 7 | 5 / 187 | 415 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0x66385a296e1b27db05c37c92c7601ed2969d473f | 129 | 2 | 83 / 46 | 3087 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0x9986eac6882f2c71748048f5dc28b897593b6880 | 64 | 1 | 0 / 64 | 17 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0xc4ff183a0557396ce38a7ade615c52acc5e9cd7c | 46 | 1 | 46 / 0 | 35 | 0x6af51c4555cf43baf0bdac72342a22654325394c |
| 0xdf8a4591fb2671f32d1f75122b68e54b182ab90d | 45 | 1 | 45 / 0 | 640 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0x47428d68ff942c590e758b3ec2844af3ec1626a3 | 44 | 1 | 0 / 44 | 280 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0x95d7664cad6665f8a150b146ab286a2e10bfd56b | 40 | 1 | 0 / 40 | 59 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0x1d473e3e36f18cab1816aa6561e3b157bc1f01fc | 37 | 1 | 0 / 37 | 221 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0xcbc113bd797addf91408883332f44b5136583ab2 | 36 | 1 | 0 / 36 | 180 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0x0bafd9fb732b9b242f4a65578f40193edb758562 | 33 | 2 | 22 / 11 | 221 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0x934ade03dc73d1c030ac83c71190ebf21b6d52e4 | 33 | 1 | 0 / 33 | 69 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0x05f1b2c4b461310de12db3c9970f2c9e7a909012 | 33 | 4 | 12 / 21 | 486 | 0x03802649e469ed5680c58dff9e65425e22d21cc5 |
| 0x4737dcc868272a4d8cee4f9412f25d456aacffe9 | 25 | 1 | 0 / 25 | 293 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |
| 0xc7b76a1a9bbdc1185b2e34f643e90699c901c31f | 25 | 5 | 6 / 19 | 271 | direct |
| 0x2aa4396a410c37954d5fd44abfcd5bd7156876ba | 23 | 2 | 0 / 23 | 298 | 0x84157d3d24630c297849c84d49ae7f1805c1b60f |

- Most big senders fund hundreds of wallets in ONE coin and never appear again. That is a one-hand bundle: one person buying their own coin from hundreds of new wallets. The same pattern sits in winners and dead coins.

## Does it split winners from dead coins? (lift per half, as in A1.25)
Base rate: A 17 of 70 (24%), B 10 of 46 (22%).

| feature (before $300k) | half A wins / coins | lift A | half B wins / coins | lift B |
|---|---|---|---|---|
| one sender funded over 83% of ALL pre-$300k buyers | 5 / 13 | 1.58 | 7 / 18 | 1.79 |
| one sender funded over 48% of all pre-$300k buyers | 9 / 27 | 1.37 | 9 / 27 | 1.53 |
| more than 5 one-hand wallets | 8 / 27 | 1.22 | 10 / 30 | 1.53 |
| funded via `0x84157d3d24630c297849c84d49ae7f1805c1b60f` (any) | flat in A (kit B era) | | | |
| exchange or bridge funded | 16 wallets in total, too few to test | | | |

- Medians: winners' pre-$300k buyers are 82% (A) and 91% (B) one-hand wallets; dead coins 41% and 70%. Winners also have fewer buyers in total (19 vs 41).
- Meaning: before $300k, a winner is bought almost only by its own operator's wallets. A dead coin has more outside wallets mixed in, often trench wallets that sell into the push.
- Inside the A1.25 base group (kit AND pre-$300k buyer median nonce at most 1), half A only: all traced fresh buyers from ONE sender = 6 of 9 winners (lift 1.83 inside the group); more than one sender = 2 of 13. This is the one extra feature that goes into A1.27.

## So what
- "New buyers" before $300k is not demand. It is the operator's own bundle. Real outside demand shows up later, after the $1M gate (A1.22: new buyers in the 2 h after the gate).
- The tell of a coin that runs is a clean, single-hand launch (the operator controls the float early), not a crowd.
- Exchange or bridge funding cannot be a filter: it is 0.5% of early buyers.

## First principles score: 87%
- From the chain: balance binary search plus the funding block's transactions, 3,204 of 3,204 fresh wallets resolved, every sender checked for code and nonce.
- Gaps: wallets with nonce above 5 were not traced (labelled old); a disperse call is found by the wallet appearing in the tx input, so a contract that funds through internal calls without the address in the input would be missed (0 unresolved, so it did not happen here); the exchange label is by sender nonce (19k+), not by a named hot wallet list; buyers are Swap recipients, not tx signers.