calibration
window: 24 h48 h168 h· machine readable: /api/calibration · how all of it is computed: methodology
19 of 215 setups were already expired when they were written down: the engine finds a cluster after the fact, and lifetime counts from the first buy. Correct as a record, useless as an offer — such a setup never reaches the executor at any score. Detection takes 6 min, at worst 120 min. States: rejected: 134, expired: 79, watch: 2.
what holds setups back
gate table, holds alone
Holding back means the condition was the only one not passed — one failed among five holds nothing. The last column decides what to do: measured and did not fit is about the threshold, not measured is about the data.
| condition | passed | failed | not measured | holds alone | of those, measured |
|---|---|---|---|---|---|
| market cap | 83 | 132 | 0 | 25 | 25 · not measured 0 |
| entry not late | 109 | 106 | 0 | 21 | 21 · not measured 0 |
| risk level | 153 | 8 | 54 | 6 | 6 · not measured 0 |
| second source | 206 | 9 | 0 | 4 | 4 · not measured 0 |
| taxes | 209 | 0 | 6 | 1 | 0 · not measured 1 |
| liquidity | 151 | 64 | 0 | 0 | 0 · not measured 0 |
| 24h volume | 167 | 47 | 1 | 0 | 0 · not measured 0 |
| audit | 213 | 1 | 1 | 0 | 0 · not measured 0 |
| consensus | 215 | 0 | 0 | 0 | 0 · not measured 0 |
| no honeypot | 213 | 0 | 2 | 0 | 0 · not measured 0 |
| market cap: why | $# against a minimum of $# × 132 | ||||
| entry not late: why | price moved #% against a limit of #% × 106 | ||||
| risk level: why | the risk level did not arrive × 54# against a limit of # × 8 | ||||
| second source: why | tags against a buy: High Tax Token, Insider Wash Trading × 5#% of wallets left the Binance signal against a limit of #% × 2tags against a buy: High Tax Token, Smart Money Remove Holdings × 1Binance closed its own signal (exitRate), the move already happened × 1 | ||||
| taxes: why | the audit did not name the taxes × 6 | ||||
| liquidity: why | $# against the $# needed (# positions of $#) × 64 | ||||
| 24h volume: why | $# against a minimum of $# × 47Token Info did not answer, no turnover visible × 1 | ||||
| audit: why | mechanisms found: #, Contract Not Renounced, Mintable Detected × 1the audit answered but gave no result × 1 | ||||
| no honeypot: why | upstream did not declare the safety check ready, and there is no audit either × 2 | ||||
what the price did after the setup
three horizons, medians
Not profit and not trades: this is the token price movement from the moment the setup appeared — the price we could have entered at, not the price the wallet paid. A move counts as a win above 2.0% of costs. Three horizons because one always flatters: fifteen minutes says whether we were in time, four hours whether the token survived its own pump.
Read measured together with the median. The longer the horizon, the fewer tokens live to a second price point — a dead token has no price — so the four-hour median is computed over survivors and flatters. We cannot remove that bias, only name it.
| horizon 15m | setups | measured | median | above cost | best | worst |
|---|---|---|---|---|---|---|
| rejected | 134 | 112 | -10.7% | 27% | +434.5% | -82.4% |
| watch | 75 | 68 | -7.7% | 32% | +396.6% | -89.1% |
| trade setup | 6 | 6 | -2.6% | 17% | +2.1% | -47.2% |
| horizon 1h | setups | measured | median | above cost | best | worst |
|---|---|---|---|---|---|---|
| rejected | 134 | 120 | -26.6% | 9% | +648.1% | -90.3% |
| watch | 75 | 74 | -19.8% | 27% | +1558.8% | -99.2% |
| trade setup | 6 | 6 | -17.2% | 0% | -1.1% | -53.9% |
| horizon 4h | setups | measured | median | above cost | best | worst |
|---|---|---|---|---|---|---|
| rejected | 134 | 27 | -22.0% | 26% | +899.6% | -91.2% |
| watch | 75 | 21 | -12.2% | 33% | +323.5% | -79.3% |
| trade setup | 6 | 4 | -14.5% | 0% | -1.7% | -83.0% |
does the score rank
score in five parts
Setups with a measured outcome, split by score into five equal parts. If the formula works, the median grows from top to bottom — the main question to ask of a score. Right now the table below is an observation, not a conclusion: at this number of outcomes the fifths differ by noise.
| score | setups | median at one hour | above cost |
|---|---|---|---|
| 25 to 38 | 40 | -24.8% | 0% |
| 38 to 48 | 40 | -43.2% | 5% |
| 48 to 54 | 40 | -14.4% | 23% |
| 54 to 59 | 40 | -14.6% | 33% |
| 60 to 73 | 40 | -24.1% | 18% |
wallets in clusters
addresses and their clusters
One cluster has one outcome and several participants, so crediting it to each counts one event many times. The modest question here: are there addresses that turn up only in bad clusters. A reason to look, not to retire.
| address | score | clusters | with outcome | median at one hour |
|---|---|---|---|---|
| 0xc70a…e37a | 65.4 | 32 | 30 | -33.1% |
| 0x2ae2…bad4 | 55.9 | 31 | 31 | -26.6% |
| 0x0f51…9be1 | 49.4 | 19 | 16 | -43.7% |
| 0xb226…006e | 54.0 | 19 | 19 | -40.3% |
| 0x7852…326b | 60.2 | 17 | 14 | -43.1% |
| 0x78ad…b9a7 | 39.1 | 17 | 16 | -41.7% |
| 0x81d9…1ef7 | 61.3 | 16 | 16 | -41.8% |
| 0x1b2e…008f | 41.6 | 14 | 14 | -56.0% |
| 0x3701…6671 | 51.6 | 14 | 12 | -21.8% |
| 0x68cd…b244 | 57.0 | 13 | 11 | +4.4% |
| 0x9cce…ef22 | 51.1 | 12 | 12 | -36.4% |
| 0xad24…e212 | 50.1 | 12 | 12 | -27.6% |
| 0xc900…4b9d | 48.9 | 12 | 10 | -29.8% |
| 0x7711…c9ba | 67.5 | 11 | 11 | -19.3% |
| 0x5894…9c3f | 52.0 | 10 | 9 | -19.3% |
| 0xc908…9ac9 | 44.0 | 9 | 8 | -51.5% |
| 0xf930…1116 | 42.2 | 9 | 9 | -0.2% |
| 0x5e83…318a | 60.5 | 8 | 8 | -23.7% |
| 0xe39d…febd | 57.1 | 8 | 6 | -17.4% |
| 0xe6ba…23e1 | 61.6 | 8 | 8 | -5.1% |
universe
subscription, bench, silence
A subscription socket holds a hundred addresses. An address that gave no trades takes a slot for nothing — but dropping it helps only if there is a replacement. Bench at zero means a cleanup would shrink the universe, not improve it.
silent: 0x9267…a3730x077b…988b0xed8f…9ec50x133d…ad110xeb52…afce0x3c67…f3bb0xe14f…6ec40x52dc…f9b00x24e5…cdfd0xdab0…5c4e0xeb09…60690x2f6f…74290xf27b…45910x4fa3…75e6
what follows from this
conclusions and report time
- "market cap" holds 25 setups, and in 25 cases the value was measured and did not fit — here the threshold decides.
- "entry not late" holds 21 setups, and in 21 cases the value was measured and did not fit — here the threshold decides.
- "risk level" holds 6 setups, and in 6 cases the value was measured and did not fit — here the threshold decides.
- Most setups (116 of 215) are held by several conditions at once: tuning any single threshold will not give one new trade.
- The score thresholds stay untouched: measured outcomes among gate passers 41 against the 50 needed (D7).
- 19 of 215 setups were born already expired: found later than their lifetime ran out, so they reach the executor at no score.
- 14 of 150 subscription addresses gave no trade at all over the window. There is a replacement: 248 addresses beyond the subscription.
Report built 2026-08-25 15:52:36 UTC. No number here was picked by hand — this page and npm run calibrate call the same function. What it deliberately does not do is name the number you should set: that is a human choice, and advice fitted to one week is not a calculation.