Skip to content

Pattern Recognition

Models trained on every recorded launch and tested on launches they never saw.

897,725launches learned from · system 4 of 6

Illustration, generated with AI
Reads
Recorded launches with what became of them a day later.
Hands on
The pattern a launch matches and how launches with it ended.
Used in
MeTerm, Network
Runs
Trained each night at 08:00 UTC
Keeps
Every model version with its test
Access
Account, Free check

How it works

5 stations, from what it reads to what it hands on.

  1. 01

    Labels

    What became of each recorded launch within 24 hours.

  2. 02

    Features

    What the dev, the first buyers and the first minutes showed at the reading.

  3. 03

    Train

    Regression and tree ensembles on earlier days only.

  4. 04

    Test

    Scored on later days the models never saw, day by day.

  5. 05

    Gate

    Three nightly runs above the base rate and the best single rule before anything shows.

Measured

Counted from the recording and the research runs. Each figure names its date and source.

Skill per held-out day, crash within 24 h read at minute 5Brier skill score
Tree ensembleLogistic regression
0.800.53
31 Oct3 Nov6 Nov9 Nov12 Nov15 Nov
Show the 17 values
DayLaunchesTreesRegression
31 Oct 20267,9650.7540.709
1 Nov 20267,5930.6960.591
2 Nov 20268,8320.7150.651
3 Nov 20269,3820.7710.720
4 Nov 202611,7550.6760.654
5 Nov 20268,8590.7180.634
6 Nov 20269,4200.7400.680
7 Nov 20268,0270.7460.654
8 Nov 20268,6380.7170.622
9 Nov 20269,1530.7530.695
10 Nov 20268,2110.7470.682
11 Nov 20268,1230.6830.631
12 Nov 20268,2250.6490.585
13 Nov 20268,7310.7200.666
14 Nov 20267,8470.6640.667
15 Nov 20267,1700.5810.600
16 Nov 20264,9040.6180.563
Walk-forward: each day scored by models trained only on the days before it. Trees beat the regression on 14 of 17 days and lost on 0. Skill is the Brier skill score: 0 means no better than the base rate, 1 means every launch called right.
Every class on launches it never sawholdout
ClassLaunches testedBase rateSkillHit rate
Crash within 24 h, read at minute 263,5284.1%0.5988.7%
Crash within 24 h, read at minute 555,2473.8%0.6383.7%
Crash within 24 h, read at minute 1544,7662.5%0.6890.4%
Crash within 24 h, read at minute 6034,4311.0%0.5575.2%
Dev sells out, read before launch147,56890.7%0.1598.1%
Measured 18 Nov 2026 on the newest days, held out of training. Hit rate: share that came true among the launches ranked highest, as many as catch a quarter of all cases. Skill is the Brier skill score: 0 means no better than the base rate, 1 means every launch called right.
Similar launches: forecast against outcome50 nearest
Forecast 0%Forecast 100%

Up: share that crashed. Dashed: forecast equals outcome.

Show the 6 values
Forecast bandLaunchesForecastCrashed
0% to 1%56,0290.4%0.1%
1% to 3%1,6372.2%2.4%
3% to 10%1,4806.0%7.0%
10% to 30%1,43119.3%23.4%
30% to 60%1,40643.0%45.4%
60% to 100%1,90576.4%83.5%
Each held-out launch gets the crash share of its 50 most similar earlier launches. 63,888 launches, measured 19 Nov 2026. Close to the diagonal means the share can be taken at its word; it is shown as a reason, the trees give the rate.

Checks

The rules it holds before anything it finds reaches a page, and where each stands.

  • HoldoutThe newest days are never in training. Each class is scored on them with the Brier skill score.Every night
  • Gate3 runs in a row with skill above zero on held-out days and better than the best single rule.Required before a class shows
  • ParityThe live code must compute the same features and scores as the research code.Trees match to 1e-9; features 98.6% to 99.99% identical
  • Drift watchEach live class is scored on the day before. Three days under half its holdout skill raise an alarm.Every night since 19 Nov 2026
  • Similar CoinsA share is shown only with enough matches and with its interval.Shown to members
  • Signal ScorecardShown only above the minimum sample. Rules are versioned; nothing is overwritten.Shown to everyone
  • Auto CalibrationA proposal passes a gate and a shadow run before staff approve it. Every version stays on record.Shown to everyone

Limits

What it cannot see. Read these before you lean on a result.

Similar Coins
Similar is not the same. No probability for a single coin until the holdout test passes.
Signal Scorecard
Past precision is no forecast for one coin.
Auto Calibration
Calibration follows the past market. A new pattern needs time to show.

Changes

Dated, newest first. Every correction to data is on the corrections page too.

  1. Nightly run trains five-seed tree ensembles beside the regression and checks every class for drift.
  2. Class dev_exit went live in the check.
  3. Dev exit record live in the check after its holdout test.
All corrections