Every model, tested on launches it never saw.
A young record: the first model went live on 18 Nov 2026 and each test covers weeks of launches, not years. Measured by CON, not audited by a third party.

Cards
Skill is the Brier skill score: 0 is no better than the base rate, 1 calls every launch right. Hit rate: share that came true among the launches ranked highest, as many as catch a quarter of all cases.
crash24_m2
Crash within 24 h, read at minute 2
- Skill
- 0.59
- Hit rate
- 88.7%
- Base rate
- 4.1%
- Tested on
- 63,528 launches
- Trained on
- 148,232 launches
- Walk-forward
- trees better on 16 of 17 days
In the gate: shown nowhere until it passes
crash24_m5
Crash within 24 h, read at minute 5
- Skill
- 0.63
- Hit rate
- 83.7%
- Base rate
- 3.8%
- Tested on
- 55,247 launches
- Trained on
- 128,909 launches
- Walk-forward
- trees better on 14 of 17 days
In the gate: shown nowhere until it passes
crash24_m15
Crash within 24 h, read at minute 15
- Skill
- 0.68
- Hit rate
- 90.4%
- Base rate
- 2.5%
- Tested on
- 44,766 launches
- Trained on
- 104,453 launches
- Walk-forward
- trees better on 14 of 17 days
In the gate: shown nowhere until it passes
crash24_m60
Crash within 24 h, read at minute 60
- Skill
- 0.55
- Hit rate
- 75.2%
- Base rate
- 1.0%
- Tested on
- 34,431 launches
- Trained on
- 80,339 launches
- Walk-forward
- trees better on 7 of 17 days
In the gate: shown nowhere until it passes
dev_exit
Dev sells out, read before launch
- Skill
- 0.15
- Hit rate
- 98.1%
- Base rate
- 90.7%
- Tested on
- 147,568 launches
- Trained on
- 344,323 launches
- Walk-forward
- trees better on 15 of 17 days
Live in the check since 18 Nov 2026
A rate describes a group of launches, never one coin. No class forecasts winners; that would read as a buy signal, so none is built.
How they are tested
The rules every class meets before it reaches a page.
- HoldoutThe newest days are kept out of training and used only to score.Every night
- Walk-forwardEach past day is scored again by models trained only on the days before it.17 days per class
- Gate3 nightly runs in a row with skill above zero and better than the best single rule, then review.Before a class shows
- ParityThe live code computes the same features and scores as the research code.Trees to 1e-9, features 98.6% to 99.99%
- Drift watchEach live class scored on the day before; three days under half its holdout skill raise an alarm.Every night
Results
The figures behind the cards.
| Class | Launches tested | Base rate | Skill | Hit rate |
|---|---|---|---|---|
| Crash within 24 h, read at minute 2 | 63,528 | 4.1% | 0.59 | 88.7% |
| Crash within 24 h, read at minute 5 | 55,247 | 3.8% | 0.63 | 83.7% |
| Crash within 24 h, read at minute 15 | 44,766 | 2.5% | 0.68 | 90.4% |
| Crash within 24 h, read at minute 60 | 34,431 | 1.0% | 0.55 | 75.2% |
| Dev sells out, read before launch | 147,568 | 90.7% | 0.15 | 98.1% |
Show the 17 values
| Day | Launches | Trees | Regression |
|---|---|---|---|
| 31 Oct 2026 | 7,965 | 0.754 | 0.709 |
| 1 Nov 2026 | 7,593 | 0.696 | 0.591 |
| 2 Nov 2026 | 8,832 | 0.715 | 0.651 |
| 3 Nov 2026 | 9,382 | 0.771 | 0.720 |
| 4 Nov 2026 | 11,755 | 0.676 | 0.654 |
| 5 Nov 2026 | 8,859 | 0.718 | 0.634 |
| 6 Nov 2026 | 9,420 | 0.740 | 0.680 |
| 7 Nov 2026 | 8,027 | 0.746 | 0.654 |
| 8 Nov 2026 | 8,638 | 0.717 | 0.622 |
| 9 Nov 2026 | 9,153 | 0.753 | 0.695 |
| 10 Nov 2026 | 8,211 | 0.747 | 0.682 |
| 11 Nov 2026 | 8,123 | 0.683 | 0.631 |
| 12 Nov 2026 | 8,225 | 0.649 | 0.585 |
| 13 Nov 2026 | 8,731 | 0.720 | 0.666 |
| 14 Nov 2026 | 7,847 | 0.664 | 0.667 |
| 15 Nov 2026 | 7,170 | 0.581 | 0.600 |
| 16 Nov 2026 | 4,904 | 0.618 | 0.563 |
| Class | Tested on | Skill before | Skill with holders |
|---|---|---|---|
| Crash, read at minute 2 | 66,483 | 0.624 | 0.660 |
| Crash, read at minute 5 | 57,928 | 0.657 | 0.700 |
| Crash, read at minute 15 | 46,990 | 0.677 | 0.687 |
| Crash, read at minute 60 | 35,981 | 0.522 | 0.523 |
Studies
Questions the research runs answered on the way, with the data that answered them.
- All launches4.3%
- Bundled first block43.1%
- Top 10 hold over 35%38.7%
- Top 10 hold 22 to 35%20.9%
- Largest holder over 10%19.5%
- Dev holds over 10%13.8%
- 3+ snipers in the first second7.2%
- Largest holder 5 to 10%6.4%
| Flag | Launches with it | Crashed with it | Crashed without | Times the base rate |
|---|---|---|---|---|
| Bundled first block | 3.4% | 43.1% | 2.9% | 15× |
| Top 10 hold over 35% | 2.3% | 38.7% | 3.4% | 11.3× |
| 3+ snipers in the first second | 54.4% | 7.2% | 0.7% | 10.1× |
| Top 10 hold 22 to 35% | 7.2% | 20.9% | 3.0% | 7.1× |
| Largest holder over 10% | 3.0% | 19.5% | 3.8% | 5.1× |
| Dev holds over 10% | 2.9% | 13.8% | 4.0% | 3.5× |
| Largest holder 5 to 10% | 14.8% | 6.4% | 3.9% | 1.7× |
- Read at minute 2
- Read at minute 5
- Read at minute 15
- Read at minute 60
| Read at | Crashed | 10% | 25% | Median | 75% | 90% |
|---|---|---|---|---|---|---|
| minute 2 | 8,928 | 0.8 min | 2.5 min | 8.1 min | 23.1 min | 68.3 min |
| minute 5 | 7,109 | 1.8 min | 4.6 min | 11.7 min | 32.9 min | 103.2 min |
| minute 15 | 3,863 | 3 min | 7.4 min | 20.7 min | 60 min | 175.5 min |
| minute 60 | 1,283 | 8.6 min | 27.6 min | 79.8 min | 251.7 min | 763.3 min |
Up: share that crashed. Dashed: forecast equals outcome.
Show the 6 values
| Forecast band | Launches | Forecast | Crashed |
|---|---|---|---|
| 0% to 1% | 56,029 | 0.4% | 0.1% |
| 1% to 3% | 1,637 | 2.2% | 2.4% |
| 3% to 10% | 1,480 | 6.0% | 7.0% |
| 10% to 30% | 1,431 | 19.3% | 23.4% |
| 30% to 60% | 1,406 | 43.0% | 45.4% |
| 60% to 100% | 1,905 | 76.4% | 83.5% |
What did not work
Tried, measured, dropped. Kept here so nobody has to take our word for what is in the models.
- Coordinated buyer groups
- Wallets that bought the same launches early again and again were grouped. Launches bought by a group whose earlier launches crashed 30% of the time or more crashed in 34% of cases, against 0.22% without such a group. Strong alone, but the models already see it in the first minutes of trading: added to them, the groups raised no score on held-out days. Not in the models.
- Operator trace
- Dev wallets were linked through shared funding to find one operator behind many launches. Launches marked this way did worse on every measure than a plain control: dev sell-outs 38% against 43% for bundled launches without any history, crashes 8.3% against 10.3%. The dev wallet's own record alone was more precise, 78% against 67%. Dropped.
- Narrower time windows
- A model was trained to give each launch its own window for the time to a crash. Its windows hit as often as promised but were no narrower than the fixed spread per reading time. The fixed spread is what the pages show.
Research runs of 18 and 19 Nov 2026 on the CON market recording.