Research record · Historically supported
Trap Scanner: Breakout Trap Backtest Results
Trap, called Trapdoor until October 2026, tracks a common “trap and reverse” sequence: price breaks out of a range, fails, snaps back inside and runs to the other side. Research over about two years and 95 coins found that much of the move happens before the final breakout. The sealed holdout was encouraging but inconclusive.
- About one in three first breakouts reached the opposite value boundary; 79.5% of those went on to the opposite range edge.
- Measured from the first breach, the move averaged +1.004 ATR after 3 bars; waiting for final confirmation left +0.169 ATR.
- In the sealed holdout, average results were positive but the uncertainty intervals crossed zero.
- Fees, spread and slippage were not included.
What Trap watches
Trap works on 15-minute ranges with 1-minute detail. It follows each range through stages, from the first close outside the range to a possible full reversal to the other side. It is built to show you where a chart is in that sequence, so you can inspect it yourself. It does not place trades or set stops.
ATR (average true range) is a coin’s typical movement per candle. Results in ATR units can be compared across coins with very different prices. Each range has an outer edge (its high and low) and an inner value area where most trading happened; the developing stage is when price breaks the value area on the opposite side.
How often a first breakout turns into a reversal
This discovery analysis followed the first closed-candle breach of a range edge. Counts are episodes, not unique coins.
| Step | Episodes | Share of first breaches |
|---|---|---|
| First close beyond the range high or low | 15,753 | 100% |
| Reached the opposite value boundary | 4,917 | 31.2% |
| Continued to the opposite range edge | 3,907 | 24.8% |
Of the episodes that reached the opposite value boundary, 79.5% (3,907 of 4,917) carried on to the opposite range edge.
Where the move actually happens
The same confirmed moves were measured from three different starting points, three bars later.
| Measured from | Average 3-bar outcome | What it means |
|---|---|---|
| The first breach | +1.004 ATR | Much of the move is underway before the final breakout |
| The developing stage | +0.271 ATR | Catching it at the opposite value breach still leaves continuation |
| Final confirmation | +0.169 ATR | Waiting for full confirmation leaves less movement |
This is the main practical finding: confirmation costs movement. That is why Trap shows the developing stage instead of only alerting at the end.
The sealed holdout: encouraging but inconclusive
The holdout covered 95 coins from 6 December 2025 to 30 May 2026, after the rules were fixed. Results varied with a volume-flow filter setting.
| Setting | Events | Positive rate | Mean / median return | Mean best run / worst drawdown |
|---|---|---|---|---|
| 5% (default) | 238 | 50.4% | +0.265 / +0.007 ATR | +1.257 / −0.935 ATR |
| 10% | 170 | 46.5% | +0.278 / −0.092 ATR | +1.275 / −0.965 ATR |
| 15% | 109 | 46.8% | +0.573 / −0.023 ATR | +1.504 / −0.912 ATR |
| 20% | 69 | 47.8% | +0.960 / 0.000 ATR | +1.867 / −0.826 ATR |
| 25% | 39 | 41.0% | +0.682 / −0.207 ATR | +1.591 / −0.995 ATR |
Averages were positive, but medians sat near zero and the uncertainty intervals crossed zero. The 20% setting had the highest average on only 69 events, which is an interesting data point, not a proven best setting. Opposing flow at the first breach did not reliably predict reversal strength.
Limits of this research
- About two years of data, 95 coins, late May 2024 to late May 2026.
- No exchange fees, spread or slippage.
- Trap identifies the stage of the pattern. It does not predict direction, give win odds or manage trades.
Frequently asked questions
Is Trap a fakeout indicator?
It tracks a specific fakeout pattern: a breakout that fails, returns into the range and runs to the other side. It shows the stage a chart has reached. It does not label any single breakout as fake in advance, and it does not tell you what to trade.
Why does waiting for confirmation leave less movement?
Because much of the reversal has already happened by the time it is confirmed. Measured from the first breach, the move averaged +1.004 ATR after three bars; measured from final confirmation it averaged +0.169 ATR. Earlier entries carry more risk of being wrong.
Is Trap the same scanner as Trapdoor?
Yes. The scanner was renamed from Trapdoor to Trap in October 2026. The definition and this research record did not change.
What does “inconclusive” mean here?
The average results in the holdout were positive, but the statistical uncertainty was wide enough to include zero. The evidence supports the pattern as a structural guide, not as a proven trading edge with a known win rate.
Educational content, not financial advice. Historical results do not guarantee future outcomes.