Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when a Polymarket outcome appears to overreact and then stall away from recent filled-price range.
.claude/skills/superior-trade-probability-mean-reversion/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -43% | 0% |
Use this when someone asks for fade, overreaction, mean reversion, range trading, panic buy, euphoria selloff, or probability jumps with weak follow-through.
POST /v3/markets/search.Moderate fit. Filled TradeTick history can test whether probability moves have a measurable reversion pattern, while using actual fills as the execution proxy.
Limit: backtests cannot validate resting liquidity, spread paid, or what is missed in the maker queue.
Enter when the outcome trades below a lower band and exit near the rolling median. If your implementation permits, do the inverse for above-band conditions when downside overextension appears.
json{ "window_ticks": 40, "entry_deviation": 0.08, "exit_deviation": 0.02, "order_size": 10, "max_holding_ticks": 80 }
| Knob | Effect | |---|---| | window_ticks | Larger windows produce a smoother baseline. | | entry_deviation | Higher values wait for stronger overreactions. | | exit_deviation | Lower values demand tighter reversion before exit. | | max_holding_ticks | Prevents stale positions through stale conditions. |
"This is an overreaction fade. It works best in noisy markets without new decisive information. I’ll backtest it on filled prices first, then verify trade frequency and liquidity before suggesting live use."
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 34,823 | 14,851 | -57% | 1 | 1 | 0% | 3,049 | 2,538 | -17% | 0 | 0 | — |
case-02 | fail→pass | 33,924 | 24,635 | -27% | 1 | 1 | 0% | 4,727 | 4,195 | -11% | 0 | 0 | — |
case-03 | fail→fail | 15,994 | 7,091 | -56% | 1 | 1 | 0% | 2,741 | 1,948 | -29% | 0 | 0 | — |
case-04 | pass→pass | 22,497 | 15,872 | -29% | 1 | 1 | 0% | 2,419 | 2,678 | +11% | 0 | 0 | — |
case-05 | pass→pass | 16,164 | 11,166 | -31% | 1 | 1 | 0% | 2,446 | 1,947 | -20% | 0 | 0 | — |
case-06 | fail→fail | 19,264 | 15,433 | -20% | 1 | 1 | 0% | 2,845 | 2,435 | -14% | 0 | 0 | — |
case-07 | fail→pass | 11,079 | 3,894 | -65% | 1 | 1 | 0% | 1,935 | 1,072 | -45% | 0 | 0 | — |
case-08 | fail→pass | 29,856 | 11,912 | -60% | 1 | 1 | 0% | 2,565 | 2,371 | -8% | 0 | 0 | — |
case-09 | fail→pass | 21,706 | 7,354 | -66% | 1 | 1 | 0% | 2,736 | 1,563 | -43% | 0 | 0 | — |
case-10 | fail→pass | 20,446 | 50,562 | +147% | 1 | 1 | 0% | 2,458 | 2,723 | +11% | 0 | 0 | — |
case-11 | pass→pass | 15,933 | 6,080 | -62% | 1 | 1 | 0% | 2,123 | 1,422 | -33% | 0 | 0 | — |
case-12 | fail→pass | 18,190 | 13,848 | -24% | 1 | 1 | 0% | 2,495 | 2,526 | +1% | 0 | 0 | — |
case-13 | fail→pass | 14,058 | 10,115 | -28% | 1 | 1 | 0% | 2,176 | 1,785 | -18% | 0 | 0 | — |
case-14 | fail→fail | 14,276 | 7,977 | -44% | 1 | 1 | 0% | 2,181 | 1,655 | -24% | 0 | 0 | — |
case-15 | fail→pass | 15,025 | 2,642 | -82% | 1 | 1 | 0% | 2,086 | 925 | -56% | 0 | 0 | — |
case-16 | pass→pass | 17,070 | 11,909 | -30% | 1 | 1 | 0% | 2,337 | 2,461 | +5% | 0 | 0 | — |
case-17 | fail→pass | 15,706 | 11,437 | -27% | 1 | 1 | 0% | 2,291 | 2,010 | -12% | 0 | 0 | — |
case-18 | pass→pass | 24,105 | 13,127 | -46% | 1 | 1 | 0% | 3,124 | 2,140 | -31% | 0 | 0 | — |
case-19 | pass→pass | 16,336 | 10,387 | -36% | 1 | 1 | 0% | 2,423 | 2,079 | -14% | 0 | 0 | — |
case-20 | fail→pass | 14,706 | 2,696 | -82% | 1 | 1 | 0% | 2,672 | 898 | -66% | 0 | 0 | — |
case-21 | fail→pass | 23,305 | 14,209 | -39% | 1 | 1 | 0% | 2,997 | 2,290 | -24% | 0 | 0 | — |
case-22 | pass→pass | 29,357 | 21,713 | -26% | 1 | 1 | 0% | 3,724 | 3,343 | -10% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +55 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.