Question 1

Does the ATR Expansion improve RSI(2) Mean Reversion performance on Crypto?

Accepted Answer

Not decidable — the runs disagree: the 90% interval of the effect straddles the ±1 percentage-point line, so claiming a direction would assert more precision than the data supports. Based on 186 independent backtest runs, the ATR Expansion changes median CAGR by +1.50% vs. a no-filter baseline. Multiple-testing check: the filtered Sharpe clears the threshold SR̂₀ = 0.119.

Question 2

How is multiple testing handled?

Accepted Answer

When 11 variants (no filter plus every filter of this strategy) are tested on the same backtest data, one of them can look good by luck. SR̂₀ = 0.119 is the Sharpe the best of these trials would reach by chance alone, computed from the spread of their Sharpe values with the extreme-value formula of Bailey & López de Prado (2014). The filter passes when its Sharpe lies above SR̂₀ — a threshold test, not a probability. Sharpe values are annualized over 365 calendar days.

Question 3

How is this analysis survivorship-bias-free?

Accepted Answer

Each filter is evaluated in isolation across all real backtest runs saved by users on the platform — with no look-ahead in trade signals. Sample size is shown transparently; verdicts with fewer than 30 runs are marked as insufficient and excluded from search indexing.

Does ATR Expansion help RSI(2) Mean Reversion?

CAGR Comparison

Risk-Adjusted Performance

How does this filter work?

Multiple-testing threshold