Survivorship Bias Is a Hall of Mirrors

Funds close. Strategies decay. Traders blow up and stop posting. The internet remembers survivors.

This distorts expectations: it makes rare success look normal and normal failure look like personal incompetence.

Without base rates, you misprice risk and time horizons.

Testing reality: backtests hide execution, spreads, and regime clustering. If your edge survives only in clean data, it’s a data story, not a trading edge.

• Test with variable spreads and slippage by session.
• Stress the tails: worst 1% days matter more than average days.
• Prefer rules you can execute consistently over rules that optimize the past.

Overestimate success probability → oversize → normal drawdown feels ‘unfair’ → strategy hopping → blow-up.

How it kills accounts:

1. Model works on clean history.
2. Regime changes and execution friction increases.
3. Performance decays slowly, so you rationalize.
4. You optimize parameters instead of reducing risk.
5. Drawdown becomes the teacher.

Assume base rates are worse than the highlight reel.
Judge strategies over full cycles, not screenshots.
Prefer boring survivable progress over heroic targets.

Rule that survives:

• Stress test tails and execution, not averages.
• Prefer robust plateaus over optimized peaks.
• When performance decays, reduce risk before “fixing” the model.

You see a few traders claiming outsized returns and assume it’s common. You take risks that make sense only if those returns are normal. They aren’t. The mirror lied.

Tell: if small parameter tweaks flip your results, your model is fragile.

Base rates protect you

Once you accept base rates, you stop trying to force outcomes and start engineering survival.

Related: High win-rate strategies die quietly and Sequence risk beats your average.

Glossary: survivorship bias, base rate.

Field checklist

• Stress-test the tails. The worst days define survival.
• Use variable spreads and slippage in testing.
• Prefer robust plateaus over optimized peaks.
• Look at drawdown shape, not only profit.
• If a tiny parameter change breaks the model, the model is fragile.