Survivorship Bias: Why Backtests Only See the Winners
A backtest on the current coin list ignores every dead coin. What Brown, Goetzmann, Ibbotson and Ross showed, with an example and a checklist.
Backtesting & overfitting
How to test a trading idea without fooling yourself: out-of-sample checks, multiple testing, forward tests.
Why Most Backtests Overstate Results
Three landmark papers explain why a great backtest is usually a statistical accident, and what to do: out-of-sample tests and a higher bar.