Base rate (trading strategies)The share of comparable ideas that held up out of sample after costs. Start from it before trusting your own backtest.Placebo testRunning random signals through the same test on the same markets. A real signal has to do clearly better than this range.Out-of-sample testJudging an idea on data that was not used to build it. In the register: built up to 2014, judged on 2015–2026.Trial countThe number of variants tried before a result was found. The more attempts, the higher the bar a result has to clear.In-sample upper boundA backtest over the full history including the tuning period. The most favourable view; if a rule loses here, it will not do better on new data.Dead endAn idea class that failed repeatedly after costs: fewer than 20% of tests beat the benchmark and the median result was clearly negative.