Previously, I asked a question about number of years of backtest required for results to be trusted. On second thoughts, I think that was the wrong question. I think the right question would be what is the minimum number of trades in a backtest, not number of years, for backtest results to be...