Probability of backtest overfitting
Forge API 1
A low figure is good: it says the winner of the search is not just the luckiest of many.
over the ways to choose half of {pbo_n_blocks} blocks of the history as the fit, the share in which the best setting on the fit ranked below the median on the other half; the run passes when it is at most {pbo_max}
Was this page helpful?