Trading Concept
Overfitting
Tuning a strategy so precisely to historical data that it captures noise instead of signal, performing well in backtests but poorly in live trading.
Overfitting is the most common cause of backtest-to-live performance gap. A strategy with 15 parameters, each tuned to maximize historical returns, has learned the specific noise of that dataset — not the underlying market dynamics. Change the dataset and the strategy fails.
The signature of overfitting: exceptional backtest performance that degrades rapidly in live trading. Complex strategies with many parameters are more susceptible. Simple strategies with few parameters are more robust but may underperform in backtests.
The botwir3 builder constrains overfitting risk through module design. Each module has a small, fixed set of configurable parameters. The builder does not allow arbitrary indicator stacking or parameter multiplication. This is a design choice, not a limitation — fewer knobs to turn means fewer opportunities to overfit.