What criteria do you use to qualify a Breaker Block over an Inversion Fair Value Gap?

Both concepts capitalize on failed structural levels. Which setup model has shown higher historical win rates and cleaner price delivery in your backtesting data?