You can never be certain that a trading strategy, which you have patiently developed using historic market data, will continue to work as the market changes. Our approach is to formulate large numbers of strategies and keep track of the successes and failures of each one. Most of them are never trusted; only if a strategy manages to acquire a strong record of making good decisions, and avoids almost all bad decisions despite having had a number of opportunities to do so, will it be listened to. And it will only be trusted as long as its performance continues to be good.

We call these evidence-based strategies. As the market changes, old strategies may fail and be discarded, but the system continually explores the range of possible strategies looking for new ones that will suit the current conditions.