Whenever going for testing or for a testing purpose of a trading idea on historical data, it would be suggested to also reserve the period of the time of the historical data. This historical data using for testing purpose and for optimisation is usually known as in-same data and,
The data set that has been reserved is referred to as out-of-sample data. This setup is very crucial for the evaluation process because it enables traders to test the trading idea on the data that has not been the part of optimisation model.
Resulting in the zero biasedness of the idea by the out of sample data and traders will be able to evaluate how the system performs on new data i.e., real-life trading.
Note: Prior to the initialization of any backtesting or optimisation, traders have the option of setting aside a percentage of historical data to be reserved for out of sample data. It is highly suggested to reserve one-third of the historical data to be used in out of sample testing. In the sample, data should only be used for initial testing and optimization only.