As described on https://en.wikipedia.org/wiki/Backtesting, Backtesting is a term used in modeling to refer to testing a predictive model on historical data. 

Since we have a model created by Machine Learning, backtesting is a good way to see how the model performs on past data. In addition, we have filters on the backtest page where you can test the odds, probability and expected value ranges you prefer and save them too. Afterwards, you'll only get tips matching the newly saved criteria.

Any tips after the date of the first publication of a model will be shown on the backtest too, however for dates before we can show simulated tips (which is basically how backtests usually work).

In short, it can show the performance of the given model. As mentioned, we created a model and with backtest we are able to find settings where our model performs the best. "Best" is subjective though: for example, some prefer high yield, others prefer prefer high success rate.

Please see the article Backtest in practice for practical information on how to do backtesting and changing filters at TopTennisTips.

Continue with Backtesting in practice.