The traditional way to backtest or develop strategies is to take a very long history and run your rules over that history. The objective is to find rules that produce stable or consistent returns. However, I have recently been experimenting with another technique for developing intraday strategies where I load from only a single to a few days worth of history. On this short history, I seek to find the optimal trading rules or discover patterns. My original objective was simply to find rules, filters, or patterns that might be helpful to understand as a discretionary trader.
However, surprisingly, some of the rules I developed continued to perform when I extended my backtest history as far as 5 years back, the maximum that I checked. I am not sure how useful this approach will prove to be but it looks like it may be useful as a starting point given that it is especially difficult to develop systematic intraday trading strategies.