I have been doing backtesting on my own over the summer. I focused on the gbpusd forex cross. Basically in going for backtesting, some will tell you the classic version about the past not being a guarantee for the future, while others will swear to backtesting. The answer most likely lies somewhere in between.
First of all test something that makes sense. This means don't set up 5 indicators and try to find a golden forecasting rule. Think about what would make sense - for example a correlation between the volatility of the prior day and the profit target on any given day. On the basis of these thoughts(what I gave you was just one of many), you should set up a small set of rules. In backtesting you will use these rules and see what would have happened over different periods, had you been trading with your set of rules.
This brings the question of which periods to test with. You have to test data from a longer period, and not just the past week. Also it might be good to test with two periods separated in time, one very recent and one further back in time - if the results are positive in both cases, you can expect this tendency to stay around.
Remember to follow your rules strictly when backtesting - avoid over optimization, ie. 'i wouldn't have taken that trade as I could have easily seen it not working out'. And expect that whatever results you get, your eventual live trading results will probably be less profitable as commisions, spreads and psychological factors step in.
The guys at MasterTheGap have studied open range breakouts significantly, but the 15min and 60min breakouts, not 30min. Their published setup statistics are based on
1) size and direction of the opening gap
2) has the gap filled or not
3) which quartile on the opening range the price closed in at the 15 or 60min (30min in your case) time frame
4) "market conditions" on the daily time frames, usually is the MA20 > MA50, or is MA10 > MA200, or some combination of MAs above/below longer-term MAs to weigh bullish/bearish market conditions.
Also, I think you also have to consider the size of the opening range relative to ATR On a daily chart. For example, if the opening range > the ATR on the daily chart it's probably not a good idea to trade a breakout because the odds of range expansion beyond the opening range are less probable.
For more info on their methodology you can look at this page and watch the video
A few years ago I did quite a bit of backtesting myself on CL opening range. I could never get much better than 55% win rate and 1.1 profit factor which is not very comfortable for me. Also, since CL is a more 24x5 market, the open is not as relatively volatile as equity markets since most stocks don't trade much pre/post RTH.
The following 6 users say Thank You to shodson for this post: