Topic: Improving strategies with infrequent indicators
I have been adding some new logical conditions to a strategy to try to improve it. Some of these indicators change maybe 1-4 trades out of ~850. They improve the total profit, sometimes only cutting out 4 losing trades while not changing the total number of winning trades. Sometimes they change 1-2 of each but increase total equity. Are these alterations worth working on? After all, if they only change a few trades per year (I had around 850 trades over a 12month back-test) then there is a distinct possibility of them just being lucky in the trades they cancel out.
Would any of you bother including a new indicator that only changed, say 4 trades out of 1000, but increased total profit in back-testing?
I am talking about conditions such as Stochastics slow %D being above a level of 2.