<![CDATA[Forex Software — Improving strategies with infrequent indicators]]> https://forexsb.com/forum/topic/3450/improving-strategies-with-infrequent-indicators/ Thu, 05 Jul 2012 22:36:01 +0000 PunBB <![CDATA[Re: Improving strategies with infrequent indicators]]> https://forexsb.com/forum/post/15341/#p15341 Run the analyzer and see what effect small changes in the indicator values make.

]]> Thu, 05 Jul 2012 22:36:01 +0000 https://forexsb.com/forum/post/15341/#p15341 <![CDATA[Re: Improving strategies with infrequent indicators]]> https://forexsb.com/forum/post/15339/#p15339 I tried replacing each main indicator with the infrequent ones, didn't work out very well though.
I think the best thing to do will be to work out the probability of the trades being "blocked" being luck, with a 0.47 win rate (0.53 loss/break even) then the probability of 5 trades being blocked by a new indicator all being bad trades should be around 4.2%. I don't think I would settle for much higher than 5%.

]]> Thu, 05 Jul 2012 18:56:39 +0000 https://forexsb.com/forum/post/15339/#p15339 <![CDATA[Re: Improving strategies with infrequent indicators]]> https://forexsb.com/forum/post/15338/#p15338 What if you replace one of the standard indis with the infrequent as you call it? Does it make strat worse or better? I think the latter is the main question. Cluttering the strat with many indis might not be the smart play here, but that's only my opinion.

]]> Thu, 05 Jul 2012 18:35:16 +0000 https://forexsb.com/forum/post/15338/#p15338 <![CDATA[Improving strategies with infrequent indicators]]> https://forexsb.com/forum/post/15337/#p15337 Hey,
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.
Regards,
Matthew

]]> Thu, 05 Jul 2012 18:23:43 +0000 https://forexsb.com/forum/post/15337/#p15337