This is a bit complicated but I will try to explain what I am considering doing as best as I can.
I have written an indicator that returns six possible results for each candle. I have then written an indicator that basically does the same calculations as the first one and then analyses them. It looks at each candle and then compiles a set of results as a probability survey. So for example if indicator one returns a 1 for this candle what is the historical probability that the next candle will be an up one or a down one. It also gives historical performance for each possible option and returns a total of twenty four results for the various possibilities. Then I’ve written a strategy that uses the probability results from the second indicator (by CALL) to make decisions on when and and in what direction to trade for the highest probability of a positive return. My only one issue is that the probability indicator uses one variable and that is a time based one and I hate time based variables as they are so easily curve fitted. So I was wondering if there was any reason why I shouldn’t have say five or more identical probability indicators (but with different names obviously) with the variable set differently in each and then CALL them all from the strategy and then allow the strategy to analyse the results and pick a mid set of values to prevent curve fitting. A sort of continuous optimization in a running strategy.
I’ve not used the CALL function much in the past due to the fact that it has always been so slow but it seems quite quick with the probability indicator that I have created. Any thoughts on why I shouldn’t continue down this road would be much appreciated before I waste a day of my short time on this spinning globe trying to complete it.
.
