'Weighting' Indicators influence to either Open or Close trades?

 

I'd like to give 'weighting' and control the amount of influence that different indicators and chart info has on making the decision to either enter into a trade or not and to close and exit a trade or not. Hence one indicator might be weighted at 100%, while another indicator might only be weighted so that it only has 50% of its info 'influencing' the decision. We tend to do these ourselves (subconsciously?) in prioritizing different indicator output and chart info when Manual trading. But how to do it in an EA? If all EAs only had outputs of say 0 to 100, then this would be relatively easy. However as we know, most indicators don't have this sort of output. There is a vast array of different types of outputs and information from indicators.

So how to achieve and automate this?

 
FourX:

We tend to do these ourselves (subconsciously?) in prioritizing different indicator output and chart info when Manual trading.

I don't use Technical Indicators . . . how do you do this ? figure out how you do it, write it down, then you can code it.
 

Hi Raptor,

Interesting. What do you use as criteria? Likely 'weighting' the factors that you do use. I think most do.

figure out how you do it, write it down, then you can code it

Haven't 'brought it down to Earth' yet and figured out how to do it. Especially as indicators don't tend to utilize the same, functions, variables, arguments and outputs etc. All manor of things. I think that I have to come up with a 'standardized' weighting factor. Such as 1 to 10 and then convert the results of the indicators to this scale. Sounds like a nightmare to me! )< 8). But I can see that in doing so, a lot of the usefulness of the various different types of data & info returned by these indicators would be lost. Course I certainly could, and would most likely have to distinguish between types of data. Such as specific TP, SL, TSL etc while this same info and others might only be factored into if and when to open up a new position and when to close them out etc. But even the various levels of TP, SL etc get weighted into if to enter into a new trade or not.

Once I have these entry and exit info 'scaled', I would then had to give a weighting to my initial scaling and set a threshold to either act or not from the cumulative total or some other mathematical summing of these factors.

Gotta admit, this sounds pretty inadequate, cumbersome, unwieldy and ineffective. Have to think about different approaches and also get more concrete about how I currently do them now.

As we hear over and over again in these realms, until one can lay things out clearly in concrete terms, one can't code it. Otherwise just GIGO and a heck of a lot harder to come up with effective solutions. But I can see this being a series of iterations: get the major building blocks and functions, then work my way through accomplishing them: 'Uroboros'

Again: If you don't utilize indicators Raptor, what criteria do you use and how do you prioritize it?

 
FourX:

Hi Raptor,

Interesting. What do you use as criteria?

I use Fibs.
 
RaptorUK:
I use Fibs.

I don't want to get into an argument over semantics, but those are indicators Raptor. So call and classify them however you like.

I use Fibs as well, but don't find using just them to be adequate. Likely I'm not using them to the extent that you are Raptor. I have one EA that using them but reverses them. Said he found this to be more accurate.

 
Sure you can call Fibs whatever you like . . . . I said I don't use Technical Indicators . . . and I don't consider a Fib to be a Technical Indicator, neither do Metaquotes as far as I know.
 
FourX:
Once I have these entry and exit info 'scaled', I would then had to give a weighting to my initial scaling and set a threshold to either act or not from the cumulative total or some other mathematical summing of these factors.

Gotta admit, this sounds pretty inadequate, cumbersome, unwieldy and ineffective. Have to think about different approaches and also get more concrete about how I currently do them now.
Sounds like Fuzzy Math, Part 1, The Theory « Synaptic
 

The starting point has to be the question, when manual trading, are your decisions regarding which indicator values are the most influencial at the time, based on any definable logic or not.

If it is, you might say the reason I sway more to fibs than pivot points at some times is when fibs have been successful recently .. so then you could code to test fibs over the past nPeriods and pivot points over the same and whichever is more successful might be part of your criteria to decide which to use. Im not suggesting that would actually work, its just an example of turning your manual trading decisions to an EA.

I would imagine, in reality the scenario I described wouldnt work because just when the tests show fibs have had a run of success would be just when they quit working lol