10points3 review - flawed - shall I restart?

[PipCop]

So I was looking at the 10points3 robot (http://www.PipCop.com/forums/previous-robot-reviews/63-10-points-3-robot-review-busted-4-547-a.html) and noticed something interesting. The SL is 25 pips, but when you look at all the orders the SL is more like 100+ pips. Why is this?

I looked at the code and found that while the SL is set to 25, it appears that it's not put onto the order properly - it's put on as price + SL . That means it's WAY off, because the typical price of 1.2589 + 25 comes out to something crazy. I suspect a code review would be in order, and at that point the robot may perform well.

Does anyone want to tackle that? Here's the code.

[PipCop]

It also appears there are inconsistencies with the TS as well... although I am not a good enough programmer to confirm, the calculations are clearly different when they should be similar.

[1ka]

Er, no.

If the EA is broken, then that's what PipCop is here to find, isn't it?

It's up to the author to fix the EA.

[PipCop]

Technically I want to find the best robots. However, if one that has great potential has some simple bugs, why not fix it? The initial performance was stunning, it just had way out there values for the SL and now I know why. Why not patch it up if it's simple? What if the fixed version is a real money maker?

I'll ping the author ... see what we can get...

[1ka]

Good point - A good EA would be great!

Having looked at the code (I just can't resist) then the SL is multiplied by Point before it's added to Price. Point is a metatrader function that returns the current point size (0.0001 or 0.00001 I would think). So if that's the case, then this code looks fine.

The SL also gets set dynamically as the EA runs and is based upon the property TrailingStop. Perhaps this is wrong?

This (queue another code moan) author is another keen amateur. Everything runs in the start() procedure!

I think Frodo had a phrase for this "One routine to ruin us all.."

[PipCop]

Thanks for explaining Point - I didn't realize that. I can't "see" the MQL4 values, so they escape me sometimes.

Yeah, that's the thing I noticed - somehow the SL was not working right. Look at the first few trades in the review thread - the SL is like 108 points below the opening price. Why? Had it been 25 points then those massive losses would have been trivial.

So who do we have to kidnap around here to get someone to fix this code?

[1ka]

PipCop, I don't think the code is a problem, just that it has an initial SL set at 110. Obviously with a lot size of 2+ this is a significant amount of cash!

I think it needs to be run with a lower SL , the value to choose would need to be based on some strategy testing. However, this is another very simple EA based around a limited check of MACD, my guess is that if you start to limit SL you'll hit more SL calls than TP calls. Might be worth a try with a value of, say, 30 pips for SL ?

[PipCop]

I didn't see that 100 SL , which makes sense that you mention it. I agree that a SL of around 30-40 pips would be better. I'll see if I can take a look at it this weekend and start it up again. It makes no sense (to me) to have a SL that is over 10x as large as the TP . That is not good money management IMO.