Phoenix optimization

 

Now it has it's own part of forum I will take up one of the most popular question and hopefully someone could answer. How do I optimize phoenix?

This is the one of the faq how to optimize the phoenix by VinceTheBeast.

https://c.mql5.com/forextsd/forum/162/phoenix-2007-how-to-optimize-phoenix.pdf

If you have any more tips or links, please post in this thread.

 

Thank you Prankie. We are hoping to have NewDigital move and Sticky the 2 threads that we are planning to use as informational Posts. So hopefully in a day or two we will be getting things up to speed in here.

Just as an aside. Phoenix will always be Hendricks baby. However from his last post in his main thread he admits he has suffered extreme overload with the popularity of his EA. A couple of us have been talking about how best to carry on with the development and education for Phoenix_EA. This is not a takeover, we are simply trying to find ways to keep this awsome tool alive.

If you want to help Contact either myself or Daraknor.

 

I've been posting in the threads instead of writing the merged code I promised. I don't want to moderate the threads really, but it would be useful to have the information all in one place.

 
 

Vincethebeast,

I agree with you about standard sizing of accounts for backtesting. That would be very useful.

Right now a major concern of mine (I consider it a big bug) is the different results from using the same broker, same time. Once we tackle that problem, we might be able to properly address the question of Alpari's data stream versus another. Also, I don't know if the MT Build200 datastream is broker specific or Alpari only. I need to download a few and see if they are identical.

It may turn out that once we fix the 'varied results' issue that we automatically account for different data streams. I have been staring at the indicators and thinking. The individual broker streams shouldn't have vastly different moving averages. The likely main reason for differences is the timing of entry points.

 
daraknor:
I don't know if the MT Build200 datastream is broker specific or Alpari only.

I had a look at the ip Build 200 tries to reach whilst downloading and it seems to be pointing to this address: 64.151.112.148 which is the domain of Metaquotes. Maybe there are other brokers which do point to their own server... If so, that data would be more reliable I suppose.

Vince

 
daraknor:

Also, I don't know if the MT Build200 datastream is broker specific or Alpari only.

It uses Metaquote data. They also disabled option to use external data (no more Alpari).

Their historic data differ from Alpari's grrr and beta testing results are different. (add another grrrr) lol

More about this subject you can find here: https://www.mql5.com/en/forum/175528

---------------------

Vince, I was thinking the same way. Since most of backtesters will use Alpari historic data I've already decided, when I will open live account, I will open with Alpari for start.

Now, there is the way, to use 99% modeling quality for backtesting. Msg nr 4:

https://www.mql5.com/en/forum/175561

I've lost the link where I downloaded it from (still searching lol), but I have data for one pair for 99% modeling quality. I couldn't use it, but it was probably because of MT b200. Will try to reload it today with MT b198. I hope it works. Would really be great, if we can all set up our computers to 99% modeling quality and at least eliminate one worry/issue - data for backtesting. There are already enough issues to worry about IMHO lol

Mario

 

Bank data not broker data, interesting. There seems to be all sorts of strangeness regarding Build 200 backtesting.

I think the biggest question facing Phoenix is if reoptimization leads to better results now.

 

I've found the link, it was on the sister's thread.

I've downloaded again and try, but loks like it is no good hist file.

http://rapidshare.com/files/1333387/EURUSD1_0.zip

I got only 25% modeling quality

Mario

 
daraknor:

I think the biggest question facing Phoenix is if reoptimization leads to better results now.

We can still use Alpari's historic data (90% modeling Quality) and MT4 B198, right?

Mario

 
alamanjani:
We can still use Alpari's historic data (90% modeling Quality) and MT4 B198, right? Mario

I think you could use Build 200 too if you wanted to.

You just need to make sure you don't have 2 sources of ticks 'polluting' the history files.

I have a MT4 which isn't connected to the internet (great idea Hendrick!) so no new data gets in.

I've deleted all tick data (all the *.hst and *.fxt files) first of this MT4 and then imported the history files the old fashion way (so NOT with the download button).

Reason: