Phoenix 2007 (new thread) - page 104

 

Good article ipixtlan, I have enjoyed reading a few articles you post in your blog before but that was very very good. The changing of market specifics means you need to retrain your indicators, the changing of market types means you need to learn to ignore them or listen to them, and the fact that everything changes in patterns means you need to correlate relevant information to your decisions. Those are GA, NN and Bayes respectively Sean (physics friend) was originally convinced that wavelet analysis would be very important and accurate - because it works so well in physics. After a few hours of experimentation he threw it away because he realized this was simply a different reality (my words not his).

I originally thought I could do some systems math and determine specific facets of finances using those models. That is partially correct but it depends on on accurate data being used to make decisions. The reality is that our Signal is drowned out by quasi random Noise, but we react to the quasi random Noise as signal. Not collectively, but in discrete disconnected pools (brokers).

That understanding is part of why I think Phoenix has a chance at long term viability, but only if we can adapt it regularly to change. Laying awake at night I have been trying to figure out what conditions I could use to tell the user "Don't trust me anymore, I think I'm broken. Please get new settings." The article you posted is similar - the quantative finance people don't know when to stop trusting their models.

 
 
 

Daraknor

daraknor:
Agreed on all counts psnw!! Exactly right. I hope that we can make a system that functions reliably. That is my idea of relatively bug free. We will still need new settings files regularly, but reliability of trading is very important. Same broker, same currency, same time should equal same/similar results. I hope that isn't breaking the mold of Phoenix :/ I do worry about changing too much. Multiple threads may fragment the community - and as much as I do want a website (3 offers now) I hope we can mend and heal as a community here. If we make a website now the fragmenting and confusion may only be greater. NewDigital and the other admins were kind enough to give us a separate section, and I would like to stay here on a forum if we can. Once there are dozens of "stable settings files" then we may need a database to organize them all. Do you think the community, hurting and fragile as it is, could be transplanted to a website? (Honestly asking opinion here.)

I agree with maintaining new settings for the different currencies while working the "bugs" out. To alter the code too much would take away from the heart and soul of what Hendrick created. I mean Phoenix works!

As for starting a website, I agree the community seems very fragile at this point. I am not sure if it would survive another major change or not. perhaps a poll may be in order to better answer that question. What I am in the progress of doing right now is creating a web page for myself. It is sort of a easy referance of links which lead back to the the threads here in the forum. I am not a Webmaster and in no way could build a high speed site needed to maintain Phoenix. But, If anyone is interested I would be willing to share and am still looking for ways I can be productive here.

From the reading I have done, Yes....there are a those here who are taking up the torch and carrying on the spirit. My post was in no way meant to slander the community, it was meant to hopefully pull us together. United we stand, Divided we fall.

Thank you Newdigital for granting us our new section!!

 

What an amazing weekend for the future development of Phoenix, finally a structure is in place that will hopefully make running the community easier, i just hope that daraknor and friends are able to concentrate more on development than running these threads now that they are set up.

Now the community is taking the development structure seriously, the next sticky question must be development funding, ie whilst backtesting and forward testing on demo accounts is useful for pointing the right direction in creating better settings until they are proven in live testing with a variety of brokers they are simply unproven.

I do not know how many people in the community actually have live accounts but considering the number of possibilities in confirming good/bad settings with a variety of brokers i am sure there are not enough, therefor would it not be a good idea to set up a Phoenix development fund for the creation of live accounts just for testing purposes, any new business needs development/start up funds i see no reason why Phoenix should be any differant, whilst i can offer no help when it comes to coding etc i and i would like to think others would be happy to contribute funds to help set up live accounts for development purposes, perhaps a monthly membership fee like the membership fee for the private section here, obviously if these accounts overall made money then perhaps some kind of dividend payment to the members at the end of say 6 months or something could be paid out, with a percentage sent to Hendrick, just a thought.

I would also like to thank those that have made this community possible it is truly inspiring.

 

Thanks. This link works.

Francoc

jarcrocker:
They have just recently modified the websites and moved files. History Downloads are here. Keep following the manual directions. http://www.alpari-idc.ru/en/dc/databank.php Thanks Hendrick for putting that together.
 
francoc30:
Thanks. This link works. Francoc

Here's an alternate link to the UK Alpari History Centre:

http://www.alpari.co.uk/en/dc/databank.html

and the US History Centre:

http://www.alpari-idc.com/en/dc/databank.html

 

change

I don't see any need for relocation to a website personally. This seems for all intent and purpose to be just that the way it is now arranged. Excellent work by all who took responsibility. Kudos!

Greg

 

Is ph0enix still Using TimeFrame 15MIN ?

 
ForexMan:

Is ph0enix still Using TimeFrame 15MIN ?

ForexMan, see msg 4 FAQ:

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

We need to reduce these keep repeating questions in other threads.

Although it is easy to just answer "yes" it is hard to keep repeating the same thing

Mario

Reason: