Fratelli EA by Dragos and magft

[MickyMouse]

WOW

I am glad you figured that out.

Thank you

[MickyMouse]

Dragos and magft

when you are around guys, please say so... need to have a chatter

[fratelli]

Hello fellow traders

I am happy to see so much brain here, trying to backtest and find better settings for my little baby.

Ken I am happy you want help with backtesting also.

Soon I hope to learn do backtest with EA also.

Its a Micky project. LOL.

Keep up the good spirit here.

Peter

[magft]

Right guys i'm back, kids in bed and so i'm settled down for the evening to get my head around all that has been happening and to do some "work" before Monday

The pm in outbox is a feature i saw mentioned elsewhere. It stays in outbox until recipient reads it so you know when they do, i think.

Is it worth chatting on skype?? pm me details.

Mike

[Panamamike]

I think back testing serves a purpose, but I feel like too much emphasis is being placed there. IMO, we need to have the EA be able to duplicate the signals from the Fratelli Main indicator that Mickey has put together. Once we have accomplished that to a 95% accuracy over 50+ trades, then those who wish to backtest can do so.

While that is being done, I would like to forward test the EA on a small $1000 account risking 2% per trade. I will rack the results and check things like optimal trading times etc. to improve performance. I will also do a demo simultaneously and stop trading at a specified daily goal. I will compare the results on an equity basis after a two month period. I think there are limitations in Empty4 gacktesting and that forward testing, including small live tests are necessary to prove the long term viability of the system. I will gladly share the results as they happen.

[SteveHopwood]

I think back testing serves a purpose, but I feel like too much emphasis is being placed there.

I am not up with what is happening here, so ignore my comments if they are irrelevant.

In my experience, backtesting is useful only for testing code. ST works with the current spread, so given that most bastard crims vary their spreads on a tick by tick basis, then the results of backtests will also vary from one test to another. You can argue that this does not matter much, but try backtesting at the weekends......

The spread is the st manipulation we know about. 12 months ago, I did not know about this. What else do we not know about that the bc's have up their collectively smelly sleeves? Edit on reading the posted thingy: imagine actually going up the smelly sleeves of the bastard criminals? Yurrrrgggghhhhhh

Empty4 is crap. ST is total crap. Results gained from ST are irrelevant utter total crap. Unless you are flogging a useless ea via Ebay, in which case curve fitting (otherwise known as 'optimisation') is invaluable.

[johnjwong]

I think back testing serves a purpose, but I feel like too much emphasis is being placed there.

I am not up with what is happening here, so ignore my comments if they are irrelevant.

In my experience, backtesting is useful only for testing code. ST works with the current spread, so given that most bastard crims vary their spreads on a tick by tick basis, then the results of backtests will also vary from one test to another. You can argue that this does not matter much, but try backtesting at the weekends......

The spread is the st manipulation we know about. 12 months ago, I did not know about this. What else do we not know about that the bc's have up their collectively smelly sleeves? Edit on reading the posted thingy: imagine actually going up the smelly sleeves of the bastard criminals? Yurrrrgggghhhhhh

Empty4 is crap. ST is total crap. Results gained from ST are irrelevant utter total crap. Unless you are flogging a useless ea via Ebay, in which case curve fitting (otherwise known as 'optimisation') is invaluable.

Well said, backtesting does not consider the change in spread. In the weekend it got lock up to sometimes 20 pips for some brokers. However, good thing we have the mt4i spread changer that we can change the spread for ST purposes to an amount we see reasonable, then perform the backtest.

[MickyMouse]

We are backtesting to make sure the code works ok ...
we added a lot of features and we are not sure anymore if everithing works as it should ...
Also , the system started with a set of rules , tat were chanced on the basis that somebody, experienced, back tested for us 17months of data so, with everybody interest at herth, peter changed the rules . While everybody was happy and exausted that they had new more profitable rules, the only thing running in my head was. how... how did that person backtest so fast what we needed days to backtest 3 months of data ... in the end, turned out that person backtested on 15 min tf, not 10 as the system is on ... so an ea needed to be done . and also a backtest needed to be done . turns out , faulty , buggy , weekend and so on ... back tests clearly show that the first set of rules were the appropriated ones . we are not optimizing, nore changing anything, probably we will just return to our first set of rules with more confidence in our hearts .

I know back tests ar crap .... especially on Empty4 , but we play with what we have .... should the crims offer us something better (KEEP DREAMING) we will test on something better .

[kenfen]

A couple points from my end...

First...I have Optimized more of my code into The Grail than I ever care to remember...Jackass!

The aim here is to match the EA with the Indy...hopefully to around 95%, or so. There is nothing to optimize in the system. It either is...or it ain't.

Second...spread and weekend crim shenanigans. I'm sure plenty of people here are familiar with Birt and all of the fantastic tools/patches/scripts he has, and continues to tirelessly provide...even when Metaquotes constantly breaks his tools with new builds...its almost like they don't want people using those tools! Hmmmmmmm?

In short...and this will be the only thing short about the entire process. You download RAW tick data from Dukascopy...many GBs in size. For free, you will NOT find better data anywhere...very good quality! Through the use of Birts scripts, you then convert the tick data into Empty4 formats (HST and FXT files). As you are doing this you can select "Use Real Spread" As in...the difference between Bid and Ask. Dukascopy copy has, in my opinion (and I trade live with them), middle of the road spreads these days...not the best and not the worst...pretty average. You can also "pad" the spread if you so choose. Further, you can add in a Commission value that will come off each trade. You can adjust for DST, by European or American dates. As far as Empty4 land...what a land ( )...goes...this is the cream of the crop.

If you are staying away from Optimizing and just looking to mechanically check a system...this is it! We will obviously then have to check/match them manually...until we're satisfied things are firing at the appropriate place. IF we can get there, it will save hours and hours of manually backtesting. It will make it easy to test trading hours...other pairs etc.etc. etc.

[kenfen]

Well said, backtesting does not consider the change in spread. In the weekend it got lock up to sometimes 20 pips for some brokers. However, good thing we have the mt4i spread changer that we can change the spread for ST purposes to an amount we see reasonable, then perform the backtest.

Not good enough! That will NEVER give accurate results.

You MUST use real spread that changes with every tick...the difference between Bid and Ask. If you want to additionally pad that spread, so be it. And as commissions per trade are very common these days, you must be able to account for this, too, or your results will not be very accurate. Empty4 does not allow for this OOTB...hence the use of tools to give Empty4 a kick in the butt!