THV MTF EA beta testing | Page 69 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

THV system, final edition 26,001 replies

MTF Stochastics into MTF Stoch Histogram 8 replies

THV ind 8 replies

My Journal - Based on THV V3.0 by Cobraforex 30 replies

THV RSI_V2 with Alarms --- need help! 6 replies

Attachments: THV MTF EA beta testing
Exit Attachments

THV MTF EA beta testing

  • #1,361
  • Jul 11, 2009 1:06am Jul 11, 2009 1:06am
To whom are you speaking?
  • #1,362
  • Jul 11, 2009 1:20am Jul 11, 2009 1:20am
@fxspeedster

please enable your PM or send me your email via PM to my acount.

thanks
AZBOfin
  • #1,363
  • Jul 11, 2009 1:24am Jul 11, 2009 1:24am
What you wanna say with that?

Nobody believes in backtests, only to calibrate the setting more that less it's usefull.

Quoting fxspeedster
This week the flavor is vanilla and next week is chocolate etc. Folks its time to get realistic.. please run some backtest instead of relying on useless 1,2,3 week accidental forward demo results.
  • #1,364
  • Jul 11, 2009 1:33am Jul 11, 2009 1:33am
Xux99

Would it be possible to add a "DumpToLogFile" option?

Whenever a trade is opened or closed, specific data such as ATR and trix levels, MFTTrix status, distances to cloud, coral and S/R levels, and whatever else triggered the open/close or is otherwise useful, will be written to a log file.

I know this will increase the load on the EA and CPU but it also could make the weekend job of analyzing the trades a hell of a lot easier.

AZBOfin
  • #1,365
  • Jul 11, 2009 6:22am Jul 11, 2009 6:22am
Hello AZBOFIN,

This is a very good idea.

Rgeards,
Jeff

Quoting AZBOfin
Xux99

Would it be possible to add a "DumpToLogFile" option?

Whenever a trade is opened or closed, specific data such as ATR and trix levels, MFTTrix status, distances to cloud, coral and S/R levels, and whatever else triggered the open/close or is otherwise useful, will be written to a log file.

I know this will increase the load on the EA and CPU but it also could make the weekend job of analyzing the trades a hell of a lot easier.

AZBOfin
  • #1,366
  • Jul 11, 2009 6:25am Jul 11, 2009 6:25am
Helloo Cobra,

FF is lucky to have you as a contributing member to its forum. Many thanks for everything you have contributed. You are truly a giver.

Regards,
Jeff

Quoting cobraforex
Hi Folks

Good that this newbee stepped in here, which gives me an opportunity to write what I’m thinking about this work.

I started this only because I saw that there is a high demand for a THV EA, some rat noses already tried to sell THV EA’s and I don't want any comercial suppose for THV.

I started with contracting Ronald Raygun who gave up saying that the indicators aren’t EA friendly.

Karun[font=Arial] who is professional programmer...
  • #1,367
  • Jul 11, 2009 9:49am Jul 11, 2009 9:49am
Quoting cobraforex
What you wanna say with that?

Nobody believes in backtests, only to calibrate the setting more that less it's usefull.
Ok let me explain. If you read most of the posts you will see statements like.. "I changed the PriceCrossIchimoku and the FastSlowTrixCross settings to (whatever..) and look at my screenshot of the last day (or few days of trading). So profitable! yupee!".
My question to you all is: is this the best way to find the most profitable settings for this EA? I know that Cobra you stated that nobody believes in backtests, but why is it that every other EA out there that that shows a good backtest (like MegaDroid for ex.) matches it in real time? If you have any doubt about the efficiency of the MT4 backtest please perform the following exercise. Make a printout of the last 10 trades that THV has executed live. Then run a backtest and with the same exact settings, you will match the live trades 100%. the MT4 backtest engine is known to have some issue but for most part I think people will agree that is reliable.
Also if somebody handed you an EA and told you it was a great performer.. would you forward test it to prove it OR would you first run a backtest and then demo trade it for a while? My point is that besides the longer term backtests I ran which show the EA not to perform well, there are no backtests here that show that certain settings are better than others. We need to find the good settings (and prove them) with a long term backtest.. this is my point.
If it backtests good run it live!
  • #1,368
  • Jul 11, 2009 12:05pm Jul 11, 2009 12:05pm
Quoting fxspeedster
there are no backtests here that show that certain settings are better than others. We need to find the good settings (and prove them) with a long term backtest.. this is my point.
What would recommend as a long term back test? I'm running a series of 6 month backtests ( Jan-Jun 2009 EURUSD) tests on different M15 sets files post here - including my own.

So far the one posted by Kaizasoze ( USDCAD M15 ) has performed the best :

Deposit 10000 , Net Profit 2321 , Abs. Drawdown 1532 ,

Wins 240 ( 96%) losses 8 ( 4%)

Heres one of those few big bad trades from the Kaizasoze backtest also showing the trade before and after:

407 2009.05.18 18:45 buy 204 0.12 1.35073 0.00000 0.00000
408 2009.05.18 18:55 close 204 0.12 1.35330 0.00000 0.00000 30.84 15103.95
409 2009.05.19 08:15 sell 205 0.60 1.35436 0.00000 0.00000
410 2009.05.22 17:00 close 205 0.60 1.39907 0.00000 0.00000 -2680.80 12423.15
411 2009.05.25 13:46 buy 206 0.50 1.40023 0.00000 0.00000
412 2009.05.25 14:31 close 206 0.50 1.40073 0.00000 0.00000 25.00 12448.15

As I've already mentioned these trades tend to last for several days - from 19th to 22nd March in this case. I wonder if ATR based Stop Losses get recalculated every day ? Eliminating these trades will be a big step forward IMHO.

Will post all the results when complete - probably sometime next week.

B Rgds

WR.
  • #1,369
  • Jul 11, 2009 1:42pm Jul 11, 2009 1:42pm
Quoting azjeff
Hello AZBOFIN,

This is a very good idea.

Rgeards,
Jeff

Hi Bofin and Jeff,

I agree we need to see what the EA is doing and what is triggering BUY/SELL/CLOSE on the trades.

Great idea!!!!

TradeHound
  • #1,370
  • Jul 11, 2009 1:45pm Jul 11, 2009 1:45pm
Quoting AZBOfin
have a great weekend everybody and try to stay cool while analyzing those charts from the past week

i for myself am completely stressed out. i have to go over those 345 demo accounts with the 1234 demo charts, so i can buy all the demo food and pay my demo cars and demo houses that will satisfy my 5 demo wives. that's why i have this demo rig setup.

AZBOfin

Nice trading office pic Bofin,
with all that one would think that they would have a more comfy chair, lol.


Happy pipping to all,

TradeHound
  • #1,371
  • Jul 11, 2009 2:01pm Jul 11, 2009 2:01pm
Quoting azjeff
Helloo Cobra,

FF is lucky to have you as a contributing member to its forum. Many thanks for everything you have contributed. You are truly a giver.

Regards,
Jeff

Could not agree more!!!!

TradeHound
  • #1,372
  • Jul 11, 2009 2:02pm Jul 11, 2009 2:02pm
Hello all and I hope everyone is having a nice weekend.With all this talk about optimization of the ea, I did a Google search to see what would come up with the terms Automated Optimization of a Trading Robot in Real Trading and I got a link to an ea and also an article on how to do it.Not knowing which options would be the best to optimize, I would not know if this method would be good to use with this ea since we have so many options to try?
the article can be found at http://articles.mql4.com/506
  • #1,373
  • Jul 11, 2009 2:10pm Jul 11, 2009 2:10pm
Hi Cobraforex, xux99, Caveman, anyone,

Please take a look at my post #1319 and let us have the ability to open/close trades based off of my post, please... I have been making serious $$$$ on my demo manually trading following the rules I posted. Once the trade goes into profit and hits the break even and lock in pips settings you can not lose on the trade.

Please take a look and allow us the option to open/close trades based off of the Entry/Exit rules on my post, thanks...

TradeHound
  • #1,374
  • Jul 11, 2009 2:24pm Jul 11, 2009 2:24pm
Quoting WackyRacer
What would recommend as a long term back test?...
Looking forward to your results. I will also run a backtest when Cobra will post the latest .set file. 6 months is plenty and it will give us a more reliable of what this EA can do and adjust for further more profitable settings.

Regards
If it backtests good run it live!
  • #1,375
  • Jul 11, 2009 2:29pm Jul 11, 2009 2:29pm
Quoting letizia
Hello all and I hope everyone is having a nice weekend.With all this talk about optimization of the ea, I did a Google search to see what would come up with the terms Automated Optimization of a Trading Robot in Real Trading and I got a link to an ea and also an article on how to do it.Not knowing which options would be the best to optimize, I would not know if this method would be good to use with this ea since we have so many options to try?
the article can be found at http://articles.mql4.com/506
This is an excellent article on optimization of an EA and solves some of the limitations of the MT4 optimization engine.
Perhaps Cobra can make a short list of the parameters that are considered to be vital in improving performance and we can start several optimizations cycles on those. We can then post the results for everyone and gather the most successful parameters into a standard preset file that we can all use.

Regards
If it backtests good run it live!
  • #1,376
  • Jul 11, 2009 3:06pm Jul 11, 2009 3:06pm
Quoting fxspeedster
This is an excellent article on optimization of an EA and solves some of the limitations of the MT4 optimization engine.
Perhaps Cobra can make a short list of the parameters that are considered to be vital in improving performance and we can start several optimizations cycles on those. We can then post the results for everyone and gather the most successful parameters into a standard preset file that we can all use.

Regards
I was looking at the article again, and I am wondering how hard would it be to code the ea either with the auto optimizer routine built in to the ea or have it call the optimizer program with the set parameters that are considered to be vital to the ea? Then have the ea optimize itself at a certain time?
Have a nice day!
Vince
  • #1,377
  • Jul 11, 2009 4:01pm Jul 11, 2009 4:01pm
Quoting fxspeedster
This is an excellent article on optimization of an EA and solves some of the limitations of the MT4 optimization engine.
Perhaps Cobra can make a short list of the parameters that are considered to be vital in improving performance and we can start several optimizations cycles on those. We can then post the results for everyone and gather the most successful parameters into a standard preset file that we can all use.

Regards

Thanks for looking this info up for us,

TradeHound
  • #1,378
  • Jul 11, 2009 4:02pm Jul 11, 2009 4:02pm
Quoting letizia
I was looking at the article again, and I am wondering how hard would it be to code the ea either with the auto optimizer routine built in to the ea or have it call the optimizer program with the set parameters that are considered to be vital to the ea? Then have the ea optimize itself at a certain time?
Have a nice day!
Vince
Wow, what an excellent idea!!!!


TradeHound
  • #1,379
  • Jul 11, 2009 4:08pm Jul 11, 2009 4:08pm
Quoting TradeHound
Thanks for looking this info up for us,

TradeHound
Glad I could contribute.
  • #1,380
  • Jul 11, 2009 4:24pm Jul 11, 2009 4:24pm
Quoting letizia
Hello all and I hope everyone is having a nice weekend.With all this talk about optimization of the ea, I did a Google search to see what would come up with the terms Automated Optimization of a Trading Robot in Real Trading and I got a link to an ea and also an article on how to do it.Not knowing which options would be the best to optimize, I would not know if this method would be good to use with this ea since we have so many options to try?
the article can be found at http://articles.mql4.com/506
Here is another article that does the same for indicators?
http://articles.mql4.com/766
Have a nice day!
Vince
0 traders viewing now