Old Dog with New Tricks | Page 8 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Old Dog's Taming of the Beast? 34 replies

Old Dog Fib Trading - EA Development 18 replies

Old Dog Fib Trading - EURUSD 399 replies

5 Tricks of Trend Trading, by Rob Booker 25 replies

How To Work Around A Market Maker's Tricks 0 replies

Attachments: Old Dog with New Tricks
Exit Attachments

Old Dog with New Tricks

  • #141
  • Jan 13, 2010 6:15am Jan 13, 2010 6:15am
When I download the files into meta editor they appear like this

EX4




So I can't compile them and run them, why is this?
The greatest force on earth is compound interest.
  • #142
  • Jan 13, 2010 7:12am Jan 13, 2010 7:12am
Quoting Ronald Ray gun
What did trading style say?
Did trade trigger say anything else?

My suggestion would be to set alerts, and e-mail to true (make sure you set up e-mail) then you can at least know if the EA attempts to make a trade.
I'm at work right now . . . I believe style was trend trigger not sure
I'll set the e-mail function

but in general are these settings doable ?
  • #143
  • Jan 13, 2010 8:41am Jan 13, 2010 8:41am
Quoting dryclean
I'm at work right now . . . I believe style was trend trigger not sure
I'll set the e-mail function

but in general are these settings doable ?
Yes. The only thing that stands out to me is the moneymanagement. If the account size isn't large enough, then it's likely the size will be too small that the EA rounds to 0.
  • #144
  • Jan 13, 2010 8:58am Jan 13, 2010 8:58am
would a setting of 3 work?
  • #145
  • Jan 13, 2010 9:30am Jan 13, 2010 9:30am
If your account size is ~300k
  • #146
  • Jan 13, 2010 10:21am Jan 13, 2010 10:21am
What is the minimum account size it could function off? Ive just got the EA running now. but not placing trades.
The greatest force on earth is compound interest.
  • #147
  • Jan 13, 2010 11:00am Jan 13, 2010 11:00am
Quoting sebastionay
What is the minimum account size it could function off? Ive just got the EA running now. but not placing trades.
This applies to the MM function, but 1 percent I believe requires 100,000 to run 0.1 lots. 10,000 for 0.01 lots
  • #148
  • Jan 13, 2010 12:13pm Jan 13, 2010 12:13pm
Why can't we apply a trailing stop?

Do you think it would be useful if we had a seperate option for TP/SL for breakouts as we do for counter trend trades? I would expect breakouts to be larger for example.

Ive been playing with the system for a while now, it seems good.
The greatest force on earth is compound interest.
  • #149
  • Jan 13, 2010 12:44pm Jan 13, 2010 12:44pm
Quoting sebastionay
Why can't we apply a trailing stop?

Do you think it would be useful if we had a seperate option for TP/SL for breakouts as we do for counter trend trades? I would expect breakouts to be larger for example.

Ive been playing with the system for a while now, it seems good.
Problem with implementing a trailing stop is the actual results will not be anything like the internal backtest.

As for separate TP/SL for breakouts and counter trades, that's interesting. I can see the potential for this applied on an individual basis in other EAs. I'll get around to putting together a coding template exploring this.
  • #150
  • Jan 13, 2010 12:49pm Jan 13, 2010 12:49pm
I don't know how easy this would be to implement but how about if time became included in the decision making process.

If we assume downward swing's happen faster than rallies. for counter trending shorts for example if we had the system recognise that a counter trend trade should be entered and exited faster than a breakout going north.
The greatest force on earth is compound interest.
  • #151
  • Jan 13, 2010 1:44pm Jan 13, 2010 1:44pm
Could you elaborate?
  • #152
  • Edited 7:04pm Jan 14, 2010 11:32am | Edited 7:04pm
I mean that I think sell offs happen much faster in proportion to the amount they move and also as a result happen over shorter time periods than rallies.

I would expect its just because fear has a stronger effect than faith in most peoples buying decisions.

Anyway I think it would be interesting if we had an option to decide the time of a trade dependant on its type for instance CT trades open for 60% of the time a breakout trade is open for in proportion.

Ive got a high CT trade success score here but the EA still has not placed a trade. (96% anyone got it higher?)
Attached Image (click to enlarge)
Click to Enlarge Name: high counter %.gif Size: 47 KB
The greatest force on earth is compound interest.
  • #153
  • Jan 14, 2010 12:00pm Jan 14, 2010 12:00pm
Did any of the alerts go off? (Were they set?)
  • #154
  • Jan 14, 2010 12:12pm Jan 14, 2010 12:12pm
I didn't have alerts on, changed the setting to true now, what about the email alerts and signals should I have them on aswell to trigger the most trades possible?
The greatest force on earth is compound interest.
  • #155
  • Jan 14, 2010 12:23pm Jan 14, 2010 12:23pm
Signals only isn't necessary
Email works if you set up the terminal
Alerts enables the pop-up.
Sounds aren't really useful...
  • #156
  • Jan 14, 2010 5:17pm Jan 14, 2010 5:17pm
I've been trying all sorts of combination's, and still can't get the ea to place any trades on my demo account. I even have used 0 win percent; 0 min success score; and 10 min. confidence. I've been trying different combinations for a few weeks now. Maybe the problem is I keep tinkering with it, but has the ea placed any trades for anyone during the last two weeks?
  • #157
  • Jan 14, 2010 5:22pm Jan 14, 2010 5:22pm
Quoting trader_scott
I've been trying all sorts of combination's, and still can't get the ea to place any trades on my demo account. I even have used 0 win percent; 0 min success score; and 10 min. confidence. I've been trying different combinations for a few weeks now. Maybe the problem is I keep tinkering with it, but has the ea placed any trades for anyone during the last two weeks?
In order for the EA to take a trade:

 

  1. One of the two options (breakout/counter) must be above the win percent and the other option must be below it.
  2. Trading style should not say "Stand Aside"
  3. The high and low lines should not be yellow.

I'll be back at my main computer by Sunday at the latest. I've spent the last 4 weeks taking winter classes, so things have been quite hectic. I can't wait for some normalcy.

  • #158
  • Jan 14, 2010 7:28pm Jan 14, 2010 7:28pm
Quoting Ronald Raygun
In order for the EA to take a trade:

 

  1. One of the two options (breakout/counter) must be above the win percent and the other option must be below it.
  2. Trading style should not say "Stand Aside"
  3. The high and low lines should not be yellow.

I'll be back at my main computer by Sunday at the latest. I've spent the last 4 weeks taking winter classes, so things have been quite hectic. I can't wait for some normalcy.

RR -- Thx for getting back to me. I think my problem is not having one of the options above, and the other below the win percent. I've adjusted it, and I'll let you know how it goes.
  • #159
  • Jan 15, 2010 2:35am Jan 15, 2010 2:35am
[quote=Ronald Raygun;3325825]This EA really can't backtest because it uses historical bar data to generate its statistics.


Hi RR, first of all thanks for starting an interesting thread. I was wondering when someone was going to start a thread on "smarter" EAs using simple and profitable strategies. I've done a little work myself building auto-optimisers into EAs (the kinds that find the best parameters for a particular strategy), and know how much work this can be for even a simple trading system. Nice work on such a thorough job, particularly with the performance analysis and resulting stand-aside logic.

A couple of questions; I found my auto-optimising EAs to be very sensitive to the number of bars of history you feed it, and how often it is run. Do you find that yours has similar characteristics?

I've often thought about using a MTF optimising type strategy (the code is complex enough hence I haven't yet tried it) i.e. if profitable in last month AND in last three months then place trades, otherwise stand aside. Have you tested any of these types of scenarios with any of your EAs?

Also, you mentioned that your EA can't be backtested; is there any special reason for it? I would have thought that your EA and optimiser could run in the MT4 backtester (given enough processing power and time)?

I'm also interested to hear more about your brigade of EAs that you use (I'm working on a similar thing but they are not yet profitable); are these mainly based around similar types of breakout strategies or are you using other types of strategies as well? I certainly understand if you do not wish to disclose this, and sorry if I'm getting a little off topic here.

Thanks Michael.
  • #160
  • Jan 15, 2010 4:35am Jan 15, 2010 4:35am
Quoting PipMasterMik
A couple of questions; I found my auto-optimising EAs to be very sensitive to the number of bars of history you feed it, and how often it is run. Do you find that yours has similar characteristics?
EA re-calculates its stats at the beginning of each new day ( During the period when the EA is also calculating the new day's starting range).

Quoting PipMasterMik
I've often thought about using a MTF optimising type strategy (the code is complex enough hence I haven't yet tried it) i.e. if profitable in last month AND in last three months then place trades, otherwise stand aside. Have you tested any of these types of scenarios with any of your EAs?
Yes. As I mentioned earlier, I did it with multiple EAs running simultaneously instead of trying to code everything into one EA. The resulting code is far easier to maintain.

Quoting PipMasterMik
Also, you mentioned that your EA can't be backtested; is there any special reason for it? I would have thought that your EA and optimiser could run in the MT4 backtester (given enough processing power and time)?
No. The backtesting engine is built directly into the EA. There is no interaction between the EA and strategy tester.

Quoting PipMasterMik
I'm also interested to hear more about your brigade of EAs that you use (I'm working on a similar thing but they are not yet profitable); are these mainly based around similar types of breakout strategies or are you using other types of strategies as well? I certainly understand if you do not wish to disclose this, and sorry if I'm getting a little off topic here.

Thanks Michael.
It's all in my profile. My apologies if I seem rude, but I don't think this last bit is appropriate for this thread.
0 traders viewing now