Old Dog with New Tricks | Page 44 | 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

  • #861
  • Aug 22, 2010 5:37am Aug 22, 2010 5:37am
You're not using the latest version. Also, make sure to set ECNBroker to true for IBFX.
You are in a maze of twisty little passages, all alike.
  • #862
  • Aug 22, 2010 5:54am Aug 22, 2010 5:54am
I have the lasted version posted by Ron on 4 Aug...
  • #863
  • Aug 22, 2010 7:08am Aug 22, 2010 7:08am
Quoting LeW
Hi Ron,

Your EA doesn't work correctly on IBFX broker. It opens 10 trades during one bar against one trade on each bar.
I'm using last version on M1 chart

Regards,
LeW
Try using steve's version. His code handles ECN brokers better than mine.
  • #864
  • Aug 22, 2010 11:58am Aug 22, 2010 11:58am
I am not filtering based on MinimumSS, only Win Rate and R/R.

Quoting getting
Hi Networkr75,

your MinimumSuccessScore=0.8 correct?
  • #865
  • Aug 22, 2010 2:22pm Aug 22, 2010 2:22pm
Hi networker75

Do you use a stevegee58 or Ron version of EA? On which broker?

Regards,
LeW
  • #866
  • Aug 22, 2010 4:00pm Aug 22, 2010 4:00pm
I am on a demo account from Alapari and I am using the "official" EA of Ronald.

Quoting LeW
Hi networker75

Do you use a stevegee58 or Ron version of EA? On which broker?

Regards,
LeW
  • #867
  • Aug 22, 2010 4:27pm Aug 22, 2010 4:27pm
4 days of demo and not 2.
Nothing strange with that, look at the csv files and you will see if you will get too many trades or not.
Besides, my max trades was set to 5 and not 10 like the default settings.
If you set the RR to .9 than you will get fewer trades.

Quoting getting
I'm curious to why you got so many trades( over 400) from only just two days of demo.
I'm demoing on Exness this week, and will post result on mt4i
bars = 125000 bars
WR = 75
RR = .9
test pair this week = GU nd UJ
  • #868
  • Aug 24, 2010 7:04am Aug 24, 2010 7:04am
Hey Ronald, I know I've asked you about this before - but I'm wondering why you think spread does not play a role in the optimization calculations? Surely a pip can make a difference when working out whether or not a trade would have been a win or a loss? It would be great if the optimizer could model based on the current spread, or an inputted spread even.

The reason I am concerned about spreads is that I have been running the EA successfully over the past 2-3 weeks on Alpari pro demo. Their spreads are great. The retail accounts I have, however, are anywhere between 0.5-1 pip higher on the majors. I wish I could use something like Alapri pro, but I simply don't have the capital at the moment. ThusI am wary about putting the EA onto a live account without some idea as to how the spread could affect the outcome. I guess I need to run it on a demo account with spreads more reflective of what is available to me at present. Some idea of what you think about this would be of help.

My other thought is - is it possible to implement some sort of emergency shutoff filte in the next versionr? You could set a maximum consecutive loss number, for example.

Thanks!
  • #869
  • Aug 24, 2010 7:20am Aug 24, 2010 7:20am
I think I explained how the backtester works, and assuming your stoploss / takeprofit are sufficiently large, spread should be a non-issue. I didn't design this to be a scalping EA, I designed this to be more of a swing-trading EA, which is immune to minor spread fluctuations, but not from news.

As for the other option, were you looking for something like: "if the win rate / risk reward / success score is away by a certain deviation, stop trading?"
  • #870
  • Aug 24, 2010 7:35am Aug 24, 2010 7:35am
Quoting Ronald Raygun
I think I explained how the backtester works, and assuming your stoploss / takeprofit are sufficiently large, spread should be a non-issue. I didn't design this to be a scalping EA, I designed this to be more of a swing-trading EA, which is immune to minor spread fluctuations, but not from news.
Well, for example - on USDJPY I've got my stoploss set to 15 pips, and the corresponding takeprofit ranges from 21-30 pips (winrate is 50%). Thus I've got a r/r of around 1.66. In this example, a pip or two could significantly impact the return, could it not? My other fear is that it could impact the implied win-rate.

Quote
As for the other option, were you looking for something like: "if the win rate / risk reward / success score is away by a certain deviation, stop trading?"

I guess that'd be one way of doing it. My idea was primarily aimed at significant market events where all of sudden it starts moving against what the EA believes to be the right action for that hour. e.g. it continuously breaks out while the EA keeps on counter-trading. I know you said this doesn't happen often, but would it not be useful to have some sort of cutoff feature?

e.g. Stop trading for this hour if x consecutive stoplosses have been hit.

My other thought was perhaps a news filter could help, or more an excessive spread filter. When significant news events occur the spread can spike massively and getting filled can be problematic. You could even get filled and stopped almost simultaneously.

Just some ideas...
  • #871
  • Aug 27, 2010 3:02pm Aug 27, 2010 3:02pm
Very quiet on here this week. Anyway I decided in an attempt to get a couple more trades to use the following settings this week on both EU and UJ.

SL10pips
min win rate 50%
Min success 0.8
bars tested 500,000

Its been a good week. IN total 46 trades 23 wins and 23 losses. Total profit 205 pips.

This is only on demo at the moment so will continue for a couple of weeks as I have had my fingers burnt on numerous occassions before where demo settings were great only for it not to work on live account. My only concern is the spread although most of my trades this week had a success score of over 1 therefore meaning my rr was at least 2 which ain't bad.

Paul
  • #872
  • Edited 3:17pm Aug 27, 2010 3:05pm | Edited 3:17pm
Quoting paulbt38
Its been a good week. IN total 46 trades 23 wins and 23 losses. Total profit 205 pips.


Paul
HOw is this possibile? If you have 50% wining rate (you have 23 wining and 23 losses) and wisk/reward less than 1 than you are losing and it wasn't a good week, there is no way you can have 200 pips profit with the results listed... please correct your post.
  • #873
  • Aug 27, 2010 3:08pm Aug 27, 2010 3:08pm
I got second losing week... 60%, 0.9 r/r on 200000 bars first days od the week lots of winings, last days of the week mostly losing. I am using 12 pairs.

I noticed something strange this week however....same setting 200000 bars give same results as sames settings 700000 bars... interesting...
Also can somebody confirm, same settings 30 pips stoploss gives same results as same settings 60 pips stop loss? Or there must be something wrong with my metatrader/system/EA...
  • #874
  • Aug 27, 2010 3:15pm Aug 27, 2010 3:15pm
I dont really understand why you are willing to take 40-50% losses. You dont need a self-optimizing statistical backtest for that, all you need is a coin.
  • #875
  • Edited 3:34pm Aug 27, 2010 3:23pm | Edited 3:34pm
Quoting rfking2
I dont really understand why you are willing to take 40-50% losses. You dont need a self-optimizing statistical backtest for that, all you need is a coin.
Placing orders by flipping a coin (random entries) you are losing, you know that you'll have to overcome the spread just to break even in the long term. Flipping a coin gives you about 49 % wining rate. So my logic is simple, in order to be profitable any system has to give you at least 55% winning rate, less than that you are breaking even or lossing, when you consider the spread. I tried higher settings eg winning rate but the EA hardly/if place any orders.
Please share your experience with this system.
  • #876
  • Aug 27, 2010 3:46pm Aug 27, 2010 3:46pm
My winrate is much higher, I dont have a min ss or rr. I do optimize by ss however. I trade every pair in demo watching performance. Live trade the best pairs (always changing) and I havent had a loosing week yet.
  • #877
  • Aug 27, 2010 3:49pm Aug 27, 2010 3:49pm
UJ has the highest winning rate, from looking at my trades and jupiter stats
http://www.mt4i.com/users/jupiter/stats
http://www.mt4i.com/users/getting/stats
  • #878
  • Aug 27, 2010 4:19pm Aug 27, 2010 4:19pm
Quoting iulli
HOw is this possibile? If you have 50% wining rate (you have 23 wining and 23 losses) and wisk/reward less than 1 than you are losing and it wasn't a good week, there is no way you can have 200 pips profit with the results listed... please correct your post.
No need to correct my post. Win rate was 50%. My risk/reward was not less than 1. With a minimum success score of 0.8 the minimum risk/reward would have to be 1.6. As I said on most occasions the risk/reward was more than 2.

Therefore the 23 trades that lost lost a total of 230 pips. However the 23 that won earned a total of 435 pips. Net result 205 pips. If you want I can post actual trades.

I haven't decided if these are the settings I want to use live but simply lowered then min win rate to try and get a few more trades as the week beofre with 0.65 wr and ss 0.8 I got no trades.

Paul
  • #879
  • Aug 27, 2010 5:14pm Aug 27, 2010 5:14pm
Quoting paulbt38
No need to correct my post. Win rate was 50%. My risk/reward was not less than 1. With a minimum success score of 0.8 the minimum risk/reward would have to be 1.6. As I said on most occasions the risk/reward was more than 2.

Therefore the 23 trades that lost lost a total of 230 pips. However the 23 that won earned a total of 435 pips. Net result 205 pips. If you want I can post actual trades.

I haven't decided if these are the settings I want to use live but simply lowered then min win rate to try and get a few more trades as the week beofre...
My bad... What stop loss did you use?
  • #880
  • Aug 27, 2010 5:39pm Aug 27, 2010 5:39pm
Quoting iulli
My bad... What stop loss did you use?
Stop loss was 10 pips (100)
0 traders viewing now