Old Dog with New Tricks | Page 36 | 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

  • #701
  • Aug 6, 2010 2:52pm Aug 6, 2010 2:52pm
Quoting Ronald Raygun
During that specific hour of the day.
So during another hour of the day it might "wake up" and go into breakout or counter mode?
You are in a maze of twisty little passages, all alike.
  • #702
  • Aug 6, 2010 2:52pm Aug 6, 2010 2:52pm
Quoting stevegee58
So during another hour of the day it might "wake up" and go into breakout or counter mode?
Yes
  • #703
  • Edited 3:20pm Aug 6, 2010 3:02pm | Edited 3:20pm
I am confused.. on eur/usd:

My setings (please see picture 1):

minimum win rate 60
minimum r/r 0
minimum success 0

Ea is trading and it shows:
minimum winrate 60
minimumr/r 1.2
success 72

As you can see my minimum coditions set are basic, and Ea shows me that at present time the market actually it is in a much better conditions of trading.


but if I change my settings by raising minimum a litlle bit more (please see picture 2):

winrate 60
r/r 0.7
success 60

it suppose to trade as it showed me earlier the conditions matched and are even better that what I set as minimum, however ea wont trade Keep in mind it is the same hour, same minute.

Is there a bug in the program?

p.s.(please disregard the sell order in the pictures, it was made manually)
Attached Image(s) (click to enlarge)
Click to Enlarge Name: e-u today.gif Size: 25 KB
Click to Enlarge Name: e-u today1.gif Size: 25 KB
  • #704
  • Aug 6, 2010 3:17pm Aug 6, 2010 3:17pm
Quoting iulli
I am confused.. on eur/usd:

My setings (please see picture 1):

minimum win rate 60
minimum r/r...
Interesting...

I'm not sure to be honest.
  • #705
  • Aug 7, 2010 2:24pm Aug 7, 2010 2:24pm
Quoting Ronald Raygun
Interesting...

I'm not sure to be honest.
I was going to post with exactly the same issue actually.

I was looking through the optimized csv data with default settings and decided to try setting a minimum successrate of 60 (there were many hours in the day for that pair which were marked with a higher success rate in the optimized csv). However, after restarting the EA, the optimized csv file had -1 for all entries - i.e. it could not find suitable results. I believe there may be a bug somewhere :/
  • #706
  • Aug 8, 2010 3:26pm Aug 8, 2010 3:26pm
The EA is taking hours and hours to optimize itself - never reached the end -
Although I am not having any resources issue (CPU 30%), I have tried it on two different PCs, with different MT4 compilations.
I have reduced the number of bars to optimize to 100 and still having the same problem.
I even changed the parameter EachTickMode to False to see if it is faster but same result...
I can see the animation during the optimization, with the number of cycles and ticks increasing and the smiley too!
I am on M1 chart and didn't changed the default settings.
I can see the optimization CSV files in ExpertsFiles

So wondering what's going wrong
  • #707
  • Aug 8, 2010 4:34pm Aug 8, 2010 4:34pm
Since the market hasn't opened yet you're not receiving any ticks. When the first tick comes in, it'll start optimizing and probably peg the CPU. I'd bump up the number of bars to like 50K since that's considered the minimum and it's very easy on the CPU anyway. Takes about 10 or 20 seconds on my crappy computer.
You are in a maze of twisty little passages, all alike.
  • #708
  • Aug 8, 2010 6:02pm Aug 8, 2010 6:02pm
Thanks Steve, it's working now the market is open.


Quoting stevegee58
Since the market hasn't opened yet you're not receiving any ticks. When the first tick comes in, it'll start optimizing and probably peg the CPU. I'd bump up the number of bars to like 50K since that's considered the minimum and it's very easy on the CPU anyway. Takes about 10 or 20 seconds on my crappy computer.
  • #709
  • Aug 8, 2010 6:52pm Aug 8, 2010 6:52pm
Science experiment #2

Previously I traded 1 pair (EURUSD) for 2 weeks. Everything default so MimimumWinRate = 50 and MinimumRiskReward 0. BarsToOptimize = 50000. This meant that I was taking less than favorable trades frequently and there was a large trade count.

I'm now trading 8 pairs with MimimumWinRate = 60 and MinimumRiskReward 1. This should give me fewer more favorable trades. It was a pain finding the right BarsToOptimize setting. I had been using 50000 and I wanted to increase it but my test computer only has 500 MB RAM. 400000 bars was too much and the computer started to thrash. When I dropped it to 300000 bars it was all able to fit in RAM with no paging.

The optimizing process took about 4 minutes per pair.
To pick the SL setting I used Ron's rule of thumb and set it to 10x the spread.
The pairs and SL values are:

AUDUSD,30
EURJPY,30
EURUSD,20
GBPUSD,40
NZDUSD,50
USDCAD,30
USDCHF,40
USDJPY,30

I'm trading a fixed size of 0.01. Because the SL values are quite varied my risk is also not constant. Also, the $ per pip is different for the non xxxUSD pairs so I'll have to take that into account as well. Next phase I want to equalize my $ risk per pair.

This experiment should at least give a baseline of profitability in pips.
You are in a maze of twisty little passages, all alike.
  • #710
  • Aug 8, 2010 7:15pm Aug 8, 2010 7:15pm
Quoting stevegee58
Science experiment #2

Previously I traded 1 pair (EURUSD) for 2 weeks. Everything default so MimimumWinRate = 50 and MinimumRiskReward 0. BarsToOptimize = 50000. This meant that I was taking less than favorable trades frequently and there was a large trade count.

I'm now trading 8 pairs with MimimumWinRate = 60 and MinimumRiskReward 1. This should give me fewer more favorable trades. It was a pain finding the right BarsToOptimize setting. I had been using 50000 and I wanted to increase it but my test computer only has 500 MB RAM. 400000 bars...
Nice I like your aproach, I am trying:
risk/reward 0.8
winrate 60
all pairs (10 pairs) with 30 pips sl and 30 pips TP
100000 bars.

Last week I wasn't satisfied with only minwinrate 50 bars 60000
  • #711
  • Aug 8, 2010 8:18pm Aug 8, 2010 8:18pm
Ive never seen my winrate break 60 on any pair.
  • #712
  • Aug 9, 2010 10:29am Aug 9, 2010 10:29am
Quoting iulli
Nice I like your aproach, I am trying:
risk/reward 0.8
winrate 60
all pairs (10 pairs) with 30 pips sl and 30 pips TP
100000 bars.

Last week I wasn't satisfied with only minwinrate 50 bars 60000
With the default of reverse trades true, Your 30 pip TP will be ignored by the bot.
  • #713
  • Aug 9, 2010 10:33am Aug 9, 2010 10:33am
Has anybody tried a global closure yet?
  • #714
  • Aug 9, 2010 10:48am Aug 9, 2010 10:48am
Quoting faze
Ive never seen my winrate break 60 on any pair.
Then something is wrong with your setup as I have had a couple over 70%...
We are our own best indicator.
  • #715
  • Aug 9, 2010 10:49am Aug 9, 2010 10:49am
Quoting stevegee58
Science experiment #2

Previously I traded 1 pair (EURUSD) for 2 weeks....
Thanks for sharing. I am running on ibfx, gomarkets and forex.com

waiting to go live though... So far so good on even 50% winrates....

Have it set up for 50, 60 & 70 to see where the sweetspot is. Had one at 90, no trades.

We are our own best indicator.
  • #716
  • Aug 9, 2010 1:20pm Aug 9, 2010 1:20pm
Am I correct in thinking that if after optimisation if I have the following showing that that means that the EA could not find a successful set of past trades from my requirements I gave it:

TP -1
Current Win Rate -100%
Current Risk Reward -1
Current Success Score -100
Array Win -1

etc.

I was using the following settings across 13 pairs.

Stoploss 300 (5 digit broker - with TP set to false)

Breakout Settings:
Barstooptimise 400,000
Min WR 70.0
Min RR 0.0
Min SS 90.0
ReverseTrades False

Then using optimisation based on Highers SS

Thus I wanted the EA to use a fixed stop loss of 30 pips (10 x spread on GU) and a min WR of 70% and SS > 90 (meaning a RR of 1.29). I'm using JadeFX thus have ECN set to true and am using the Stevegee58 version.

Thus, does the display I'm getting mean that my input settings are too high...e.g. I should lower the WR or SS or both?

Thanks
  • #717
  • Aug 9, 2010 4:17pm Aug 9, 2010 4:17pm
Quoting fxshadow
With the default of reverse trades true, Your 30 pip TP will be ignored by the bot.
you are right, I know that.

Can you guys post some results? All I know is that i Had this bot with default settings last 3 days last week on 10 pairs, and it lost heavily and there were also lots of orders... over 50 orders actually. I am trying different settings this week supposedly there would be less orders but more profit.. we'll see...

So let me ask you this.. What do you guys think, take profit suppose to be fixed or the stop loss? It is an interesting question... I think stop loss.... at least you know for sure how much you are going to lose in the worse case scenario
  • #718
  • Aug 9, 2010 4:20pm Aug 9, 2010 4:20pm
Quoting lovejoy80
Am I correct in thinking that if after optimisation if I have the following showing that that means that the EA could not find a successful set of past trades from my requirements I gave it:

TP -1
Current Win Rate -100%
Current Risk Reward -1
Current Success Score -100
Array Win -1

etc.

I was using the following settings across 13 pairs.

Stoploss 300 (5 digit broker - with TP set to false)

Breakout Settings:
Barstooptimise 400,000
Min WR 70.0
Min RR 0.0
Min SS 90.0
ReverseTrades False

Then using optimisation based on Highers SS

Thus...
i would lower success rate...
  • #719
  • Aug 9, 2010 4:23pm Aug 9, 2010 4:23pm
Quoting iulli
i would lower success rate...
Thanks...is that what TP -1 etc etc means then that settings are too 'high'.

I quickly tried it with very minimum SS and 0 WR and 0 RR and still got the same result?

I'll try again tomorrow with a different broker and see what happens. I'm surprised a WR>70 and SS> 90 doesn't produce any results though?
  • #720
  • Aug 9, 2010 4:23pm Aug 9, 2010 4:23pm
Quoting bluemele
Thanks for sharing. I am running on ibfx, gomarkets and forex.com

waiting to go live though... So far so good on even 50% winrates....

Have it set up for 50, 60 & 70 to see where the sweetspot is. Had one at 90, no trades.

How good was it on WR 50%? I am curious because I lost heavily last week... Was it a standard deviation of a coin toss or something more substantial?
2 traders viewing now