Great EA in backtest! - page 8

[Deleted]  
integrityfx:
I found one broker that doesn't allow Cyberia to be run, i had to past the code into a new file and rename to get it to work... very strange but thats what did it... Please link me to 1.60 and i'll get on testing that as well. Thanks!

I downloaded 1.60 and found the same problem. Once i copied the code and pasted into a new expert it tested fine. This is using FXDD.

fxspeedster
434
fxspeedster  
integrityfx:
I downloaded 1.60 and found the same problem. Once i copied the code and pasted into a new expert it tested fine. This is using FXDD.

I find this really hard to believe! How do they control what backtests are ran and which EA is being tested? Makes no sense to me. But hey, it may be so. Are you able to do a backtest with 1.60 by renaming the EA? I now have come to realise that certain brokers do not like this 2-3 pip scalping activity for real trading. Those are brokers that bucket the trades and go against your positions. So of course they don't want to waste a dealer's time by going against these numerous small trades..it would kill one on the other end and they would get slaughtered even if they attempted. So in a way it tells us who the honest (oxymoron!) brokers are!

slam
19
slam  
fxspeedster:
well, I ran many backtests with version 1.50 (default settings) on M5 timeframes and I get very good stats with only 0.10 lots traded. See attachement. For some reason I cannot get the backtest with 1.60 running. It provides no output. Can you guys take a look and see why this is? pls post your recommendations to be able to run backtests on 1.60

Well, wouldn't you think forward testing is more indicative to actual real results than backtesting? Just a thought...

It seems there is a divergence in forward testing results from GP2X (who supposedly did "700k starting with 10k in 20 hours"), to having severe drawdowns from last nights testings using the same default settings and currency pairs. Curious to peoples thoughts as to why the dramatic difference from one days results to the next.

fxspeedster
434
fxspeedster  
slam:
Well, wouldn't you think forward testing is more indicative to actual real results than backtesting? Just a thought... It seems there is a divergence in forward testing results from GP2X (who supposedly did "700k starting with 10k in 20 hours"), to having severe drawdowns from last nights testings using the same default settings and currency pairs. Curious to peoples thoughts as to why the dramatic difference from one days results to the next.

The backtest of Cybertrader are totally meaningless. I explained why in an earlier post. The only valid testing is forward testing. If you can do backtests on M1 with accurate, high quality data going back 3-4 years then this would be valid. Any backtests ran on M5 or higher are meaningless since the intrabar M1 direction is unknown in the backtests.

fxspeedster
434
fxspeedster  
yeoeleven:
I am a novice trader with no knowledge of codes and programming but I have downloaded Cyberia Trader 1.6 and the presets.

I have it opened it on a specially created account on a thumb drive with a demo opening balance of $5000 using the presets as supplied on EUR/USD & USD/JPY using 1 minute charts.

The Journal tells me that both are loaded successfully and the Experts tell me that calculations are taking place starting at 8.45PM EST. Wed. Aug 2nd.

I may not be able to contribute by way of expertise but can test and report the results so expect an update after 12 hours. If this is inappropriate please tell me I am new around here and am trying to help.

Next to the Cyberia Trader in the top right of the chart the little emoticon is not smiling it is definitely frowning is this a problem?

John

You will need to enable the Expert trading. Click on the Expert Advisors on the top menu bar and make sure it is green. You should see the smiley face and the expert will start trading.

fxspeedster
434
fxspeedster  

backtest equity

For the skeptics, check out the equity test on a M1 backtest from 2000. it trades here 0.10 -0.30 lots and the test is not even 1/4 of the way through. Do you think this can make money??

slam
19
slam  
fxspeedster:
For the skeptics, check out the equity test on a M1 backtest from 2000. it trades here 0.10 -0.30 lots and the test is not even 1/4 of the way through. Do you think this can make money??

fxspeedster,

Again, backtests are great in "theory", but forward testing is more indicative to real life results (Please see post #65 for real forward testing results last night...BTW those demo accounts balances are even lower now since this morning).

I am not questioning this EAs possibly be a winning one. It is just there is such an extreme difference in results from everyone who has done backtesting and forwardtesting during the past two days. We hope to find the most quantifiable/accurate results, which will reflect a real live trading account environment. It seems we have not found that yet. Hopefully we will...

yeoeleven
840
yeoeleven  

Thanks fxspeedster

fxspeedster:
Attached find the latest Cybertrader and the presets for EUR and JPY M1. According to the deleloper's documentation it needs to be traded on M1 for best effectivess. Please post your comments and suggestions.

I am a novice trader with no knowledge of codes and programming but I have downloaded Cyberia Trader 1.6 and the presets.

I have it opened it on a specially created account on a thumb drive with a demo opening balance of $5000 using the presets as supplied on EUR/USD & USD/JPY using 1 minute charts.

The Journal tells me that both are loaded successfully and the Experts tell me that calculations are taking place starting at 8.45PM EST. Wed. Aug 2nd.

I may not be able to contribute by way of expertise but can test and report the results so expect an update after 12 hours. If this is inappropriate please tell me I am new around here and am trying to help.

Next to the Cyberia Trader in the top right of the chart the little emoticon is not smiling it is definitely frowning is this a problem?

John

Edited to add now that a trade has started the emoticon has started to smile!!!

fxspeedster
434
fxspeedster  
slam:
fxspeedster,

Again, backtests are great in "theory", but forward testing is more indicative to real life results (Please see post #65 for real forward testing results last night...BTW those demo accounts balances are even lower now since this morning).

I am not questioning this EAs possibly be a winning one. It is just there is such an extreme difference in results from everyone who has done backtesting and forwardtesting during the past two days. We hope to find the most quantifiable/accurate results, which will reflect a real live trading account environment. It seems we have not found that yet. Hopefully we will...

I agree. It is all about the settings. If you look at my results, you will see that the settings I have made money and continued to make money today. But it is impossible for me as just one person to test everything and come up with the best settings. We need to work together and get a lot more done, thats the only good option.

slam
19
slam  
fxspeedster:
I agree. It is all about the settings. If you look at my results, you will see that the settings I have made money and continued to make money today. But it is impossible for me as just one person to test everything and come up with the best settings. We need to work together and get a lot more done, thats the only good option.

Then there is definitely a discrepancy here. I have been running for the past 24hrs six different IBFX MT4's running Cyberia 1.50 and 1.60 in M1, M5, M15, M30, and H1 with all the presets that have been provided here on this EA's thread so far (with the exception of the Cyberia generic preset you posted today), and ALL have been consistently losing money. So, I do not understand how we are arriving to such different results.