Firebird EA - fixed version - | Page 22 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Firebird EA, please test this one, it is one of the greats 1,228 replies

which version of firebird are you using on live account? 12 replies

Firebird V1.0 an amazing EA ! 33 replies

Going live with Firebird 38 replies

My firebird is a roast cornish game hen 3 replies

Attachments: Firebird EA - fixed version -
Exit Attachments

Firebird EA - fixed version -

  • #421
  • Aug 28, 2006 7:49am Aug 28, 2006 7:49am
Quoting goran
I am testing now FB v1.0b2 modified by Wackena and I only changed percent to 0.05 because I use it on 1M chart.I did the same with the previous version.Cheers.
HI Goran,

Can you please post this version..I can not find it ......can we please post all these in one place...it will help us all...

Thanks
Babar
  • #422
  • Aug 28, 2006 8:47am Aug 28, 2006 8:47am
Quoting babarmughal
HI Goran,

Can you please post this version..I can not find it ......can we please post all these in one place...it will help us all...

Thanks
Babar
Hi Babar,

for a version Firebird v1.0b2 see post 373 on page 25.For all updated versions check regularly this site: http://www.accrete.com/fx-mirror/mt4-firebird/
  • #423
  • Aug 28, 2006 10:15am Aug 28, 2006 10:15am
Greetings - I was wondering if anyone could tell me why one of my trades ticket #10625131 closed at -$7.80 GBPUSD...why -78 pips? I'm running FBv1.0b2 right out of the box at IBFX on $1000 demo. Also do I need to change percent=1.0 to .10? Or .05 percentlike goran said in post#418

Thanks for your help!

balancer
Attached File(s)
File Type: zip statement1.zip   2 KB | 384 downloads
balancer
  • #424
  • Aug 28, 2006 11:18am Aug 28, 2006 11:18am
Quoting piperwannabe
Greetings - I was wondering if anyone could tell me why one of my trades ticket #10625131 closed at -$7.80 GBPUSD...why -78 pips? I'm running FBv1.0b2 right out of the box at IBFX on $1000 demo. Also do I need to change percent=1.0 to .10? Or .05 percentlike goran said in post#418

Thanks for your help!

balancer
I believe most are using Percent = 0.05 or 0.1.

Wackena
  • #425
  • Aug 28, 2006 11:24am Aug 28, 2006 11:24am
Thanks Wackena...any ideas about the -78 pips?

balancer
balancer
  • #426
  • Aug 28, 2006 11:27am Aug 28, 2006 11:27am
I think we should move this forward testing...forward. Another question that needs answering is the MA TF and the percent value. My live trial is using v63G with 0.05 and doing nicely with 25-0 trades in 2.5 days.

Count me in for one of the trials but best if it's one which is 0700-2200 GMT (which is atrial in itself avoiding the asian session)
TRADE FOR SWAP
  • #427
  • Aug 28, 2006 11:39am Aug 28, 2006 11:39am
I agree cgldsmth, I'm ready to start with the formal forward testing so who
ever is taking the lead to divvy out various parameters just let me know b/c I'm ready to go!
balancer
  • #428
  • Edited 2:01pm Aug 28, 2006 12:16pm | Edited 2:01pm
Yes, ready to go. Had three MT4 IBFX demos running through the Sunday open yesterday running 63G-TS, FBv1b2, and FBv1c without a hitch. If there is going to be someone compiling a list of "sets" count me in on helping with one or two at least.

: ) Thom

The lastest "Draft" of the FireBird Resource page:
http://www.accrete.com/fx-mirror/mt4-firebird/
This is a permenant link and will not change so it is usable in the forums here as a pointer to the EA Archives and files.

This is a PostScript thought: For myself, and i don't mean to be a pain, i would actually prefer is there was an actual list of "To Dos" that included the exact set-parameters for a given EAversion. That was i would know i had not thrown in any of my bias or goof-ups in setting the test up.

Maybe??? >>> There could be a new thread set up by one of you FB-Leaders/Mods and in the first post you could list the posting rules that would make such a thread usable, i.e. keep all the general Q's over here? Then simply list each of the dozen or so "test-set-rules" and each of the indicators? Better yet, how about a zip file for each test that includes the exact EA with the settings for that test as the defaults? That would mean names like 63G-01, 63G-02, FBv1b2-001, etc??? But surely, if this was done we would have exact data to go by without the human factor involved? [and i am volunteering to do this last part also...so there would actually be a downloadable EA version with the name of each individual "forward-test-setting".]

Also?? >>> Remember that there is a resource page available and all the files could be placed there for the beta test. I'll spend some time over there today and include the PDF file and such. It's fun for me, and a blessing to help in such a way. : )
  • #429
  • Aug 28, 2006 12:21pm Aug 28, 2006 12:21pm
What does the "percent" setting signify?


Thanks!
  • #430
  • Aug 28, 2006 12:21pm Aug 28, 2006 12:21pm
Quoting cgldsmth
I think we should move this forward testing...forward. Another question that needs answering is the MA TF and the percent value. My live trial is using v63G with 0.05 and doing nicely with 25-0 trades in 2.5 days.

Count me in for one of the trials but best if it's one which is 0700-2200 GMT (which is atrial in itself avoiding the asian session)
Do I understand that you shut off your EA during the hours of 2300-600 GMT?

How do you do that? Do you close all open trades? Or, do you leave open trades but allow no new ones?

Thanks for any advice here, just trying to understand your methods.
Don Life is expensive, but includes a free trip around the sun.
  • #431
  • Aug 28, 2006 12:25pm Aug 28, 2006 12:25pm
Quoting forexpips
What does the "percent" setting signify?


Thanks!
A moving average is calculated.
Then that average is shifted x% upwards and x% downwards.

x=percentage
Just like envelope
May the pips be with you.
  • #432
  • Aug 28, 2006 12:27pm Aug 28, 2006 12:27pm
Thanks TraderSeven
  • #433
  • Aug 28, 2006 12:44pm Aug 28, 2006 12:44pm
Quoting piperwannabe
I agree cgldsmth, I'm ready to start with the formal forward testing so who
ever is taking the lead to divvy out various parameters just let me know b/c I'm ready to go!
I'm still working on it.

I really need help on coming up with the different variables. The people that have studied the statements can probably be of most help. I'm really kinda shooting in the dark with these, some may be unnecessary or obviously faulty. I need to know which settings logically have merit. Otherwise, I'm afraid we're going to have too many combinations to properly test.

The global ones (accross all versions) I've come up with so far are:

1. Time Period (on the chart)
2. Pairs - Here's on I need help on.
I think that we should use different setups with the top 2,
top 4,6,8, and 10 pairs. I've ordered the pairs from 1-10 based on
ATR, but is that the best way? Should we number them according to
risk, win ratio, what?
3. Brokers
4. MA Length
5. MA Timeframe
6. MA Type

With 4,5, and 6, we're talking hundreds, or maybe thousands, of
permutations.
7. Percent


Do we need to use different TP,SL,Pipstep settings too? Cause now we're talking even more. What about trading times? Do we need to set it to only trade in certain hours for some of the tests?

There's a section of the code dealing with divergence, iFX Analyzer I believe, that has multiple settings too (fast period/price, slow period/price). Do those warrant manipulation?

Then there are the settings exclusive to the multiple versions themselves, which shouldn't add too much to the complexity, so I've been saving them for last.

As if this weren't enough, we should probably run 2 or 3 of each setup allowing one to run unhindered, one reset every week, the other reset every other week.

I'm sure there are some more I haven't even come up with yet.

I hope everyone understands the magnitude of this. This would be completely overwhelming to try all the permutations I've outlined already, much less the ones I haven't. People that have a better understanding of the code need to tell me which ones we can eliminate or at least narrow the range. Those who have more experience in Forex need to analyze the statements and try to determine which pair/settings deserve more attention and what we can eliminate.

I think we just need to narrow it down a lot. I also think that the complexity of administrating this is so, that it's only going to happen once, so we better get it right.

Like I've said before, I'm willing to dole out the tests and track them, I'll even track the performance, but figuring out what all to test is going to take some (read a lot) help.

I've made several posts so far about the variables I've thought of and asked what people thought, but no one has really said anything. If everyone is content in doing their own thing, fine, just let me know so I stop wasting effort. Doing a test like this is hard work, but it's also necessary in my opinion if we want to determine the best way to set Firebird up and standardize it's performance. Otherwise, we'll all being doing testing accross different time frames, with different set ups (variables not under our control) and judging performance on what may be a fluke.
  • #434
  • Aug 28, 2006 12:46pm Aug 28, 2006 12:46pm
Quoting traderone
Do I understand that you shut off your EA during the hours of 2300-600 GMT?

How do you do that? Do you close all open trades? Or, do you leave open trades but allow no new ones?

Thanks for any advice here, just trying to understand your methods.
I just shut down my PC but all the open trades are left on the broker's server. there's so many variables we could test in this EA, and a decent test should be a month (at least).
TRADE FOR SWAP
  • #435
  • Aug 28, 2006 1:20pm Aug 28, 2006 1:20pm
Quoting piperwannabe
Thanks Wackena...any ideas about the -78 pips?

balancer
I don't know. This requires a little bit of investigation.

Can you determne if it was the order a main order or hedge order? In the Journal, was this order closed (close) or did it hit a stoploss (s/l)?

Wackena
  • #436
  • Aug 28, 2006 1:54pm Aug 28, 2006 1:54pm
I can help do some testing for Firebird.
  • #437
  • Aug 28, 2006 2:10pm Aug 28, 2006 2:10pm
I would be glad to help for testing too. All the best to everyone
  • #438
  • Aug 28, 2006 2:18pm Aug 28, 2006 2:18pm
Quoting Hagbard Celine
I'm still working on it.

I really need help on coming up with the different variables. The people that have studied the statements can probably be of most help. I'm really kinda shooting in the dark with these, some may be unnecessary or obviously faulty. I need to know which settings logically have merit. Otherwise, I'm afraid we're going to have too many combinations to properly test.
I think it will be very hard to decide what settings have merit at this point. You may need to limit testing to a few settings to start.

I have been testing on mainly one version using different settings on each demo. None of my tests have come close to Mono's results but that is not the point. Between back testing and now forward testing I may stumble on to a conservative safe earner in all market conditions. It takes a long time to optimize one currency pair. I won't be able to tell if any of this has helped me until I can test for a few weeks with at least one strong trend lasting several consecutive days in the test.

I hope a few people try settings completely different than what Mono and others have posted. On my live account my settings are close to Monos, but on demos I have been using v0.1a and adjusting the following settings on various time frames with most pairs:
MA_timeframe
Percent
TakeProfit
StopLoss
DivergenceLimit
PipStep
IncreasementType

Is anyone back testing and optimizing, and /or forward testing with adjustments to the more unusual settings?

Hagbard, I'm not suggesting trying to add a bunch of seldom changed setting to your test structure at this point. Instead I am hoping that if some people try a few unusual things on one demo they will be able to report any unexpected improvements after optimization and forward testing.
  • #439
  • Aug 28, 2006 2:47pm Aug 28, 2006 2:47pm
Quoting nvwine
I think it will be very hard to decide what settings have merit at this point. You may need to limit testing to a few settings to start.

I have been testing on mainly one version using different settings on each demo. None of my tests have come close to Mono's results but that is not the point. Between back testing and now forward testing I may stumble on to a conservative safe earner in all market conditions. It takes a long time to optimize one currency pair. I won't be able to tell if any of this has helped me until I can test for a few weeks with at least one strong trend lasting several consecutive days in the test.

I hope a few people try settings completely different than what Mono and others have posted. On my live account my settings are close to Monos, but on demos I have been using v0.1a and adjusting the following settings on various time frames with most pairs:
MA_timeframe
Percent
TakeProfit
StopLoss
DivergenceLimit
PipStep
IncreasementType

Is anyone back testing and optimizing, and /or forward testing with adjustments to the more unusual settings?

Hagbard, I'm not suggesting trying to add a bunch of seldom changed setting to your test structure at this point. Instead I am hoping that if some people try a few unusual things on one demo they will be able to report any unexpected improvements after optimization and forward testing.
hi nvwine,

have you any suggestion what for settings i can forward test? i ask because i started today some platforms (17 ) with FB on my other notebook, but the only thing i have changed was the timeframe. i test now the 5 versions on the 1min,30min,1h and 4H chart with the 30-300-30, and monos settings.

so if you can give me any suggestion what for for settings would be worth to test then i would start them today.

thx
  • #440
  • Aug 28, 2006 3:26pm Aug 28, 2006 3:26pm
Quoting ART_1_
hi nvwine,

have you any suggestion what for settings i can forward test? i ask because i started today some platforms (17 ) with FB on my other notebook, but the only thing i have changed was the timeframe. i test now the 5 versions on the 1min,30min,1h and 4H chart with the 30-300-30, and monos settings.

so if you can give me any suggestion what for for settings would be worth to test then i would start them today.

thx
I would suggest helping out Hagbard when his testing program starts. I'm not trying to compete or steer people away from an organized forward testing program. My point was it can be useful to experiment on one demo with some unusual settings.
You can try increasing Percent until you have far less trades or changing PipStep to get more or less trades. I think Mono was trying out the PipStep a couple of weeks ago. I can't really say for sure in this area because I am just trying different things. I take a setting and back test and optimize with it for a given pair. If it varies or improves the results I then try forward testing. You can be looking for improvements in the float, stoploss amount, how fast it earns money vs. how safe it is, how different pairs react to PipStep for example, or how it responds to different types of markets.
0 traders viewing now