SPOUTNIKH1V1.1 = 600 PIPS A WEEK (per currency) TRADING SYSTEM IS LOOKING FOR EA !!! - page 15
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Hi all,
My last contribution to this topic.
Again this posting is only to help if I sound hard, it is the only way to make myself clear.
I don't understand that here and in so many other topics, people come up with a system and there they sit, wachting like ducks on a pond to each other and wait what would happen...looking if someone will make an EA ..watching what the results would be if they start to trade this system....demo testing it for some time ...etc...
Instead of losing time why do most of you find it so hard to go over historical data manually and calculate what the results would have been ? (we don't talk about executing a brainsurchery)...
Don't fool yourselfs. Pick out periods were you don't see heavy trends but were price is moving more or less sideways. Januari and februari are good months to do some manual backtesting. Go over the chart take a pencill and write down every trade or signal that is been given. Take this results in to a spreadsheet and calculate what the total result was over that period. Look how many consecutive losing trade you would have. Calculate what the MaxDD would have been of those consecutive losses, look to the worst periods and ask yourselfs if you would have sticked to this or any other method wich gives you X amount of consecutive losing trades or X amount of MaxDD or X amount of loss.
The backtesting of MT4.0 is absolutly not to be trusted. Not even a litlle bit. The simplest test show that this backtesting is unreliable so you need to do it manually. I know that manually backtesting over a short period like 3 months doesn't proof anything so if you could do more data then do it. If you check 1 year data it will take you 1 compleet day to check that data but at the end of that day you will know if a system is crap or interesting enough to put more energy and time in it. If it proofs to be crap leave that system for what it is and move over to the next system and do the same.
People make some postings of results made over the last week. But if you would ask them what this system would have done over the last 3 months, they don't know and it's like they don't care. Is it that they hope that a pair will not get stuck in a consolidation period in the future, I don't know but I know that every system would have made profit over the last week. The system I introduced on the sbfx forum (I_FX_T) did extremely well but it underperformed over januari and februari. So start to check out the data from 1st januari till know.
My experience tells me that it is not worth while to follow this method because I does not beat a simple indicator. Maybe one only needs one indicator to have a good method ? ...But don't sit there as ducks waiting what the future will bring if the past is there to be investigated. Maybe this manuall backtest will proof that I'm wrong. If I see a posting of a spreadsheet with consistant good results or results with a bearable MaxDD, I will be the first to come back to this forum and admit that I overloocked something and that I'm wrong.
friendly greetings and keep up the good work....iGoRHey you guys,
This is my first post here. I registered recently and like the people here, hope you like me too. Now. on to what's on my mind. iGoR is correct, we should do a big manual backtest for each system that comes along. But I think that iGoR is incorrect in that each of us should test the system alone. This would take a hell of a long time and probably would still only let us see results over a period of one year or so. I think we should share the workload, let's say two members test 2000 (we need two so that we can compare the results and see who is making a mistake, if there are any inconsistencies), another two for 2001, and so on... This way we can share the load and in one week max we could tell with certainty: "this system sucks" or hopefully "yeah, this one is consistently making money". So, what do you think?
Hey you guys, This is my first post here. I registered recently and like the people here, hope you like me too. Now. on to what's on my mind. iGoR is correct, we should do a big manual backtest for each system that comes along. But I think that iGoR is incorrect in that each of us should test the system alone. This would take a hell of a long time and probably would still only let us see results over a period of one year or so. I think we should share the workload, let's say two members test 2000 (we need two so that we can compare the results and see who is making a mistake, if there are any inconsistencies), another two for 2001, and so on... This way we can share the load and in one week max we could tell with certainty: "this system sucks" or hopefully "yeah, this one is consistently making money". So, what do you think?
Neutrino,
Count me in !! I am willing to test. However in my opinion we need to ensure that the results are consistent and therefore need to employ the same methodology when testing.
I asked iGoR to provide some method for testing. Would you be able to document the steps that we should use for (manual) testing? In that way we all perform the tests using the same rules.
JJ/.
M5 Test
I test it in 5 minutes with "ask manual confirmation" ON,
and run ok.
CoyanCoyan,
When testing in the M5 do you change any settings ( besides "ask for manual confirmation")
jj/.
Neutrino,
Count me in !! I am willing to test. However in my opinion we need to ensure that the results are consistent and therefore need to employ the same methodology when testing.
I asked iGoR to provide some method for testing. Would you be able to document the steps that we should use for (manual) testing? In that way we all perform the tests using the same rules.
JJ/.That's why I'm proposing that at least two (maybe more) members test the same time period, so that if there are inconsistencies, we can always quickly double check.
Neutrion,
I can test any time period. By time period, u are refering to 00:00 to 24:00 and not Time frame i.e. M5/M15/M30 etc.
Again, I'm new to forex and Metatrader so some one will have to document the testing procedures for me. I will be able to test anything between 8:00 to 24:00 EST.
Regards...jj/.
Neutrion,
I can test any time period. By time period, u are refering to 00:00 to 24:00 and not Time frame i.e. M5/M15/M30 etc.
Again, I'm new to forex and Metatrader so some one will have to document the testing procedures for me. I will be able to test anything between 8:00 to 24:00 EST.
Regards...jj/.Yes, by time period I mean a month, 6 months, year and so on...
OK...not a problem. How do I extract and test the info via a spreadsheet as iGor suggested?
jj/.
OK...not a problem. How do I extract and test the info via a spreadsheet as iGor suggested? jj/.
Well, I'd like to know that too. Anyone with more experience care to share?
Coyan,
When testing in the M5 do you change any settings ( besides "ask for manual confirmation")
jj/.I think not.
I test it in demo and the profit was not well.
Now i am testing "Starter V6". You can find this ea in this forum.
Coyan.
I think not.
I test it in demo and the profit was not well.
Now i am testing "Starter V6". You can find this ea in this forum.
Coyan.Tanks for your reply.
I am a new newbie
. Although this is an excellent forum, I am discouraged at the amount of threads...I just can't read them all. I decided to stick to one method and see it through. This one appears to be relatively easy to follow (to start my forex education). So, I decided to stick to this one. If any one can help me document manual testing procedures (iGoR) , I would be glad to dedicate my time to this thread in order to make it work.
jj/.