Great EA in backtest! - page 110
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
just uncheck "enable live trading" in the settings window. It will still make calculations but the order won't go to the broker. Let it run 2 weeks then allow it to trade.
just uncheck "enable live trading" in the settings window. It will still make calculations but the order won't go to the broker. Let it run 2 weeks then allow it to trade.
forgive my stupidity
2.1 The one trade per bar version
Well, sports fans. Score yet another anamolie for the cyberia. I built it and they didn't come.
I had no idea that simply eliminating mulptiple orders on the same bar would have such a devestating impact on the EA but it does.
I give you this just so you can play with it and learn from the experience, not to trade with obviously.
We are going to have to identify other patterns we can filter, this doesn't give us what we want.
maybe someone can tell me what to try next?
btw, it's great to be admired, really it is. Now if I could just be profitable too.
I find it interesting that no matter when I start this backtest I see virtually the same line...the first 14 trades suck, then the wind catches it's wings and the rest is bound for glory.
I think if I had the patience and discipline to let this run in my account with risk=.01 for two weeks and not touch it and not have my account crash or the power go out or the computer need to reboot or whatever...if it could just sit there in a padded cell and run by some miracle without my touching it all would be well...I doubt that will happen. well NO actually it wouldn't need to run for two weeks just long enough to make it's first 14-20 trades...how long is that I have to leave it alone? one week? week and a half?
Aaragorn,
I can't believe you are still chugging at this man. You my friend are like a drug addict looking for the next fix, you won't stop until you get it!
Hey man do yourself a favor, and step away from the computer for the weekend, and have a good weekend. If you're anything like me, your hands are glued to this damn thing day in, and day out! So to all the cyberia followers, Have a great weekend!
Oh and great job so far, how the support/resistance thing coming along??
Aaragorn,
I can't believe you are still chugging at this man. You my friend are like a drug addict looking for the next fix, you won't stop until you get it!
Hey man do yourself a favor, and step away from the computer for the weekend, and have a good weekend. If you're anything like me, your hands are glued to this damn thing day in, and day out! So to all the cyberia followers, Have a great weekend!
Oh and great job so far, how the support/resistance thing coming along??you've got my number, cyberia addict I am. well, let's just say I'm goal oriented.
ok ok compulsively goal oriented.
I guess there's no harm in putting this on the board, make of it what you will, just please include me in the profits if it helps any ok? it's not finished.
/en/users/daraknor has agreed to assist me with it. I'm hopeful that it will become more useful with his efforts
can anyone explain this to me? I've added some comments
BuyPossibilitiesMid is declared as a variable but never given a value until
if(BuyPossibilityCount != 0 ) returns equal to zero and that never happens as long as we are modeling more than 0 bars
then BuyPossibilitiesMid is assigned zero
this variable and it's corresponding variables for sell and unknowns are used all thru the cyberia logic as comparision factors but I never see anything act upon this variable besides assigning it equal to zero? what's up with this???
// We calculate the average probabilities of the events
// Probabilities of the purchase
BuyPossibilityMid *= BuyPossibilityCount; //starts @ 0,0
BuyPossibilityCount ++; //adds 1 for each modeling period
BuyPossibilityMid += BuyPossibility; //adds 0 to BP returns market activity unchanged
if(BuyPossibilityCount != 0 ) //condition is always positive as long as modeling 1 != 0
BuyPossibilityMid /= BuyPossibilityCount; //divisor = number of modeling passes // 0 / passes always returns 0
else //repeats above logic for sell
BuyPossibilityMid = 0;I challenge anyone to put this "BuyPossibilityMid" into the find next by clicking in the metaeditor edit tab then select find.... keep clicking find next in the window that opens which will walk thru the occurrances of this variable in the code....walk thru that and tell me what this variable is actually doing? if this thing is so based on probabilities I'd like to understand what the heck this is.
I mean it's created to be a variable but nothing ever acts upon it to make it vary????? where is the icon that shows my head exploding?
ok the last two days has been totally crappy on the euro. Has anyone else had a streak of losers like I have recently? I have got to get the support resistance filter going.
Since I started Euro on live IBFX mini, 10 straight losses -87 pips. Demo running on FXDD is giving some wins. I'm now trying out putting the opposite trade on, which should give wins instead.
USDJPY with Dave's settings 25 win, 11 loss, -25 pips.
Crappy so far.
I'm having a thought here....
more drawdowns occur when the trades lose in the same bar...
however...multiple wins also occur in the same bar....
but when does a multiple loss ever occur that doesn't start with the first loss?
when does a multiple win occur that doesn't start with the first win?
you see what I'm thinking?
if the first trade in the bar is a win or a loss could be used to determine if more trades will be allowed in the same bar...rather than just saying only allow one trade per bar say...
allow only one trade per bar IF the first trade in this bar lost.
the thing is that this sort of thing happens when it hits the s/l. and reopens again
are the wins in the same bar closing on s/l or t/p?
Answer: neither they are just closing which means that the EA is closing them based on it's probability analysis.
what does that tell me? that tells me that I can use the s/l closing as an indicator to disallow more trades in the same bar...let's see what that does.
nope that doesn't get it cause often it loses the first and wins the second.
I really think the key to this EA is learning how to win these tower patterns.
This is the current focus of my playing with the ATR in this version...
something is not working with it yet. results do not match settings..hum
btw I removed the pipsator from this just for the eurusd.
ya know, I really believe that the information is THERE in the EA to untangle these problem hours if we can only get it out of it and make sense of what indentifies them as they are forming so we can make correction action.
I think it will be a good weekend. I'm hopeful.
Look around all you cyberia testers. Examine your own reports and results and tell me...
IS there an occurance where you've really gotten in trouble with this EA where you didn't pull out of it reasonably quickly that WASN'T associate with one of these tower patterns?
If not then have we not isolated perhaps the biggest culprit? That's half the battle.
Note to self: Time delay before allowing next order after s/l close.