Not quite sure why it would be any different than running an ea on a crb or renko chart. Offline charts are offline charts right? I havent had many problems with running ea's on offline charts before.
Sixths trading - an EA by macman, Bob and Steve
- #585
- Oct 29, 2010 3:59am Oct 29, 2010 3:59am
for steve and other coders,
this is how i think a basket beast Ea could be made.
first you need to check the code in renko 3.2 EA and check at the function emulateonlivechart.
this function makes it possible to add EAS on the offline chart.
add that code into T101_Indy_PriceHistory_v3-dnk.mq4.
then save the indicator as an EA.
appy T101 EA on chart. open offline chart and then add the beast EA.
ofcourse the beast EA would need to be modified for this purpose.
that means when it hits the goldine back it should open a basket of specified pairs. in my case i run 14 pairs. but as can be seen in the T101 indicator. you could also add 8 pairs etc.
i use the DC autoblock indicator as follow.
i add it on 50 blocks.
i only trade yellow lines. no green ones.
i use lots of 0.10 for primary buys.
how i close the trades i use this mechanic.
stoploss i use 310 pips.
when crosses 110 pips close at 10 pips.
when crosses 210 pips close at 110 pips.
when crosses 310 pips close at 210 pips
when crosses 410 pips close at 310 pips.
when crosses 510 pips close at 410 pips.
that is basicly what i do.
also i think with macman his calculations on this T101 pair. it could be even better.
lets keep our minds open and i am all up for ideas
this is how i think a basket beast Ea could be made.
first you need to check the code in renko 3.2 EA and check at the function emulateonlivechart.
this function makes it possible to add EAS on the offline chart.
add that code into T101_Indy_PriceHistory_v3-dnk.mq4.
then save the indicator as an EA.
appy T101 EA on chart. open offline chart and then add the beast EA.
ofcourse the beast EA would need to be modified for this purpose.
that means when it hits the goldine back it should open a basket of specified pairs. in my case i run 14 pairs. but as can be seen in the T101 indicator. you could also add 8 pairs etc.
i use the DC autoblock indicator as follow.
i add it on 50 blocks.
i only trade yellow lines. no green ones.
i use lots of 0.10 for primary buys.
how i close the trades i use this mechanic.
stoploss i use 310 pips.
when crosses 110 pips close at 10 pips.
when crosses 210 pips close at 110 pips.
when crosses 310 pips close at 210 pips
when crosses 410 pips close at 310 pips.
when crosses 510 pips close at 410 pips.
that is basicly what i do.
also i think with macman his calculations on this T101 pair. it could be even better.
lets keep our minds open and i am all up for ideas
- #587
- Oct 29, 2010 5:56am Oct 29, 2010 5:56am
- | Commercial User | Joined Apr 2007 | 8,331 Posts
EA's run fine on off line charts as they get tick data from 1 minute chart. If no tick data no chart.
That apart, I do not understand the contents of post 585 and so can do nothing about it.
If anybody here does understand it and can help, then I can think about an implementation.
- #588
- Edited 6:44am Oct 29, 2010 6:32am | Edited 6:44am
I've had a bunch of those, on exotic pairs, gbpnzd, audnok, crap like that. I just ended up removing those pairs from trading as they popped up. Great fix huh
If we removed the pairs that give bad results , that means the Bot is not able to adapt to any market conditions, and running it on "good pairs" does not garanties that I will not crash when the market will change on these pairs. I am 100% sure it will have the same problems.
For example, I removed XAU, XAG, and some nordic pairs on my bot (MAXX Ea) because I get bad results with them, I noticed that XAG has rather a square wave rather a sine wave, ok it's good for the moment, but if one day EurUsd adopt a square wave style I will get the same problem (now EurUsd have a very smooth curb but few months ago it was very hurt style).
So don't choose the pairs that looks confortable for trading now because they will not stay like that for ever. More or less it's like curve fiitting with parameters, that is always the best way to go to crash sooner or later. That is the main reason why most Bots works for few months and crash later.
I love Macman idea to try to "crash a bot". I do same.
For sure, we are far to develop an universal EA, but at this stage, we have to keep this goal always in mind.
I have the same problem with my bot, but it's an easy way and not the best method to test a bot.
If we removed the pairs that give bad results , that means the Bot is not able to adapt to any market conditions, and running it on "good pairs" does not garanties that I will not crash when the market will change on these pairs. I am 100% sure it will have the same problems.
For example, I removed XAU, XAG, and some nordic pairs on my bot (MAXX Ea) because I get bad results with them, I noticed that XAG has rather a square wave rather a sine wave,...
In fact you can almost guarantee what works today is not going to work tomorrow. What I do is try & spot an unusual day before it can do too much damage.
I use a commercial program which looks at the strength of most pairs - normally it looks 'in balance', like today. If however I start to see everything going one way or major moves I am ready to intervene.
Hope that helps -
I use a commercial program which looks at the strength of most pairs - normally it looks 'in balance', like today. If however I start to see everything going one way or major moves I am ready to intervene.
Any hints to tell us what that programme you are using could be?
How has your account been overall?
- #596
- Oct 29, 2010 11:20am Oct 29, 2010 11:20am
Max is exactly right about not curve fitting by removing pairs for short term gains.
In fact you can almost guarantee what works today is not going to work tomorrow. What I do is try & spot an unusual day before it can do too much damage.
I use a commercial program which looks at the strength of most pairs - normally it looks 'in balance', like today. If however I start to see everything going one way or major moves I am ready to intervene.
Hope that helps -
could you watch if your commercial programm give more or less the same results that my KG strenght in Master Ea 18 ?
this could help if anyone can use my code freely
- #599
- Oct 29, 2010 12:56pm Oct 29, 2010 12:56pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Here is a Basket Scalper. Take 50-100 pips every half cycle. From top to bottom is a half cycle. Then on it again when it turns around opposite direction.
If we have 1000 pips every 24 hour we can try to figure out how to get ticks to an EA. With only T101_Indy_PriceHistory_v3-dnk its not working.
What I still do not understand is where T101_Indy_PriceHistory_v3-dnk is intended to do.
Nor do I understand how the basket is triggered. Until someone is able to make this clear, there is nothing I can do.