Thread: what can we do with 90% winrate?

Originally Posted by ;

quote I meant opening gaps. Before opening the trade. The present price has already gapped away from the close since the place is open after the candle closes.

I see what you mean, let us say you wait to enter 12.00, lets instead input 11:59:59, which would remove the issue that I hope, then There'll be no opening gap because the next candle has to open before that is possible

I have yet Another idea I think could make the profitability better, if you could put that at the EA it Could be good,
the idea is Starting with Two lots or anything that can be splitted by Two, in the Goal we close Half or 1 the Place,
then we set the sl to brakeven and let the other part run to a higher tp, my Proposal Would Be to Utilize 1.1618 for the second tp,
hmm ,I guess - minus 1.618, from the other direction of This sl, I think That You get it, that would be the zero Degree plus Goal,
this will make it Feasible to score higher wins when the candle decides it wan'ts to retrace a lot,
we Can play around with different Goal and see which one is most consitent

Originally Posted by ;

I have another idea I believe could make the profitability better, if you could put that at the EA it could be good, the idea is opening with two tons or anything that can be splitted by two, in the target we shut 1 or half of the position, then we place the sl to brakeven and let another part run to a higher tp, my proposal would be to utilize 1.1618 for the second tp, hmm ,I guess - minus 1.618, from another direction of this sl, I believe that you get it, that would be the zero level plus target, this will make it possible to score higher wins when the candle decides...

Make things worse. 1.0 for the 2nd TP isn't any better

Originally Posted by ;

quote Make things worse. 1.0 for your 2nd TP isn't any better picture

okay, thats what testing is for :/

This is Version 1.10 of this EA that does not trade if the body size of the previous daily candle is greater or less than two user defined thresholds.
https://www.forexsoutheast.asia/atta...1518134960.ex4

From two above backtests we can observe that a win rate of approximately 70 percent can be anticipated from this particular egy.
If we include more powerful filters we may have the ability to increase the win rate, however it is not easy to reach to 90% win rate... I figure...

Originally Posted by ;

This is Version 1.10 of the EA that does not trade if the body dimensions of the prior daily candle is greater or less than two user defined thresholds. imageimagefile From two preceding backtests we could observe that a success rate of approximately 70% can be expected from this particular egy. If we add more powerful filters we might have the ability to increase the success rate, however it is not easy to achieve to 90% win rate...

that's great work, not that 70% is bad in any way I suspect that your trade sample isn't large enough though, I did marginally over 1000 transactions, yours is 700 something? Probability levels out around the tousands about if you filter the term bars that are long out, like if there is a candle body x amount of pips, then this transaction is avoided by the EA? I also believe the dojis should possibly be prevented since there is gains in people due to the tiny size, but maybe that is what you did? Fantastic idea in that case

Originally Posted by ;

quote That is excellent work, not that 70 percent is poor whatsoever I guess your trade sample is not big enough however, I did slightly over 1000 transactions, yours is 700 something? Probability levels out around the tousands about if you filter the term bars that are very long out, like if a candle figure is more than x amount of pips, then this transaction is avoided by the EA? I believe the dojis should be prevented because there is gains in those due to their size, but that is what you did? Fantastic idea in that situation

In Version 1.1, I've already applied a filter to not trade if the body size of the prior daily candle is larger or smaller than a user defined thresholds (i.e., it won't trade after lengthy momentum bars nor after doji's)
I shall backtest it with longer interval to around have 1000 trades.

Trend is your friend. Why dont buy green candle retracement sl?

This technique is changing the dimensions of the TP SL based on the daily candle that is present. A suggestion for testing, and simplifiion, would be to have SL dimensions and a fixed TP and base them on the ATR. This way we know the parameters of this transaction well in advance and all that we have to due is apply the filter for a candle that is too little or too large.

Just a thought.

GREEN pips to one and all.

Hi

Thanks for all the work done so far. It looks quite promising. I attempted to check out the EA MathTrader7_RoughEA. Ex4 but I am getting an error and no transactions and no data.

I've a picture of the mistake in the journal.

Thanks for your help.