what can we do with 90% winrate? | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Made a 100% Winrate EA 26 replies

Naustingfx End-Of-Day Trading System Journal @ 68.26% WinRate 34 replies

Winrate >90% - Unbelievable 4 me: omg 8 replies

MACD +60% Winrate 1:1.5 RR 63 replies

Statistical Gambling System (High Winrate, but NO EDGE) 372 replies

Attachments: what can we do with 90% winrate?
Exit Attachments

what can we do with 90% winrate?

  • #41
  • Oct 21, 2015 7:11am Oct 21, 2015 7:11am
Quoting nozzer
I am going to try this system on demo,see how it works out :]
thats great nozzer, I think you will like it​
Bulls are stupid Animals!especially when Im short!
  • #42
  • Oct 21, 2015 7:14am Oct 21, 2015 7:14am
Quoting roughtrader
{quote} with gaps there is going to sometimes be a loss, nothing we can do about it, but the good thing is that it might as well gap in our direction, the chances of that I think is 50/50 I think that's just something we have to live with
I meant opening gaps. Before opening the trade. Since the position is open after the candle closes, the current price has already gapped away from the close.
No greed. No fear. Just maths.
  • #43
  • Oct 21, 2015 7:16am Oct 21, 2015 7:16am
Quoting PipMeUp
No HG here. From April 2007 to today on E/U daily the winrate I get is 72.1%.
The pair I tested on is EURUSD, I have only tested EURUSD, maybe there is a tiny difference on other pair I don't know, please be advised that to get any reliable result, you must test it with 1000 trials minimum, after 1000 times probability tends to level out. so if you only have 500, that will not show you the real percentage
Bulls are stupid Animals!especially when Im short!
  • #44
  • Oct 21, 2015 7:26am Oct 21, 2015 7:26am
Quoting MathTrader7
{quote} You can find the EA in Post 29. :-)
Ops !sorry I was going through the posters one by one, I didn't realise you made an EA already, many thanks for your contribution both to you and also to abokwaik, this is what we need, alot of testing
Bulls are stupid Animals!especially when Im short!
  • #45
  • Oct 21, 2015 7:34am Oct 21, 2015 7:34am
Quoting abokwaik
This is a visual from EU 2015 test. @roughtrader : kindly confirm the logic. {image}
the logic, if the current days candle is up, we wait until it closes and end, on the exact moment we enter a sell order at market,
if the current days candle is down, we wait until close and enter a buy at the same second it closes, ofcourse the candles must stay up or down at the close,
if for example the dayily candle is up, and we wait to enter a sell at midnight, but before that it turns down, we would ofcourse enter a buy instead,
target you get by using the fib tool as described, note that it must be drawn from the close of the candle, not from the highs and lows as we usually do
Bulls are stupid Animals!especially when Im short!
  • #46
  • Oct 21, 2015 7:39am Oct 21, 2015 7:39am
Quoting PipMeUp
{quote} I meant opening gaps. Before opening the trade. Since the position is open after the candle closes, the current price has already gapped away from the close.
I see what you mean, let's say you wait to enter at 12.00, lets instead enter at 11:59:59, that would eliminate the problem I hope, then there will be no opening gap because the next candle must open before that is possible
Bulls are stupid Animals!especially when Im short!
  • #47
  • Oct 21, 2015 7:55am Oct 21, 2015 7:55am
I have another idea I think could make the profitability better, if you could put that in the EA it can be good,
the idea is -- opening with 2 lots or anything that can be splitted by 2, at the target we close 1 or half the position,
then we set the sl to brakeven and let the other part run to a higher tp, my suggestion is to use 1.1618 for the second tp,
hmm ,I guess - minus 1.618, in the other direction of the sl, I think you get it, that would be the zero level plus target,
this will make it possible to score higher wins when the candle decides it wan'ts to retrace a lot,
we could play around with different target and see which one is most consitent
Bulls are stupid Animals!especially when Im short!
  • #48
  • Oct 21, 2015 9:08am Oct 21, 2015 9:08am
Quoting roughtrader
I have another idea I think could make the profitability better, if you could put that in the EA it can be good, the idea is -- opening with 2 lots or anything that can be splitted by 2, at the target we close 1 or half the position, then we set the sl to brakeven and let the other part run to a higher tp, my suggestion is to use 1.1618 for the second tp, hmm ,I guess - minus 1.618, in the other direction of the sl, I think 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 is no better
Attached Image (click to enlarge)
Click to Enlarge Name: result.png Size: 85 KB
No greed. No fear. Just maths.
  • #49
  • Oct 21, 2015 10:39am Oct 21, 2015 10:39am
Quoting PipMeUp
{quote} Make things worse. 1.0 for the 2nd TP is no better {image}
alright, thats what testing is for :/
Bulls are stupid Animals!especially when Im short!
  • #50
  • Edited 12:30pm Oct 21, 2015 11:52am | Edited 12:30pm
Here is Version 1.10 of the EA that doesn't trade if the body size of the previous daily candle is less or greater than two user specified thresholds.
Attached Image (click to enlarge)
Click to Enlarge Name: BT.png Size: 45 KB
Attached Image (click to enlarge)
Click to Enlarge Name: BT2.png Size: 46 KB
Attached File(s)
File Type: ex4 MathTrader7_RoughEA.ex4   120 KB | 728 downloads


From two above backtests we can observe that a win rate of about 70% can be expected from this strategy.
If we add more effective filters we may be able to increase the win rate, but it is not easy to reach to 90% win rate... I guess...
Trading is the hardest way to make easy money...
  • #51
  • Oct 21, 2015 12:32pm Oct 21, 2015 12:32pm
Quoting MathTrader7
Here is Version 1.10 of the EA that doesn't trade if the body size of the previous daily candle is less or greater than two user specified thresholds. {image}{image}{file} From two above backtests we can observe that a win rate of about 70% can be expected from this strategy. If we add more effective filters we may be able to increase the win rate, but it is not easy to reach to 90% win rate...
That is great work, not that 70% is bad in any way I suspect that your trade sample isn't large enough though, I did slightly over 1000 trades, yours is 700 something? probability levels out around the tousands, how about if you filter out the long momentum bars, like if a candle body is longer than x amount of pips, then the EA avoids this trade? I also think the dojis should possibly be avoided because there is very small profits in those because of their tiny size, but maybe that is what you did? good idea in that case
Bulls are stupid Animals!especially when Im short!
  • #52
  • Oct 21, 2015 12:50pm Oct 21, 2015 12:50pm
Quoting roughtrader
{quote} That is great work, not that 70% is bad in any way I suspect that your trade sample isn't large enough though, I did slightly over 1000 trades, yours is 700 something? probability levels out around the tousands, how about if you filter out the long momentum bars, like if a candle body is longer than x amount of pips, then the EA avoids this trade? I also think the dojis should possibly be avoided because there is very small profits in those because of their tiny size, but maybe that is what you did? good idea in that case
In Version 1.1, I have already applied a filter to not trade if the body size of the previous daily candle is larger or smaller than two user specified thresholds (i.e., it won't trade after long momentum bars nor after doji's)
I will backtest it with longer period to approximately have 1000 trades.
Trading is the hardest way to make easy money...
  • #53
  • Oct 21, 2015 1:00pm Oct 21, 2015 1:00pm
Trend is your friend. why dont u buy daily green candle retracement sl below the low of the candle?
  • #54
  • Oct 21, 2015 1:37pm Oct 21, 2015 1:37pm
This method is varying the size of the daily TP & SL based on the current daily candle. A suggestion for testing, and simplification, would be to have a fixed TP and SL size and base them on the ATR for the last week, month or whatever length you want. That way we know the parameters of the trade well in advance and all that we have to due is apply the filter for a candle that is too big or too small.

Just a thought.

GREEN pips to one and all.
  • #55
  • Oct 21, 2015 1:54pm Oct 21, 2015 1:54pm
Quoting goldfinger
Trend is your friend. why dont u buy daily green candle retracement sl below the low of the candle?
you mean go with the trend? I guess that could be tested also, but the general idea is to play the retrace, action reaction, but no ideas are bad, if we can test that we should
Bulls are stupid Animals!especially when Im short!
  • #56
  • Oct 21, 2015 1:54pm Oct 21, 2015 1:54pm
Hi

Thanks for all the work done so far. It looks very promising. I tried to test the EA "MathTrader7_RoughEA.ex4" but I am getting an error and no trades and no data.

I include a picture of the error in the journal.

Thanks for your help.
Attached Image (click to enlarge)
Click to Enlarge Name: EA TESTING 10 21 15.png Size: 20 KB
  • #57
  • Oct 21, 2015 1:55pm Oct 21, 2015 1:55pm
Quoting SmoothTrader
This method is varying the size of the daily TP & SL based on the current daily candle. A suggestion for testing, and simplification, would be to have a fixed TP and SL size and base them on the ATR for the last week, month or whatever length you want. That way we know the parameters of the trade well in advance and all that we have to due is apply the filter for a candle that is too big or too small. Just a thought. GREEN pips to one and all.
good tip, thanks
Bulls are stupid Animals!especially when Im short!
  • #58
  • Oct 21, 2015 1:56pm Oct 21, 2015 1:56pm
I played around a bit with the tp and sl and got it close to 90% , a bit longer period too, however for some reason it takes a dive in the middle, we need to figure out why and create a filter for that whatever that might be
Attached Image (click to enlarge)
Click to Enlarge Name: report1.jpg Size: 170 KB
Bulls are stupid Animals!especially when Im short!
  • #59
  • Oct 21, 2015 2:00pm Oct 21, 2015 2:00pm
risk 3
reward 1

trade the opposite of last close / daily
  • #60
  • Oct 21, 2015 2:01pm Oct 21, 2015 2:01pm
Quoting roughtrader
{quote} you mean go with the trend? I guess that could be tested also, but the general idea is to play the retrace, action reaction, but no ideas are bad, if we can test that we should

yep we can buy/sell at the 80.9/19.1 to target the 1.618/-0.618 to get us a profit of 80.9% total....61.8+ 19.1= 80.9 or else buy/sell the second time through as shown
the win rate will be higher because we are with the hourly short term trend... + rr is better

would love to see the results for that one on 1000 trade backtest
Attached Image(s) (click to enlarge)
Click to Enlarge Name: as shown.png Size: 68 KB
Click to Enlarge Name: as shown 2.png Size: 92 KB
0 traders viewing now