What should be the ideal win rate of a successful EA? | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Features of ideal Forex Broker? 11 replies

Ideal number of pairs 4 beginner 30 replies

IDEAL PRO is driving me nuts! Please help. 29 replies

[ask]ideal target pips around news 0 replies

Ideal S/L and T/P ? 1 reply

What should be the ideal win rate of a successful EA?

  • #1
  • First Post: Jul 23, 2015 4:31am Jul 23, 2015 4:31am
[Help] Need to clear a concept about EA Results:

Suppose there are two EAs, one produce 69% Win rate over 16 years of data.
Second one produce 45% win rate over 16 years of data.

Obviously you can say 1st one is better. But 2nd one produce larger win percentage in 1 year of backtest than 1st one. So I'm confuse at the point which one is good?

Plus how to judge an EA's performance. If an EA produce 45% win rate within a decade of backtest results. Will it able to produce at least 65% win rate on forward test for next 6 months?

According to you what should be the win rate of a successful EA ?
I know when number of sample increases the probability decreases, so if EA have low win rate for long term backtesting, shouldn't it be producing good results in short term forward testing?

Please forgive, If I asked silly questions. Just want to clear my concept.
Discussion is an exchange of knowledge;argument is an exchange of ignorance
  • #2
  • Jul 23, 2015 5:01am Jul 23, 2015 5:01am
Win rate does not mean anything if the ea loses money. For example,

90% win rate vs 50% win rate

50% win rate can be a winner and 90% can be a loser. The 50% win rate EA had risk reward ratio of 1:20 and 90% had risk reward ratio of 40:1.

For 50% 1 winnner would eradicate 20 losses and for 90% 40 winners would be eradicated with 1 stop loss.

The best way to judge an EA is ROI and maximum relative drawdown. It is all about percentage return with risk taken to achieve return, in mind.
  • #3
  • Jul 23, 2015 5:09am Jul 23, 2015 5:09am
Quoting Supertrader9
Win rate does not mean anything if the ea loses money. For example, 90% win rate vs 50% win rate 50% win rate can be a winner and 90% can be a loser. The 50% win rate EA had risk reward ratio of 1:20 and 90% had risk reward ratio of 40:1. For 50% 1 winnner would eradicate 20 losses and for 90% 40 winners would be eradicated with 1 stop loss. The best way to judge an EA is ROI and maximum relative drawdown. It is all about percentage return with risk taken to achieve return, in mind.
Thank you Supertrader9 for pointing it out.
Discussion is an exchange of knowledge;argument is an exchange of ignorance
  • #4
  • Jul 23, 2015 5:32am Jul 23, 2015 5:32am
better look at expectancy instead of winrate
  • #5
  • Oct 9, 2015 8:47pm Oct 9, 2015 8:47pm
Quoting PipPirates
{quote} Thank you Supertrader9 for pointing it out.

You are welcome
  • #6
  • Last Post: Edited 7:53am Oct 10, 2015 7:11am | Edited 7:53am
Quoting PipPirates
Suppose there are two EAs, one produce 69% Win rate over 16 years of data. Second one produce 45% win rate over 16 years of data. Obviously you can say 1st one is better.

Not "obvious" at all; maybe not even true.

Quite likely not to be true, actually: authors like Van K Tharp (Trade Your Way to Financial Freedom) and others take quite a bit of space in their books to explain why systems with win-rates around 35-45% tend, overall, to be a much better and more reliable resource than systems with win-rates around 55-65% or higher.
0 traders viewing now