Win rate seems to be random

[metatrader54]

Expectancy is definitely positive, but it differs from different currency pairs to different instruments and different years.

I guess in the end a huge part of it depends on "luck", after you've done and control what you can, the rest is for the market to decide.

For example, if a trending strategy coupled with a good year of volatility and trends would produce more money, compared to a year with low volatility and sideways.

As time passes, it feels like win rate is almost an insignificant stat. Just focus on good RR and probability trades as well as discipline, then that's all you can do.

Am I right?

 

[danielc1]

You are right. Win rate is over rated. If you know that you know what works, you need a mind set that understand how luck and skills are two separate things. If you know the difference, then you can let go of your personal win rate and just focus on your skills. A good book is thinking in bets from Duke.

 

[Overnight]

...

As time passes, it feels like win rate is almost an insignificant stat. Just focus on good RR and probability trades as well as discipline, then that's all you can do.

Am I right?

Win rate is 100% irrelevant in all things. Except for bragging rights. But the bottom line is profit.

If you have 99 wins for 1 dollar, and 1 loss of 100 dollars, your win rate is 99% but your net gain is negative 1 dollar.

 

[ValeryN]

Expectancy is definitely positive, but it differs from different currency pairs to different instruments and different years.

I guess in the end a huge part of it depends on "luck", after you've done and control what you can, the rest is for the market to decide.

For example, if a trending strategy coupled with a good year of volatility and trends would produce more money, compared to a year with low volatility and sideways.

As time passes, it feels like win rate is almost an insignificant stat. Just focus on good RR and probability trades as well as discipline, then that's all you can do.

Am I right?

"significant" or not depends on what you are using it for and how large is trades sample used to compute that number.

Since most people don't have good quality input to calculate those, it if pretty safe to say that regardless of use they will be insignificant.

Statistical significance requires thousands of samples.

Here are some examples:

A strategy that had 6779 trades over last 10 years. That is pretty significant. Or not. Depends on what you're planning for. But to keep it simple let's assume that it is.

Overall key stats look like that:

Now if we break it down by year:

You can see how much expectancy and profit factor fluctuates. Win rate is not too much since that strategy has a lot of trades in sample. But that is per year.

Just for fun, I broken down win rate by months for those 10y. And monthly expectancy is on the second pic.

And that is an excellent strategy that I am trading live.

Val