Minimizing noise in Trend identification?

[ironchef]

A poster from way back in 2003 acrary mentioned a minimum of 100 trades and putting that into 1/(sqrt(# trades)) = a standard error of 10% which would be acceptable.

Thank you.

 

[ironchef]

A poster from way back in 2003 acrary mentioned a minimum of 100 trades and putting that into 1/(sqrt(# trades)) = a standard error of 10% which would be acceptable.

In my day trading method, the profit/loss, a small number, is the difference of two large numbers and normally the R:R and the win rate are slightly above 1:1, 50/50. To defeat the null hypothesis (R:R=1:1, 50/50) may require more than a sample of 100 trades. 6 weeks of winning consistently could be an illusion and I don't want to scale up then find out it is an illusion.

Thanks again. It is my problem and I shouldn't bother you with it.

 

[Jzwu2017]

If you scale up 2x to $2k size you won’t lose a lot even you’re wrong. I’d say scale up to 2x and see what happens.

 

[ironchef]

If you scale up 2x to $2k size you won’t lose a lot even you’re wrong. I’d say scale up to 2x and see what happens.

I actually decided to go the other way, reduce size by 5x and stress test the hack out of the system. So starting Monday, I am trading for a cup of Starbuck instead of a Big Mac. But instead of making 5-10 trades and call it a day, I want to make 40-50 trades a day, from sun up to sun down and see what I get out of the system.

 

[Jzwu2017]

Either way just keep testing.

 

[themickey]

.......and see what I get out of the system

A headache and prove nothing.

 

[MarkTr]

Only increase timeframe is reliable to reduce noise.

 

[maxinger]

Only increase timeframe is reliable to reduce noise.

And when you increase the timeframe, you will notice that you will be missing
tons and tons of trading opportunities.

Then traders will become very angry/frustrated/mad and suffer sleepless nights.
Then traders will do revenge trading and those things.

 

[Jzwu2017]

Use long term timeframe to determine trend, i.e., long or short. Then, use local PA to determine entry point. Best of both worlds. Lol it’s not common in the trading world.

 

[murray t turtle]

And when you increase the timeframe, you will notice that you will be missing
tons and tons of trading opportunities.

%%
I, also noticed that, on increased time frame;
but with a lot of trend study, lots of trend measure tends to increase tons + tons of profit.
Longer time frames can also increase pound + pounds of profit;
but may not be quite as fun or have as much bid\ ask+ slippage fun
Noise may not be a bad thing;
shooting sports, hunting seasons are full of noise, ear muff makers sell plenty of ear muffs LOL.[Edit, Iron chief/ MACD =too far from price [for me], its below\ volume on charts ;
a moving average is near price, or easily glanced @ even if 200dma is far from price ]