1 Topic by geektrader (edited by geektrader 2022-12-14 00:40:08)

  • From: Germany
  • Registered: 2010-04-05
  • Posts: 835

Topic: Strange/wrong(?) R-Squared value

Hi Mr. Popov,

please take a look at this screenshot:

https://i.ibb.co/GP42msm/Untitled.png

It´s odd that the second strategy, which deviates so much from the ideal (red) line, still has a very high R-Squared value of >98, just like the other strategies (where it seems correct). Is this a bug or a calculation issue of the R-Squared? I´ve attached the strategy.

Thank you :-)

Post's attachments

Strategy Collection 1 GBPJPY H1 (2).json 5.81 kb, 4 downloads since 2022-12-13 

You don't have the permssions to download the attachments of this post.

geektrader's Website

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Strange/wrong(?) R-Squared value

Hello GeekTrader,

Thank you for the report!

Please check the questionable strategy in Express Generator with the same data.
The R-squared formula in Express Generator is calculated on the scaled balance chart in order to prevent overloading on higher number of bars.

Let's see the result and will decide what to do.

Popov's Website

3 Reply by geektrader (edited by geektrader 2022-12-14 13:42:46)

  • From: Germany
  • Registered: 2010-04-05
  • Posts: 835

Re: Strange/wrong(?) R-Squared value

Hi Mr. Popov,

the strategy was 100% generated in EA Studio. But I can re-evaluate it in Express Generator if that helps anything?

geektrader's Website

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Strange/wrong(?) R-Squared value

Yes, exactly. I wan to see if the  Express Generator formula works better for the same data.

Popov's Website

  • From: Germany
  • Registered: 2010-04-05
  • Posts: 835

Re: Strange/wrong(?) R-Squared value

I´ve compared them and it makes no real difference:

https://i.ibb.co/ncBNRNC/Untitled.png

geektrader's Website

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Strange/wrong(?) R-Squared value

My only explanation is that the effect appears when the values are lower.
It looks like for the higher profits the R-squared values are more conservative.

Popov's Website