[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

PR or Luck? Here is the real answer if you are looking for (Long but informative)

Posted By: Schigolch
Date: Thursday, 18 February 2016, at 11:24 p.m.

In Response To: PR or Luck? Here is the real answer if you are looking for (Long but informative) (__1ERROR1__)

That's a very interesting question, and an interesting approach to gain insight!

A few years ago, I've done some research of that kind myself. Other then you I measured the outcome with 1 (victory) and 0 (defeat). Then I took skill and luck from GNUBG analysis. I also separated skill into checker play and cube decisions like GnuBG does. While I like XG better, I still don't like its luck notion, the luck terms GnuBG applies suits much better.

Then I took a few other factors like match length etc.

I put all these explanatory variables in a logit model (sorry, now it comes to methodology), applied it to a database of about 30000 experience points and tried to produce significant findings.

My general finding was that in the short run both luck and skill add about 50% to victory.

Interestingly, match length was not significant, neither were other variables but one, see below. My feeling was that I should do some additional tests but I didn't find anyone with a deep knowledge in statistics to discuss the ensuing problems of the methodology with.

I had the feeling I should include a variable for "easy game" or "difficult game" but I didn't find a way to build a proper indicator variable. However, the variable I've computed (cumulated equity differences between the best move and the 2nd, 3rd move etc.) seemed to be significant - the lower this variable, the likelier a win. Meaning that if you have a match where you can't make errors that hurt, you're more likely to win it than when you have a match that allows you to produce serious errors).

As you I didn't find any correlation between skill and luck - and thus could not find support for this ongoing legend in backgammon.

I haven't looked at these things for quiet a few years.

However, I would suspect that the methodology applied to the problem bears heavy on the outcome of the analysis. A others have mentioned before, I wouldn't measure the score difference, but only the outcome as 1 or 0, and the explanatory variables should be continous if applicable. Second I have the feeling that the numbers that XG poduces for luck are not informative.

Dankwart

Messages In This Thread

 

Post Response
Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:
 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.