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

BGonline.org Forums

1.0 PR = 0.066 points per game; Question for XG

Posted By: Paul Weaver
Date: Thursday, 23 December 2010, at 11:58 p.m.

In Response To: Calculating edge needed to overcome rake (eXtreme Gammon)

eXtreme gammon wrote,

"1 PR = 33 Elo

50 Elo = 0.1 ppg"

This would mean that 1 PR = 0.066 points per game. Very interesting.

Does it follow that a difference in PR of 2 indicates an edge of 0.132 ppg? Does it follow that a difference in PR of 5 indicates an edge of 0.33 ppg?

If Player A wins 33 points (on average) in 100 money games against Player B, does this mean that Player A's PR is 5 points better than Player B's?

One who plays with a PR of 5.0 should expect to lose 33 points (on average) in 100 games to a man, woman or machine playing perfect backgammon.

Magriel told me thirty years ago that the best players (of the era) would lose at least 20 points in 100 games to a machine playing perfect backgammon. I told Robertie what Magriel had said, and Robertie said it was more like 30 points in 100 games. In 1984 Danny Kleinman said the best players (of the era) would come up short by only 5 points in 100 games. Danny was way far off the mark.

Although Malcolm can produce one of his recorded matches from the 1980s in which he played with a PR in the low 3s, I doubt if anyone in the early 1980s played better than 4.0 on average.

Question for eXtreme Gammon: A bot always gives itself a PR (or error rate) of 0.0. What do you estimate eXtreme Gammon's true PR to be?

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.