| |
BGonline.org Forums
Luck
Posted By: Joe Russell In Response To: Luck (Stick)
Date: Monday, 12 April 2010, at 5:14 a.m.
When I talked with Stick I had a mixture of 5,7, 13, and 15 point matches, I took all the non seven pointers out and replace them. I posted this in another thread because I had not seen this one:
I did a quick, unscientific, survey of my last 100 7-point matches where there was a difference of .1 or more in the XG ER. I counted how many times the player that played the best won. The answer was 53. The lowest ER in each match averaged 3.3 and the highest ER averaged 4.8. There were five matches of the 100 where luck was graded insignificant by XG and the player that played the best won all five. Of the other 95 matches, where one player had more than insignificant luck, the player that played best won only one where he was unluckier. The player that was luckier won 94 out of those 95 matches. In 100 matches skill determined the result 6 times. I know there are some flaws with my methodology- I need to go back and count how many times the better player in the match also had more than insignificant luck-and the sample size was small, but I found it interesting that with a skill differential of 3.3 and 4.8, the better player was only a 53% favorite. I have been told each point of XG ER is worth ~26 ELO. 1.5 X 26 = 39. I checked to see what the ELO calculator in XG lists the advantage for a player with a 39 ELO edge. I used ELO's of 1980 and 1941 The answer was the lower ELO player should win 52.97% which, surprisingly, was in close agreement with my small sample survey. If you increase the match length to 15 points the advantage of the stronger player is only 54.34%
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.