| |
BGonline.org Forums
XG: Performance rating as a yardstick?
Posted By: Klaus Evers
Date: Sunday, 11 April 2010, at 3:52 p.m.
I am very new to XG.
What I want out of new bot is a good measure of my ER in matches, how it compares to WC players and how it compares to my performance 2 or one year ago.
Therefore I need a superb anaylysis tool and a scale to look at.
For years, this has been GnuBG 4 ply with huge movefilter for me. My scale I looked at were checker errors per move, Snowie equivalent and finally with a grain of salt cube errors per move.
Now I wonder if I should use XG. Lets look at the nice table below that Iancho posted some days ago (Thanks!):
There are two columns that seem to be a good scale to compare the skill of players: PR (maybe THE replacement for the very familiar Snowie ER?) and match play performance.
Now lets look closer to some details, namely the lines of Michy, MCG and Mochy.
Michy and MCG have pretty similar numbers when it comes to PR, moves and cubes. But in match play performance, Michy is ~100 points lower than MCG!!!
This is even more shocking when you look at Mochy: his numbers in PR, checker and moves are worse than those of Michy and MCG, but his match performance is gigantic!
How can this be? (An even more obvious example is John Clark in the last line that confuses me pretty much)
I would have liked to use the match play performance as my new yardstick to measure and compare skill. But from what I see in this table it seems to be very inaccurate and not reliable.
Any thoughts? Am I misssing something? Thanks.
p.s.:
Where exactly can I find match play performance in XG and what is the difference to my Current Elo ?
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.