BGonline.org Forums

Error Rates and Edges

Posted By: Daniel Murphy
Date: Monday, 24 May 2010, at 9:56 p.m.

In Response To: Error Rates and Edges (Matt Cohn-Geier)

There is no way to translate ER to MWC statistics.

Why do you say that?

Supposedly ER advantage translates to Elo advantage. 36 or so Elo points per ER point difference, I'm reading. And Elo advantage does translate to MWC, not perfectly, but pretty well. If two equally rated players play, say, a few hundred 5-point matches, and at the end of that session one is rated 100 points higher, you know that he won 56.4% of the matches. The obvious converse is that if those two players' true Elo ratings differ by 100, the one is a 56.4% favorite in 5-point matches, and an X% favorite at other match lengths.

There isn't a perfect correspondence, of course, that fits all pairings. One player may be better at DMP, or at 5-pointers, or 25-pointers, so the Elo formula doesn't apply perfectly. And ER covers a multitude of errors, and don't necessarily predict how a particular player will do against another.

Messages In This Thread

• Error Rates and Edges
  Bill Calton -- Monday, 24 May 2010, at 6:08 p.m.
  • Error Rates and Edges
    Stick -- Monday, 24 May 2010, at 7:51 p.m.
    • Or use XG's Elo Calculator
      neilkaz -- Monday, 24 May 2010, at 8:14 p.m.
      • Or use XG's Elo Calculator
        Leonardo Jerkovic -- Monday, 24 May 2010, at 9:31 p.m.
        • Or use XG's Elo Calculator
          Daniel Murphy -- Monday, 24 May 2010, at 10:07 p.m.
          • Or use XG's Elo Calculator
            Leonardo Jerkovic -- Monday, 24 May 2010, at 11:37 p.m.
            • XG use the same Elo formula than FIBS
              eXtreme Gammon -- Monday, 24 May 2010, at 11:49 p.m.
              • XG use the same Elo formula than FIBS
                Leonardo Jerkovic -- Tuesday, 25 May 2010, at 12:29 a.m.
                • XG use the same Elo formula than FIBS
                  Daniel Murphy -- Tuesday, 25 May 2010, at 12:57 a.m.
                  • XG use the same Elo formula than FIBS
                    Leonardo Jerkovic -- Tuesday, 25 May 2010, at 1:18 a.m.
                    • XG use the same Elo formula than FIBS
                      eXtreme Gammon -- Tuesday, 25 May 2010, at 2:07 a.m.
      • Or use XG's Elo Calculator, yes 1 PR=33 Elo
        eXtreme Gammon -- Monday, 24 May 2010, at 11:47 p.m.
  • Error Rates and Edges
    Matt Cohn-Geier -- Monday, 24 May 2010, at 8:52 p.m.
    • Error Rates and Edges
      Daniel Murphy -- Monday, 24 May 2010, at 9:56 p.m.
      • Error Rates and Edges
        Matt Cohn-Geier -- Monday, 24 May 2010, at 10:00 p.m.
        • Error Rates and Edges
          Daniel Murphy -- Monday, 24 May 2010, at 10:05 p.m.
          • Error Rates and Edges
            Matt Cohn-Geier -- Monday, 24 May 2010, at 10:08 p.m.
            • Error Rates and Edges
              Bob Koca -- Tuesday, 25 May 2010, at 3:06 p.m.
              • Error Rates and Edges
                Matt Cohn-Geier -- Tuesday, 25 May 2010, at 6:36 p.m.
                • Error Rates and Edges
                  eXtreme Gammon -- Tuesday, 25 May 2010, at 6:49 p.m.
                  • Error Rates and Edges
                    Matt Cohn-Geier -- Tuesday, 25 May 2010, at 6:54 p.m.
                    • Error Rates and Edges
                      neilkaz -- Tuesday, 25 May 2010, at 7:36 p.m.
                    • Error Rates and Edges
                      Daniel Murphy -- Tuesday, 25 May 2010, at 7:53 p.m.
                      • Error Rates and Edges
                        Matt Cohn-Geier -- Tuesday, 25 May 2010, at 8:01 p.m.
                        • Error Rates and Edges
                          Bill Finneran -- Tuesday, 25 May 2010, at 8:18 p.m.
                        • Error Rates and Edges
                          Daniel Murphy -- Wednesday, 26 May 2010, at 7:09 p.m.
                        • Error Rates and Edges
                          Christian Plenz -- Thursday, 27 May 2010, at 1:08 p.m.
                      • Error Rates and Edges
                        Timothy Chow -- Tuesday, 25 May 2010, at 8:11 p.m.
                        • Error Rates and Edges
                          Timothy Chow -- Wednesday, 26 May 2010, at 10:05 p.m.
                • Error Rates and Edges
                  Bob Koca -- Tuesday, 25 May 2010, at 7:13 p.m.
            • Error Rates and Edges
              Daniel Murphy -- Tuesday, 25 May 2010, at 4:11 p.m.
              • Error Rates and Edges
                Matt Cohn-Geier -- Tuesday, 25 May 2010, at 6:41 p.m.
  • Error Rates and Edges - - What is "average error"
    Bill Finneran -- Tuesday, 25 May 2010, at 2:23 p.m.