PBR - Beginning with this weekend's series against the Braves, the Review will publish Sabermetric Scorecards prior to every series. We did something similar to this last year and had some success, so it's time to bring the feature back.

As I noted in a post earlier this week, I've done extensive statistical research over the past year and found the combination of OPS (on base plus slugging), WHIP (walks plus hits divided by innings pitched) and RDiff (run differential) to strongly correlate to winning percentage. 

My data-set included every club since 1900 and the correlation between winning percentage and these three categories was 92%, so it's not some untested theory - these three statistics are vital to winning.

The graph below shows the raw figures for each club in each of the three categories.  To compare the clubs (within the National League) I calculated how many standard deviations each club was from the average in the three categories. I then added the three standard deviations together for the Total Stan. D (which you see below). 

Using the regression coefficients from my earlier research, I also place the individual standard deviations into the following formula: (0.4999+(0.0052*OPS s.d.)+(0.0724* RDiff s.d.)+(0.0061*WHIP s.d.)). This formula produces the Estimated Winning Percentage you see below. You can then compare the Estimated Winning Percentage to the Actual Winning Percentage and determine how much luck - good or bad - has impacted a particular team.



Post a Comment

Previous Post Next Post