An idea is to compare batters with the top batters from their era. To do this here, I took the top 30 batting averages for each season and took each season's value as the average across the 5 seasons centered there. For example, the 1950 value here is the top 30 batting averages for 1948, 1949, 1950, 1951, and 1952. Here is that plotted with +/- 1 standard deviation:
That gives us a baseline to compare against in each season. Effectively, that value tells you how well the best in the league performed, so comparing with it should give you some correction for batting difficulty changing over time. You can clearly see that batting averages among top players were higher in the 1920's - 1940's than they are now. Where do the best players in each year sit on that plot? Here's the same thing with the season leaders plotted also:
You can see some clear spikes. 1941, 1977, and 1980 really jump out. How can you normalize those values?
A way to do it is to see how much of an outlier a given performance was. Simply subtracting average performance from actual performance and dividing by standard deviation gives you number of standard deviations from average. This is a reasonable metric for 'how much better than the best of your era were you?' Doing that, I get the following scores:
| Player | Year | Stdevs Above Avg |
|---|---|---|
| George Brett | 1980 | 5.8 |
| Rod Carew | 1977 | 5.4 |
| Ted Williams | 1957 | 4.7 |
| Ted Williams | 1941 | 4.6 |
| Tony Gwynn | 1994 | 4.3 |
| Wade Boggs | 1983 | 4.1 |
| Rod Carew | 1974 | 3.9 |
| Ichiro Suzuki | 2004 | 3.8 |
| Stan Musial | 1948 | 3.6 |
| Joe Torre | 1971 | 3.6 |
Here is the plot from earlier with the top-3 called out:
From this metric then, George Brett's 1980 season was actually the most impressive in terms of batting average when adjusted for batting difficulty of the era. Interestingly, Williams' 1941 season was actually less impressive than his 1957 one.
0 comments:
Post a Comment