Team Win Values (Again)

As we mentioned this afternoon, adding up our individual win values for an entire team correlates very well with a team’s pythag win%, which should be expected since both are based on runs scored and runs allowed. However, while RS/RA does not account for the inning/score part of leverage, it does include the base/out portion of leverage. When looking at how many teams a run scores, a single with a runner on third will add to the total while a single with no one on will not. However, in our context neutral win values, all singles are treated the same – we do not account for the base/out context or the inning/score context.

Thus, you will still see some variation between team win values and team pythag wins. The Boston Red Sox are the best example of this from 2008. They won 95 games, had a pythag win expecation of 96 games, and had a team win value of 106 games. Their pythag matched their actual wins very well, but their win values had them as the best team in baseball, bar none.

Where does the difference lie? Performance in the base/out contexts. For instance, here’s how the Red Sox performed offensively in various base/out states, and their rank within that base/out split.

Overall OPS: .805 OPS, #2 in MLB
Bases Empty: .797 OPS, T-#1 in MLB
Men On Base: .814 OPS, #4 in MLB
Bases Loaded: .854 OPS, #7 in MLB

Only the Chicago Cubs could match the Red Sox in bases-empty hitting. Those two clubs started rallies more often than anyone else. However, Boston wasn’t as good at clearing the bases and capitalizing on their opportunities to score runs in bunches. The average MLB team had an OPS 18% higher with the bases loaded than with the bases empty – Boston’s OPS only rose 7% with the bases juiced.

This should actually be a bit scary for the rest of the AL East. What we’re saying, essentially, is that the Red Sox offense should have scored more than the 854 runs they actually compiled last year. On a context-neutral talent level, they performed even better than their 96 pythag wins would suggest. Their results suggest they were very good – the process behind those results suggest they were great.

Very few teams strayed as far from their win values in terms of their pythag wins as Boston. Almost all of a team’s RS/RA can be predicted by their context-neutral performance – the remainder is pretty marginal. That’s why win values correlate so well with pythag wins, and why they’re a pretty good estimator of team talent levels.