I ran into this right after I posted, and I usually don't like posting so close again, but I have to acknowledge it because it ranked Bochy 4th best in history. It was posted on Beyond The Boxscore by adowrowski, titled "Manager Wins Above Expectancy"
As the author noted, this is all preliminary noodling with a way of measuring the value of a manager, as the author also tried the above Pythagorean analysis, which I also happened to do many moons ago, but discarded because, one, the conclusions, while strong, was not statistically significant (I think the best it got to showing managers having a positive Pythaorean differential - that is, the team won more games than Pythagorean projected - was 60-70% level, not the 90% people want to see), and two, I realized that a manager could also influence the team positively to either produce more runs or to prevent more runs or both, and neither would show up in the Pythagorean difference (I explained much of this in my comment there).
I also noticed a connection between losing teams and negative differential, as well as winning teams and positive differentials, so that made me think that some flaw or something lacking in the Pythagorean was causing the differential, lessening the meaning and value of the metric.
I also noted that the methodology for WAR for Managers could be as flawed as the defensive system used to measure the individual players' WARs that was used to make this calculation for the manager. There is no concensus on what is the best defensive metric yet, where one system might have one player negative, the other might have him positive. I am sure on the whole, they probably match pretty well, but there will be those mis-matched stats for certain players, and I have no idea what the proportion of those are out there, so am hesitant to argue vociferously (as many do in the blogosphere) using WAR as the underpinning of my stance.
Still, the managers who show up well in the WAR analysis did match many of the managers in the Pythorean analysis, as well as the names of managers who popped up in my analysis of one-run games above .500 records for managers during Bochy's managerial career. So there is a lot of correlation happening across these various managerial metrics. And that, as the author noted, is the goal, to initially test out ideas and try different approaches to answering the question of which managers was talented in producing wins for their teams.
Giants Thoughts
Bochy was 19th in Pythagorean, but he could be as high as 14th if he has a good 2012 season in terms of that. As noted above, he was 4th in WAR (where WAR is totalled for the team and the differential against the team's actual wins is credited to the manager) and he should pass up Connie Mack for 3rd place after April or May, if he continues to do well. However, he might never catch up to 2nd, Mike Scioscia, who not only is still active, but is much ahead of Bochy. Though I suppose if he manages enough, Wilbert Robinson, who is currently #1, will become #2 when Scioscia passes him (he is close), and Bochy could pass him up given enough years as manager. At his current rate, it would take him roughly 5 seasons to pass up Uncle Robby (if I got Wilbert's nickname right).
As I noted in the comment area, the Giants were lucky, as the rumors was that Lou Piniella was the first guy approached for the managerial job, which he turned down as he wanted to be closer to his mom in Tampa (but really, Chicago vs. SF, not a huge difference, both are very far away), and he was among the worse managers on one of the lists. Since he turned it down, then the Giants went to Bochy.
Also, the author replied to my post, noting that Dusty Baker did make the Top 30 (so he must be 26-30 since the top 25 were listed), which matched my finding via Pythagorean and 1-run win differential, that he added value in that way.
Lastly, I would note that Felipe Alou was also among the top managers, 9th in Pythagorean (above Bochy) and 8th in WAR (below Bochy), but in either case, among the best all-time. The Giants actually improved in changing from Baker to Alou to Bochy.
No comments:
Post a Comment