A good major-league manager is, generally, one who wins. Managing, as with coaching in general, is probably the least empirically quantifiable part of baseball, because it’s about relationships. No cognitive scientist has stepped forward to propose Relationships Above Replacement. (Ahem, Russell?)
So most of us base our assessments of the best coaches in the league on who wins and who loses, because that’s something we assume they’re at least partly responsible for. There’s an award for this, too: Manager of the Year, which is given to a recipient in each league and has been awarded since 1983. The first two winners were Tony La Russa of the White Sox and Tommy Lasorda of the Dodgers. Deserving recipients, most would say.
Managing is one of the few bastions of baseball where you can say something about a guy being a great leader or a pillar of serenity in the clubhouse or something like that and you won’t immediately have somebody whack you on the head with a spreadsheet. But here at Baseball Prospectus, we’re all about taking the unblemished, naïve joy out of the game you love, so I put the Manager of the Year award through the numerical ringer.
I stuck to the most obvious numerical indicator of a manger’s quality—wins and losses—and looked at MOY results dating back to 2003, the year of PECOTA’s birth. I then looked at a team’s third-order winning percentage to get a rough idea of how much of their success came from intangible sources. (Or luck. Most likely luck.)
Ideally, a good manager would be able to bring his team to heights thought impossible based on its on-field talent, and since 2003, every MOY has done that. All 24 winners since then helped their respective teams beat their PECOTA projections.
Six of the 12 American League managers who won the award managed teams projected to finish below .500, and each guided his team to a percentage of .500 or better. Tony Peña with the 2003 Royals accounted for the largest bump, with a 117-point gap between his team’s actual winning percentage and what PECOTA projected. (That team really was a miracle.)
The same number of National League winners headed teams projected to finish below .500, and Joe Girardi of the Marlins in 2006 was the only one to fail to reach that mark, finishing with a final record of 78-84. Girardi still outshot the team’s projected percentage, however, by 43 points, which is seven wins. Girardi remains the only manager to ever win the award after finishing with a losing record.
How the winners did against their third-order winning percentages also reflects well on the voters. Ten of the 12 AL winners since 2003 beat their third-order percentage, with Joe Maddon’s 2008 and 2011 Rays falling short. Matt Williams in 2014 was the only NL MOY not to beat his third-order percentage. The managers this comparison reflects best on are Peña with the Royals in 2003, when his team's record beat its third-order winning percentage by 71 points, and Bob Melvin with the Diamondbacks in 2007, when his team finished 88 points ahead of its third-order winning percentage.
Only two of the 24 winners—Jack McKeon in 2003 and Ozzie Guillen in 2005—have won MOY the same year as winning a World Series, however. That’s the funny thing: The ballots for the award are, of course, submitted before the postseason, when a manager’s role is the largest it ever is. Every pitching change, pinch-hitting move and defensive substitution is proportionally more important than those made in the regular season, with closer games, higher stakes and unusual usage possibilities. Of the 24 managers I sampled, 20 made the playoffs. Half lost in the division series, and one lost a Wild Card game, which is to say they haven't, as a group, done any better than their non-MOY opposition. No matter: This is a regular-season award, fairly or not.
That fact makes the award easier to predict. Judging by qualities that are fairly consistent with recent MOYs — a record exceeding what PECOTA projected and an actual winning percentage that outstrips the third-order — it’s worth taking a crack at this year’s favorites.
The A.L. has two close competitors: Jeff Bannister of the Rangers and Ned Yost of the Royals. Projections were harsh on the Rangers, and rightfully so, with myriad injuries and doubt about production from guys like Prince Fielder. But Bannister has helped Texas outshoot that .481 projected percentage to a current mark of .537, which is also significantly better than the team’s .486 third-order percentage as of Tuesday night. One way of looking at this is to say that Bannister's Rangers didn't really outplay their projections–they scored and allowed more or less the number of runs they were supposed to. But they did outplay their performance, if you want to say it that way, and Bannister led them, if you want to view it that way.
PECOTA was even rougher on the defending league champ Royals, pegging them for a brutal mark of 72-90. The team’s current winning percentage is .580, nearly 140 points of winning percentage better. Yost has also helped the Royals to a winning percentage 50 points better than their third-order total.
Though a great many Cardinals fans would disagree with this, the obvious candidate in the NL is Mike Matheny. The preseason projection of a .543 winning percentage for the team put it as the best in the division, but Matheny has helped St. Louis far surpass that, already amassing 92 wins. The team’s .629 winning percentage is also solidly above its .554 third-order number. Place your Manager of the Year bets accordingly.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now
This is a decent way of looking at things, showing that the voters really do look for managers that appear to be having an effect. I'm not quite as sold on the PECOTA projection comparison; I think the third-order wins is more telling.
Still, it would be nice to see this expanded a bit, to see if in any years there were managers who performed even better by one or both measures. In other words, while the voters have been able to spot a good story, did they spot the "best" one?