I’ve been meaning to contest David’s contention that sabermetrics proves that managers don’t matter. I don’t actually think that sabermetrics has proven this, and I also don’t think it’s plausible. As it happens, I’ve just started to read Chris Jaffe’s Evaluating Baseball Managers, which (to the extent that what I’ve read so far is representative) is the most interesting work of sabermetrics I’ve seen in many years. I’ll leave my discussion of the book primarily to a more appropriate weekend slot, but since it’s come up recently I’d thought I’d make a couple initial points.
First of all, I just don’t think that a careful look at how teams develop and perform can be squared with the conclusion that managers don’t really have any impact. To take on obvious example, if you look at Earl Weaver’s teams, you’ll see some clear characteristics: a usually four-man rotation that is unusually effective, healthy, and that absorbs a huge number of innings; extensive use of of the bench, with attempts to skim cream by using matchups; 3)limited use of one-run strategies; and 4)related to the first two, using some one-way defensive players like Belanger, Blair and Dempsey without costing the team ability to score enough runs to win. You’re telling me that if someone like Don Zimmer or John McNamara — a completely conventional manager who doesn’t use the bench and has no particular ability to get good performance out of pitchers — was managing the Orioles in those years, they would have had the same success? Is Earl Weaver in the Hall of Fame just because he managed for a long time “without blowing his brains out?” I think that’s ridiculous. Or to borrow one of Bill James’ favorite example, there’s no way that a good manager — one that wouldn’t have overworked his top two starters, knew that Adock was a vastly better player than Frank Torre, etc. — wouldn’t have won the pennant with the ’59 Braves. It’s just one variable, and obviously no manager can win without talent and a mediocre manager can win in the right circumstances — but it matters.
So I don’t buy the idea that Lou Piniella is just an innocent bystander in 1998, or that management played no role in the fact that the Mariners finished more than 10 games behind a team without an obvious talent edge going into the season. One of the biggest differences between the teams is that Oates got a lot of decent-to-excellent performances out of relievers who (Wetteland aside) had very modest credentials, while the Mariners’ bullpen was a complete catastrophe. As I said, this isn’t the only or even the most important factor; the inability of the Mariners’ ownership and management to sign Johnson to a new contract and some bad trades were more important factors. But there was no reason that team should have finished more than 10 games out; the Rangers weren’t exactly solid top-to-bottom either. I hadn’t read his profile before using the example, but Jaffe shows that in Oates’ case, this was no fluke — in both Baltimore and Texas Oates consistently had overachieving bullpens. Piniella, on the other hand, just as one would expect consistently overachieved on offense but had an erratic (and, on balance, negative) effect on his pitching staffs. There are managers who have the ability to repeatedly construct a functional bullpen out of modest materials (Bobby Cox, Gene Mauch and — more about him later — Cito Gaston would be other good examples); Piniella just isn’t one of them. And in some contexts that can make a big difference. This doesn’t mean that Piniella wasn’t a good manager, but it is one reason that his overall performance isn’t Cooperstown worthy.