As my regular readers hopefully know, I wrote a book published at the beginning of the year, Evaluating Baseball’s Managers, 1876-2008. (Order it and find out why Rob Neyer says: “Way back in the 20th century, Bill James wrote the first essential book about baseball managers. Chris Jaffe has just written the second.”)
Anyhow, due to the time lag that exists between submitting my manuscript and its publication, an entire season played out (hence why the book covers managers only through 2008).
But nerd that I am, just because I finished my book doesn’t mean I finished my research. I took some of the main tools and ways of evaluating managers I used in the book and updated it for another year. It’s too late to add into the book, but hey – I got a column here, don’t I?
The main thing I used to evaluate managers is the Tendencies Database, which I’ve updated for 2009. First off, don’t expect any massive EUREKA moments as it operates under the principle of sample size: namely, you learn more about a manager by looking over his career than just by looking at one season. Each season changes the overall career, but it’s just part of the career.
The Tendencies Database
I presented this database in an excerpt here at THT, but if you don’t want to read all that, here’s a brief overview of it:
You can take any stat – sacrifice hits, park-adjusted bullpen ERA, doubles, etc – and run it through the Tendencies Database. It’s designed to figure out what the average rank for a manager’s team was in relationship to the size of the league. By that last part I mean there’s a big difference from ranking third in an eight-team league and third in a 16-team league.
An average score is 1.000. Lower means better, and higher worse. That might sound odd, but realize we’re looking at ranking. A team with the best rank in homers hit has the lowest rank (first) despite hitting the most homers. A rank of 0.500 is very low, and 1.500 very high.
Whenever possible, I adjust a stat for context. Thus if I want to look at sacrifice hits, I’ll divide SH by opportunity to call the play. (After all, a team with a .300 OBP and 100 SH is more interested in that play that a team with a .350 OBP and 105 SH.)
The Tendencies Database is only applied to managers who lasted at least a decade as a team’s primary manager. You want a decent enough sample size to feel like you’re really looking at the manager, not just the collection of talent at hand. To be sure, talent on hand always strongly influences a manager’s ranking, but it’s hard to rank at an extreme end of any category without a manager prioritizing (or de-emphasizing) that part of the game.
When I wrote the book, only 77 managers qualified for the Tendencies Database. Last year a pair of skippers hit the decade mark: Mike Scioscia and Terry Francona. Let’s see what we can learn about them.
The following isn’t every team he appears on in the Tendencies Database leaderboards. I’m tempted to write it, but it would be a bit too redundant of what I wrote about him in my book.
Oftentimes when Scioscia appears on the leaderboard the results aren’t surprising, but it’s still nice to get a sense of where he stands in historic perspective.
For example, anyone who follows the Angels likely knows Scioscia’s offenses focus largely on getting the ball in play. His hitters aren’t known for drawing too many walks, and they are damn difficult to fan. In fact, his teams are historically ball-in-play-centric, as the following lists from the Tendencies Database show:
Fewest offensive BBs Fewest offensive Ks Danny Murtaugh 1.665 Bill Virdon 0.402 Tom Kelly 1.484 Felipe Alou 0.476 Mike Scioscia 1.453 Mike Scioscia 0.520 Don Zimmer 1.407 Jimmy Dykes 0.592 Felipe Alou 1.400 Phil Garner 0.635
(If you’re curious, the rate is determined by the equation (BB-IW)/(AB+BB+SF). For strikeouts, it’s K/(AB-H). Few if any managers have been as heavily focused on getting the ball in play as Scioscia.
One offensive philosophy popular in sabermetric circles is the Three True Outcomes approach in which hitters rack up impressive walk, home run, and strikeout totals. While that approach can work, Scioscia clearly has another approach in mind.
Instead, Scioscia’s teams focus on getting balls in play. To that end, they are historically great at slapping out base hits, as the Tendencies Database’s singles query (1B/AB) makes clear:
Most singles-happy offenses Mike Scioscia 0.453 RedSchoendienst 0.487 Tom Kelly 0.524 Hughie Jennings 0.540 John McGraw 0.543
Scioscia makes this offensive approach work. In his 10 years on the job, the Angels have belted out more singles than any team in MLB four times, and were runner-up a fifth time. Not bad for someone managing in an era of 30 teams.
Other things the Tendencies Database told me about Scioscia were a bit more surprising. For example, there’s the matter of how he likes to rest (or not rest) his starting positions players.
To understand the charters below, I need to take a step back for a second. One of prouder queries with the Tendencies Database was percentage of PA each team allocated to its starting position players. I figured that out for every team, and then ranked them.
I also got a couple offshoots based on tallying that info: which managers got the most PA from their infielders, or outfielders, or catchers, or up-the-middle guys. There is a slight problem with those queries: batting order placement. If one manager puts his infielders at the top of the lineup and another puts them at the bottom, it’ll look like the former is leaning on them more than the latter, even if the two teams start the infielders as often as each other.
While that’s something to keep an eye on, it’s not as serious an issue as it sounds. It also provides some interesting evidence for who Scioscia likes to rest (his catchers), or not rest (his outfielders), as the following results from the Tendencies Database demonstrate:
Most rested catchers Most used starting outfielders Frankie Frisch 1.348 Mike Scioscia 0.547 WilbertRobinson 1.298 Jimy Williams 0.553 Jimmy McAleer 1.273 Danny Murtaugh 0.581 Mike Hargrove 1.236 Bill Terry 0.622 Buck Showalter 1.208 Harry Wright 0.707 Mike Scioscia 1.200
Fine, I cheated a bit with catchers – Scioscia is ranked sixth. Close enough.
More importantly, neither ranking can be simply dismissed as batting order potluck. In his time with the Angels, no catcher has ever appeared in more than 130 games. The other 29 teams have done it 94 times. Sure Bengie Molina ain’t great, but neither are a lot of guys who have worked well over 130 games in a season.
I wonder how much Scioscia’s own experiences as a player colors how he treats his catchers. He was an All-Star in 1989 and 1990, and in his last season in 1992. Those miles caught up on him mighty quickly.
With outfielders, Scioscia’s men started 150 or more games 13 times. Only one other team tops 11 in the last decade (Seattle, with 17). Talent certainly plays a role as he has quality players to put out there, but there’s also some decision making by Scioscia going on. He let Darin Erstad play 157 games in 2001 despite an OPS+ of 82 – and then next year gave him 150 games while he posted an 86 OPS+.
Scioscia signaling backup Jeff Mathis, in one of the 89 games he played in last year.
On the pitching side, Scioscia’s teams are known for their bullpens, which a pair of queries in the Tendencies Database makes clear: bullpen ERA+ and saves per win. The former is straightforward: which bullpen did the best job, adjusted for park. The latter looks at how much a manager likes to have his relievers close out the victory. I use saves-per-win instead of just saves because the latter unduly benefits good teams. It isn’t perfect, but it works well enough. The results show Scioscia’s bullpens are impressive by any standard:
Best bullpen ERA+ Most saves per win Mike Scioscia 0.587 Mike Scioscia 0.507 John McGraw 0.611 Burt Shotton 0.578 Jimy Williams 0.641 John McGraw 0.600 Danny Murtaugh 0.685 Bruce Bochy 0.612 Johnny Oates 0.693 Felipe Alou 0.673
(The ERA+ results are very different from my book because so much more team split info is available now compared to when I assembled it.) More importantly, Scioscia tops both lists. He’s gotten the best bullpens of any manager out there (well, at least among those with 10 years team split info – let’s see what happens in two years when Ron Gardenhire joins the mix), and relies as heavily as anyone on the save. I suppose it isn’t a surprise the individual single-season save record was broken on Scioscia’s watch.
There are some other interesting results for Scioscia: he comes in fifth best all-time in willingness to give his hitters the platoon advantage. In terms of overall platooning (combining offensive and pitching platoon scores – one nice thing about the Tendencies Database is that it’s easy to add things together because they’re all centered at one with the same high/low scores), Scioscia has the second best platoon-advantage fixation of all time, behind only Gene Mauch. But that’s enough about Scioscia for right now.
Francona is a bit harder to separate from his teams than Scioscia. Both the Red Sox and Angels have their own distinctive styles of play, but speaking as someone more than 1,000 miles away from either town, the Red Sox front office gets more attention for their club’s predilections than the Angels do.
One key trend of his teams is a lack of interest in the sacrifice hit, as the following Tendencies Database result – using the formula SH/(H+BB+HB-HR-2B-3B) to adjust for opportunity (the overwhelming majority of SH come with a runner of first) – shows:
Least likely to call a SH Buck Showalter 1.606 Bruce Bcohy 1.592 Tom Kelly 1.547 Terry Francona 1.463 Jimmy McAleer 1.455
Random comment: prior to 2009, Bochy was the most bunt-phobic manager in history. He’s doing it more in San Fran than he ever did with the Padres.
Getting back to Francona, even in Philly, he usually stayed away from the sacrifice bunt. By this equation, the 1998 team was 12th in the league in SH, and 11th in 1999. With Boston, Francona’s teams have been last or next-to-last in SH propensity every year except 2006, when they were third-to-last.
Intentional walks are another trait his teams stay away from, as the following chart (based on the equation IW/IP) from the Tendencies Database makes clear:
Least likest to issue IW Davey Johnson 1.736 Casey Stengel 1.602 Tom Kelly 1.556 Terry Francona 1.537 Charlie Dressen 1.527
In this case, Francona appears as opposed to the play in Philly as he now is in Boston. Well, almost. He was actually middle-of-the-pack with it in his rookie season, but immediately turned away, ranking in the bottom five in IW per inning every year since then.
On defense, his teams are generally pretty bad at turning the double play. The Tendencies Database handles this with the formula: DP/(H-2B-3B-HR+BB+HBP-SH-SB-CS) which is essentially double plays divided by chances to pull them off. Based on that, the following managers have had the worst middle-infield combinations:
Worst at turning double plays Hughie Jennings 1.540 WilbertRobinson 1.439 Alvin Dark 1.357 Terry Francona 1.300 Felipe Alou 1.276
Francona’s clubs only came in last once (last year’s, actually), but are routinely near the bottom. In Philly they never ranked higher than 11th place. Boston actually topped the league in this 2006 with Mark Loretta and Alex Gonzalez, but neither started with the club in any other season.
It could be that Francona has run a bunch of highly fly ball-prone staffs, but I’m skeptical of that over two teams and 10 years. Again, player talent matters, but there’s something else going on. For example, some managers – most notably Casey Stengel and Mauch – also made the double play work for them no matter where they were or what talent they inherited. Some managers fixate on it, others don’t.
On the pitching side, Francona’s relievers typically have among the shortest outings of anyone. I ran bullpen innings divided by relief appearances (IP/RA) through the Tendencies Database, and the following guys came out with the shortest relief outings, relative to their peers:
Fewest IP/RA RedSchoendienst 1.719 Bobby Cox 1.615 Whitey Herzog 1.557 Terry Francona 1.507 Tony LaRussa 1.353
There are only six times in AL history a bullpen averaged less than an inning per appearance: Francona managed two of them.
A fairly common sight: Francona taking out one reliever for another.
When Francona gives a reliever a quick hook, it’s often to give his team the platoon advantage. Another Tendencies Database inquiry makes that clear, as the chart below looks at what percentage of batters-faced saw the team’s pitching staff possess the platoon edge over the opposing batters:
Most frequent pitcher platoon advantage Gene Mauch 0.511 Mike Hargrove 0.516 Terry Francona 0.545 Tony LaRussa 0.593 Bill Vidon 0.615
Francona came in second in the league in this category in each of his first pair of seasons in Philly. He led it once in Boston, and finished third two other times with his current club.
Going by the Tendencies Database, Francona also has hitters who draw walks, older pitchers, and teams who have trouble keeping an even pace all year round.
There are other things I can say about managers into 2009 based on the research from my book, but this column has gone on long enough. That will have to wait for another piece.