March 9, 2012
The Stats Go Marching In
The Hidden Helpers of the Pitching Staff
The best pitcher handlers since 1948
I won’t describe my methods again here, since you can look at my previous article if you need a refresher. Suffice it to say that a With-Or-Without-You approach has been used here, and that the effect of the pitcher, batter, ballpark, and defense has been removed in order to evaluate that of the catcher.
For this analysis, I have used moving four-year windows. In other words, when I wanted to rate Carlton Fisk’s 1975 season, I calculated four separate values: one derived from 1972-1975 data, the second from 1973-1976, the third from 1974-1977, and finally one from 1975-1978. The final value I assigned to Pudge in 1975 is the average of the four. In this way, I tried to balance between using as many seasons as possible in order to have a wider comparison group, while not having the numbers influenced by the aging of pitchers and batters.
The first thing that will probably strike you is Mike Piazza’s ranking. Piazza has always been considered a poor defensive catcher because of his inability to throw out basestealers. However, he fared well at avoiding passed balls and wild pitches (as Tom Tango showed in With Or Without You in The Hardball Times Annual 2008) and now emerges as one of the best ever at handling the pitching staff.
Carlton Fisk is the only member of the Hall of Fame to appear on the list (until Piazza gets in).
Johnny Bench was average (two runs prevented in over 60,000 PAs); Yogi Berra, whose first couple of years are not included in the analysis, improved his batterymates by 57 runs throughout his career (roughly five runs per 5,000 PAs); Roy Campanella is not too far out of the top 10, with 123 runs prevented (15 per 5,000 PAs); Gary Carter was also very good (94 runs for his career, six-and-a-half per 5,000 PAs). The remaining receivers with plaques in Cooperstown played before 1948, so we currently don’t have a way to rate them.
Note: 5,000 PAs is roughly the number of PAs for which a catcher is behind the plate in 130 games and thus is used here as the proxy for a season of play.
On a rate basis, considering only catchers with at least 15,000 PAs, Jose Molina easily tops the competition with 38 runs saved for every 5,000 PAs; Lenny Webster and Mike DiFelice trail him from a considerable distance at 30 runs.
The Dodgers of the ’80s and ’90s are an interesting case. Mike Scioscia became the starting catcher for LA in 1981. As you can see in the above table, he was one of the best pitcher handlers ever. When he left after the 1992 season, Piazza, also among the top 10, took over the role. Finally, when Piazza left the team during the 1998 season, Paul Lo Duca was already on the Dodgers roster, though he didn’t become the starter until a couple of seasons later. Lo Duca is estimated to have saved 107 runs in his career, an average of 16 per 5,000 PAs.
Thus, except for the brief reigns of Charles Johnson (one of the worst ever in my estimation), who served for part of the 1998 season, and Todd Hundley (who is rated as slightly above average), the Dodgers had 20 years of excellent field generals behind the plate.
Handling a new staff
I divided catchers into three buckets: one for players debuting in the big leagues, one for those in their first year with a new team, and one for everyone else. After accounting for the players represented in each bucket and their playing time, I estimate that rookie catchers are roughly four runs worse and catchers who have changed uniforms are about three runs worse than catchers staying with their teams. (Numbers are again per 5,000 PAs)
Looking at service with the team as a continuous variable, I found that catchers improved at handling their pitchers by 0.70 runs per 5,000 PAs every year they spent with the team.
I used the so-called “delta method” to compare players in adjacent seasons. The numbers of every player in every season are compared with the numbers of the same player in the previous season. Then the average (weighted for the players’ PAs) of the improvements and declines is calculated.
The chart below shows how the handling-of-pitchers skill varies throughout the catchers’ careers.
The thicker smoothed line shows that there’s not much of an aging factor. The decline is hardly detectable.
Calling the shots from the bench
How much of the value we assigned to catchers (both today and a couple of weeks ago) should be credited to the manager instead?
I analyzed the catcher/manager combinations in the Retrosheet database by performing another With-Or-Without-You analysis. I calculated how much a catcher improved his pitching staff with a particular manager on his side and compared that value to what he accomplished without that manager.
Below is the list of the 10 managers who most helped their batteries.
The first thing to note is that managers appear to have a smaller impact than catchers. Where the latter can reach as much as 30 runs saved per 5,000 PAs, Bobby Cox, who is head and shoulders above his peers, is credited with just a couple of runs per the same amount of PAs.
Before commenting on the names appearing on the list, let’s also look at the bottom 10.
Among managers with shorter careers (who hardly had a chance to appear on either list), Carl Ermer, Herman Franks, Harry Crafts, and Cito Gaston are rated very high, while Charlie Dressen, Tony Muser, Wes Westrum, and Don Gutteridge are considered rather poor by the model.
Note that older managers are rated for only the seasons covered by Retrosheet. Casey Stengel’s numbers, for example, come out of his New York stints (Yankees and Mets), while his Brooklyn and Boston (Braves) years are not considered.
Now, do the rankings I just showed make sense? Seeing Bobby Cox at the top is reassuring, as is finding Dusty Baker, who has a reputation for being bad at managing pitchers, near the bottom.
As a check for my numbers, I went to the best reference book on managers, Chris Jaffe’s Evaluating Baseball Managers. Using completely different methods than mine, Chris also ranked managers on how much they improved their pitchers. His list includes skippers of older times and is headed by Joe McCarthy, who is not included in this analysis. However, it features Bobby Cox right behind him, and the following all among the top 20: Lopez (third), Alston (fourth), Richards (10th), LaSorda (13th), Williams (14th) and McKeon (20th). The agreement between Jaffe’s ratings and mine is less perfect at the bottom. However, he also has Leyland, Kelly, and Francona rated very low.
Looking at other individual values, our lists often agree, but it must be noted that there are cases where the results are quite dissimilar, including Lou Piniella (ranked high here and low in Jaffe’s book) and Phil Gardner (who had the opposite fate). Overall, I consider my numbers on managers at least adequate.
As I said, the manager effect on the course of a season is rather small. However, the ratings of catchers like Jason Varitek, who was long linked with Terry Francona, would surely benefit from a revision.
The Mazzone Effect
Since my analysis removes the value of pitchers (and hitters as well) from the equation, I saw an opportunity to reexamine the issue. I repeated my WOWY catcher/manager analysis, but this time I used different identifiers for Bobby Cox, one for years when he paired with Mazzone, and one for the rest of his career. Luckily, the two groups have a comparable number of PAs, 81,000 for Cox alone and 88,000 for the Cox-Mazzone tandem.
Here are the numbers:
Looking at the above lines, it seems that Mazzone’s supporters have a case. Unfortunately, his numbers in Baltimore are both small in terms of sample size (3,361 PAs) and underwhelming in terms of performance (0.1 runs per 5,000 PAs).
Ultimately, we’re left wondering whether some of the Braves’ success was attributable to Cox, Mazzone, or the combination of the two men.
Bringing the playing experience to the bench
The Pearson’s coefficient is very low, at 0.11: there is no evidence implying that catchers who are good at handling pitchers retain the skill when moving to the bench.
I also looked at how managers fared at handling pitchers based on their positions during their playing days. Note that the primary position is calculated based on only games played in the big leagues. (I repeated the analysis adding minor-league playing information where available, and the ranking remained the same).
It makes sense that former pitchers and catchers come out as the best in this department. On the other hand, it looks like teams employ too many former infielders and managers without playing experience.
Obviously, this is a limited point of view, as there are many other areas in which guys like Billy Martin and Earl Weaver excelled. However, I’d be more careful in my choice of a pitching coach should I appoint someone without pitching or catching experience as the manager of a ballclub.
Analysis similar to what you can read here at Baseball Prospectus and at other sabermetrically-inclined sites is also going on behind the curtains of MLB front offices, as the signing of Jose Molina by one of the most analytically-oriented clubs indicates.