Managerial differentialsby Chris Jaffe
August 17, 2009
As some readers may have noticed, for the last two years or so, my "about the author" blurb at the bottom of my columns notes that I've been contracted to write a book for McFarland titled, Evaluating Baseball's Managers, which should come out later this year.
I mention this because this week's column is an offshoot of that project. By "offshoot" I mean it's something I didn't think to do until after submitting the manuscript back in March. Ah well.
Specifically, it looks at something I call differentials, which I do use in the book, but not the ones covered here. I actually briefly discussed differentials in a column earlier this summer on Clint Hurdle. The idea behind them is simple: take a stat that appears for hitters and pitchers, such as walks or home runs or strikeouts, and see how a manager's teams do in that given category over time.
For example, if you looked at walks surrendered and taken by the teams of Earl Weaver, it turns out his hitters earned 9,868 free passes while his pitchers gave up 8,139. That amounts to a differential of +1,729, the fourth-best total in history. (Note: for partial seasons, I include campaigns the manager lasted at least half the season, but not otherwise, so this walk differential includes Weaver's 1968 season. It's not perfect, but it does a good enough job. I'm a general gist guy than an absolute perfection person.)
Please note that any differential, whether it is walks or homers or whatever, primarily informs you about the players. There is no overt managerial impact on these things. However, managers have some influence over almost all these stats. They hire the coaches, determine who plays and what criteria exists for playing time, instill a philosophy of baseball with the club, and often have some pull with the front office as to who was on the team.
Don't get me wrong, the overwhelming majority of times a differential tells you little to nothing about a manager, but it's mighty darn hard to be the fourth-most extreme at any tendency without learning something about the manager. In this case, the differential perfectly fits what we know of Earl Weaver, who greatly prized getting on base.
Anyhow, I figured differentials for all managers in stats I thought would illuminate managers who appeared at either extreme end of the results: home runs, walks, strikeouts, and double plays. After submitting the manuscript, I decided to run a few more, just because I thought it would be fun rather than believing they would inform. Well, once I had the results, I wished I'd done the work before it was too late.
Negative HBP differentials
One differential I belatedly looked at was hit by pitch. According to it, the following managers had the lowest differentials (meaning their teams gave up far more HBP than they accumulated):
Manager Dif. Hughie Jennings -232 HBP Lou Piniella -164 HBP Jim Leyland -150 HBP Bruce Bochy -145 HBP Branch Rickey -139 HBP
(Note: this only goes through 2008, because that's as far as the book goes. If you're curious, through August 13, the numbers for active managers are: Piniella -170, Leyland -128, and Bochy -132. All are still in the top five, despite some declines by Leyland and Bochy.)
Hughie Jennings is about 40 percent ahead of second place, and only seven managers in baseball history had a differential half of his. (The others are Connie Mack, Frank Robinson, and Casey Stengel.) Rarely will the leader in any category have such a substantial edge.
Jennings managed the Tigers from 1907 to 1920, and they compiled a negative HBP differential in every campaign. That is the only time in MLB history a team went 14 straight years with a negative HBP differential. Next best is 12 by the 1962-73 Red Sox.
That's nice, but it could just be a coincidence. In fact, that 14-year streak of negative HBP differentials when Jennings ran Detroit was just part of a longer 19-year stretch spanning before and after he took over. Also, I looked at some of his main hitters and pitchers and found out that their HBP rates rarely changed much upon arrival in or departure from Detroit.
That isn't as important as it sounds, though. When Jennings managed, the front office hadn't fully evolved and skippers generally decided who the players would be. He might have intentionally sought out guys who fit his HBP notions.
In fact, I feel strongly that this particular differential not only sheds light on Jennings as manager, but also tells us quite a bit about what Jennings thought of his own playing career. You see, the striking thing about the leaderboard for negative differentials isn't the gap between first and second place. Nay, it's the identity of the man atop it.
The reflections of an aging Hughie Jennings
Ever heard that Craig Biggio owns MLB's all-time HBP record among players? Not quite. He owns the modern (re: since 1900) record. Overall, his 285 HBP place him two behind the real leader, Hughie Jennings.
It's amazing that Jennings is first in any career category because while he's a Hall of Fame player, he had one of the shortest careers of any HoFer, at 1,285 games. Yet he still found time to get plunked 287 times. Not surprisingly, Jennings is also the single-season record holder, as he took 51 plunks in 1896, one ahead of Ron Hunt's post-1900 record. Though MLB history records only four times a player took more than 40 HBP in a season, Jennings personally did it three times—and topped 45 in each of those seasons. Overall, Jennings took a plunk once every 20 times up. OWWW!
As a player Jennings clearly believed in taking one for the team, but as a manager his teams took the diametrically opposite approach. Weird. Why were his teams so unlike himself?
Allow me to answer this by first asking a different question: if Jennings played only 1,285 games, why is he in Cooperstown as a player? Well, he's an odd Veterans Committee pick. Such bizarre decisions have occurred frequently.
That's not quite fair to Jennings, though. Despite his short career, Bill James' New Historical Abstract calls him the 18th-best shortstop of all time. James isn't alone in this unexpectedly high appraisal, either. Baseball Think Factory contains a section called the Hall of Merit which attempted to construct a parallel version to Cooperstown, only without the blown picks. They also elected Jennings. Even if Cooperstown's VC picked Jennings for the wrong reasons (namely cronyism), he is a credible pick, despite his career length.
Jennings was a peak monster who was arguably the game's best player for his too-brief days of glory. According to advanced metrics like Win Shares, in his prime Jennings was one of baseball's finest defensive forces. Meanwhile, his bat was nearly as good. Jennings was the Sandy Koufax of shortstops.
Riddle me this, dear reader—if Jennings was such a Mr. Everything in his prime, how come he played fewer games than Harold Reynolds?
Simple: Jennings took too many HBP to remain effective. I lack medical reports from 1899, but look at the record. From 1896 to 1898, when Jennnings was a powerhouse, pitchers drilled him 143 times. Then he played only 69 games in 1899, and it was swiftly downhill from there.
Part of Jennings' offensive value lay in those OBP-boosting HBP, but they ruined him. A decade later, he became manager and his teams never seem to have many HBPs while his pitchers frequently drilled opponents. This strikes me as a bit of self-reflection on Jennings' part. I think he decided his approach was penny-wise but pound-foolish. To help his team score an extra 10-20 runs a year he paid the price of a career shortened by perhaps 1,000 games.
If the HBP was an offensive weapon, its main casualty was still the man getting hit. As manager, Jennings didn't want his players to make that tradeoff. None of his players ever got hit more than 12 times in a season. In comparison, from 1907 to 1920, when he managed players on other teams received at least 13 bruises 63 times.
Positive HBP differentials
Let's look at the other end, the managers with the highest HBP differential (through 2008 for active managers):
Manager Dif. John McGraw +708 HBP Al Lopez +270 HBP Joe McCarthy +246 HBP Earl Weaver +203 HBP Bobby Cox +190 HBP
Quite a lead Mr. McGraw has. Yeah, he lasted a long time and all, but his score is almost as large as the three next highest combined.
Fun fact: John McGraw and Hughie Jennings were old teammates in the 1890s. Their 1898 Baltimore team still owns the single-season record with a +99 HBP differential. In comparison, since 1900 the highest is +60 by the 1998 Pirates.
Back in the 1890s, when hurlers hit Jennings 40-50 times a year, McGraw's ribs felt sore 10-20 times per season. Though not as extreme as Jennings, McGraw clearly saw the value in getting hit by pitch. Based on the teams he managed, McGraw had no qualms about his past.
The Giants underwent a HBP revolution when McGraw took charge. From 1898 to 1902, New York scored a negative HBP differential every year, averaging nearly -30 per season. That included a -55 HBP differential in 1899, which is still the lowest ever (though the 1996 Angels tied it). To be fair, McGraw managed the last 65 games in 1902, but he hadn't yet had a chance to establish his print on the club.
When he did, the Giants vaulted from an HBP differential of -18 in 1902 to one of +52 in 1903. That is the most substantial change in baseball history of any team's HBP differential from one year to the next. Also, 1903 proved to be the first of 27 consecutive seasons the Giants received more plunkings than they doled out, easily the record.
As a player, McGraw's best feature was getting on base, and he wanted teams set up like that. The above charts shows that would much rather take the base by HBP than give it up. That displays a concern for the advantages of getting on base.
Another way to look at it is hit differential (which I also did after sending out the manuscript). These managers had the best hit differentials:
Manager Dif. John McGraw +3,113 hits Joe McCarthy +2,527 hits Jim Mutrie +1,883 hits Frank Selee +1,879 hits Walter Alston +1,650 hits
(Note: as of mid-August 2009, Bobby Cox, who began the year with a hit differential of +1,623, has passed Alston for fifth place.)
Sure, this tells you about team quality of players McGraw possessed, but it also says something about how McGraw constructed his squads: get on base more than the other guy and you'll win. To that end, he also had the best walk differential (which is in the book and thus won't be repeated here).
When walk, hit, and HBP differentials are summed, only Joe McCarthy and Joe Torre are within half of McGraw's career mark, and Torre barely so. On-base percentage didn't exist as a stat back then, but John McGraw believed in it as a philosophy more than any other manager.
Finally, here are the worst hit differentials:
Manager Dif. Frank Robinson -1170 hits Billy Barnie -1150 hits Zach Taylor -954 hits Gene Mauch -949 hits Art Howe -924 hits
I don't think this tells you much about the managers in this case, just the quality with which they had to contend. That said, I like this for one reason: it illustrates how incredibly untalented Frank Robinson's clubs were. His fate was to be hired continually by teams undergoing their worst stretches in decades.
He ran the Cursed by Colavito Indians, helmed the Johnnie DiSaster Giants, and managed the Montreal Seligs. Do you realize his best opportunity may have been in Baltimore, who was in the midst of the longest losing streak in AL history when he took over? Since WWII, I don't think any manager has found himself in such dire situations with so many clubs.
A final bit of navel-gazing
I should note the book hasn't come out yet and technically I could add everything here to it, but that's not nearly as good an idea as it sounds. One thing I've learned in writing both this book and my dissertation, is that as long as you are examining a subject, new thoughts will crop into your head regardless of how much time you've previously thought about it. Ideas are infinite. From a purely philosophical point of view, there's no such thing as finished.
That's the problem. If I submit these additions, it'll likely gum up the works (it apparently takes several months for a manuscript to become a book). By the time the Jennings correction is made, I might have another tweak. Then another one. And so on. I can either accept that the book will not contain my complete thoughts or hold off publication until after I did. I'd rather live to see it come out.
Much of this would've just amplified points made in the book already. The Hughie Jennings stuff is what I really wish I had a time machine for. The good news is my THT column allows me to inflict this info on the general public anyway.
References and Resources
Quick note: until well into the 20th century, league-wide HBP differentials rarely added up perfectly. That said, they were usually off by only a small bit. For example, from 1903-31, when John McGraw managed the Giants, the NL was off by an overall total of -47 HBP. I suppose McGraw's HBP differential above actually underestimates him by a half-dozen then. Unreal.
Retrosheet served as the basis for the differentials.
Baseball-Reference's Play Index helped with some information as well.
The New Historical Baseball Abstract and Win Shares by Bill James (and Jim Hentzler co-authored the latter) helped provide some reference of Jennings' ability. So did BTF's Hall of Merit
History instructor by day, statnerd by night, Chris Jaffe leads one of the most exciting double lives imaginable; with the exception of every other double life possible to imagine. Despite his lack of comic-book-hero-worthiness, Chris enjoys farting around with this stuff. His new book, Evaluating Baseball's Managers is available for order. Chris welcomes responses to his articles via e-mail. Oh, and now he's on twitter.