Managers and leveraging (SP Lev, part 7)by Chris Jaffe
May 29, 2007
Welcome back to The Hardball Times' ongoing series examining starting pitcher leveraging. If you know the gist of these suckers, skip this paragraph and the next. For the uninitiated, starting pitcher leverage refers to the once common but now extinct practice of a team intentionally using one of its pitchers disproportionately against particular opposing teams. It could be an ace starting all the time against the best opposing teams, or southpaws starting against the most left-leaning offenses.
For this study, I figured out that leveraging existed in the earliest days of baseball up to the 1960s, and I looked at the usage patterns for virtually every pitcher worth looking at. I ended up determining the leveraging for two-thirds of all games started from 1876-1969. For this, I invented a stat called AOWP+. Scroll down below to see exactly how this stat works. Short version: It's set up like ERA+ or OPS+, centered on 100. A higher score means the pitcher was used more against the best teams. A low score means more against the worst teams. If he's used evenly against all his AOWP+ will be 100.
Having previously figured out who had the best- and worst-leveraged careers, single seasons, how it affected pitchers' stats, answered a critic of this study, and looked at platoon leveraging, here I want to examine which managers did the most and least leveraging. Someone had to leverage those pitchers. They don't decide on their own when to start. Before the rise of the firmly entrenched pitching rotation in the 1960s, this was a key strategic decision for managers.
Quantifying which managers did the most leveraging is thankfully easy. Start with AOWP+. If a pitcher has an AOWP+ of 100, that's zero points for the manager. If it's 101 or 99, give the manager one point. Every succeeding digit AOWP+ moves by, the manager gets extra two points. It looks like this:
AOWP+ Points 100 0 101, 99 1 102, 98 3 103, 97 5 104, 96 7 105, 95 9 110, 90 19
Makes sense? It's such an easy stat I don't even have a name for it. Does "Leverage Points" (LP for short) work for everyone? OK. It's lame, but it works. Figure it for every pitcher with 20 starts, average it out, and ta-dah—that tells you which managers were the Leverage Kings. Let's call it Leverage Points Average, LPA.
But first, a few annoying wrinkles. If a pitcher changes teams midseason I don't count him, even if he got 35 starts with one team and one with the other. It can screw things up. (More importantly, I'd have to completely reconfigure my database, and there's no way I want to do that.) Second, managers often don't last a full season. For these guys, it's all or nothing. If a manager was around for only 80 of 162 games, I don't give him any credit because he wasn't he team's primary manager that year. If he lasts 82 games, I give him full credit. No, it's not perfect, especially given how I handle pitchers. However, it works well enough.
I found every manager who had at least three seasons as a team's primary manager from 1876-1969 and figured their LP. Some of these guys had almost no AOWPd pitchers. (Doc Prothro, for instance, had only three in his days as Phillies skipper.) This approach, however, ensured I got everyone with at least 10 AOWPd pitchers. Also, I figured LP for about 20 of the most prominent post-1969 managers. Since I didn't AOWP many pitchers over the last 35 years, their records are all extremely incomplete, but I made sure I got everyone with at least 10 AOWPd starting pitchers.
Now it's just a simple matter to figure out LPA. For the best LPA, I'm going to break it into two groups: guys with 10-19 pitchers, and managers with 20 or more. It ain't fair to your really long-lasting guys to have them go up against managers who barely got their feet wet. Out of the 72 managers in the 20+ group, here are the champs:
Manager No. SP LPA 1. Frank Chance 37 8.22 2. Hughie Jennings 40 5.35 3. Frankie Frisch 40 4.93 4. Casey Stengel 58 4.90 5. Walter Johnson 26 4.77 6. Johnny Keane 21 4.76 7. Paul Richards 27 4.67 8. Harry Wright 40 4.65 9. Cap Anson 37 4.65 10. Fielder Jones 29 4.62 11. Joe Cronin 36 4.56 12. Jimmy Dykes 60 4.55 13. Billy Southworth 30 4.40 14. Charlie Grimm 45 4.33 15. Charlie Dressen 38 4.26
How 'bout that Frank Chance? It's not too surprising. He had the best leveraged single season of all-time, with Lew Richie in 1912. Mordecai Brown had one of the best leveraged careers ever. My favorite doesn't even show up on the list. In 1912, Ed Reulbach just missed qualifying for LP with 19 starts. Only one of those 19 came against a good team. One. Bill James wrote an essay on him in the original "Historical Abstract" about what an underrated pitcher he was. Well, there's a reason he was underrated.
Also, please note what a wide range of era appears on this list. Harry Wright was the game's first great manager. Stengel, Keane and Dressen lasted until the mid-1960s.
Here's the B team, the 65 managers with 10-19 pitchers:
Manager No. SP LPA 1. Red Rolfe 13 8.54 2. Stan Hack 11 8.18 3. Bob Ferguson 13 8.00 4. Ossie Vitt 11 7.64 5. Pie Traynor 15 7.20 6. Roger Bresnahan 11 6.91 7. Pants Rowland 14 6.29 8. Billy Meyer 10 6.20 9. Hank Bauer 17 5.88 10. Eddie Sawyer 11 5.27 11. Charlie Comiske 19 5.16 12. Eddie Stankey 15 5.13 13. Bill McGunnigle 10 5.10 14. Mickey Cochrane 12 4.83 15. Bill Donovan 11 4.64
See why it's not fair to put these guys with the top 20? Guys with shorter careers would dominate. This makes Frank Chance's score that much more impressive. Only one guy tops him, and not by much, either. With Ferguson, you've got a guy who goes back to the 1870s. Hank Bauer lasted long enough to manage against the Royals.
The no-leverage zone
That's for men who OD'd on leveraging. Now for the abstainers. I need to make one qualifier on this bunch. I'm going to look only at guys who did most of their managing before 1960. Otherwise modern skippers would dominate the list. To give you an idea how times have changed, I'll give you the average LP% for the last 40 AOWPd years on the list.
Manager No. SP LPA 1. Jimmy Collins 20 1.70 2. Fred Haney 21 1.76 3. Miller Huggins 57 2.05 4. Burt Shotton 23 2.17 5. Pat Moran 28 2.18 6. Tris Speaker 20 2.20 7. Pinky Higgins 22 2.23 8. George Stallin 30 2.73 9. Ned Hanlon 41 2.80 10. Patsy Tebeau 22 2.82 11. Tom Loftus 20 2.85 12. Lee Foli 27 2.85 13. Luke Sewell 28 2.86 14. Connie Mack 123 2.93 15. Jimmy McAleer 37 3.00 1966-2005 Avg 1285 1.79
It was a very different game back then. Only Jimmy Collins and Fred Haney leveraged as rarely as modern managers did. Mayo Smith, who straddled the 1960 divide, had a score of 2.23. I'm surprised to see Ned Hanlon on the list, since his disciple, Hughie Jennings, was near the top of the other list.
If you're curious, Ralph Houk had the lowest score of all, with an LP% of 1.38. As a Casey Stengel protégé, that's rather unexpected. Really, though, once you get around/below 1.70, distinctions don't mean squat. There's a certain signal-to-noise ratio, and a manager should have some AOWP+s of 102 and 103. By random happenstance, Houk had fewer of those guys than anyone else. He wasn't leveraging any less than modern managers; the numbers just look funny for him.
Here are the managers from 1876-1960 with 10-19 pitchers:
Manager No. SP LPA 1. Dave Bancroft 10 1.70 2. Jimmie Wilson 18 2.11 3. Ossie Bluege 17 2.18 4. Dan Howley 19 2.26 5. Kid Gleason 12 2.33 6. Ed Barrow 16 2.50 7. Nap Lajoie 14 2.57 8. Horace Philli 13 2.85 9. Arthur Irwin 14 2.86 10. Bill Armour 16 2.88 11. John Ward 15 2.93 12. Gabby Street 17 3.12 12. Donnie Bush 17 3.12 14. George Gibson 13 3.15 15. Gus Schmelz 18 3.22
Even with the reduced standards, there's still almost no one with scores as low as the modern average. Neat. I like seeing Gabby Street here when his old battery mate appeared on the Leverage Kings list.
Adjusting for era
There's another way to look at this—by adjusting eras. Just because leveraging existed for decades doesn't mean it didn't have its ups and downs from the 1870s to 1960s. I can figure out the LPA for years as well (in fact, that'll be the subject of the next article), and see which managers did the least/most leveraging compared to their specific peers.
I have a manager's LPA. I can figure out the LPA for the era he managed in. It's simple—just figure out the LPA for each year in a manager's career and average it out. Then divide the former by the latter, and multiply by 100. Let's call this one Peer Number. (Skip to the next paragraph unless you really care about the methodology: When figuring average LPA for a manager's career, I weight by pitchers, not seasons. So if a manager has three AOWPd pitchers in season X, but only one in season Y, season X will have three times as much impact. Normally, it doesn't matter much either way, but it can make a difference if a manager had an unusually large or small number of AOWPd pitchers per team during one of the periodic boom periods or lulls in leveraging.)
Here are the managers with at least 20 starters who, relative to their peers, did the most leveraging:
Manager No. SP Peer No. 1. Frank Chance 37 199 2. Johnny Keane 21 155 3. Joe Torre (!) 36 153 4. Birdie Tebbetts 22 144 5. Hughie Jennings 40 133 6. Frankie Frisch 40 127.3 7. Walter Johnson 26 127.2 8. Cap Anson 37 126 9. Paul Richards 27 123 10. Casey Stengel 58 119 11. Fielder Jonens 29 118 12. Frank Selee 43 117.3 12. Jimmy Dykes 60 117.3 14. Joe Cronin 36 116.8 15. Fred Mitchell 20 115 16. John McGraw 108 114
When the No. 16 is John McGraw, you have to include him. That list's about what I'd expe—wait—Joe Torre? Joe Frickin' Torre?!?!? The heck? (Looks up Torre's numbers.) No, he really wasn't much of a leverager. His score is a fluke caused by a combination of factors: 1) the lowest score of the last 40 years (Andy Pettitte's 2002) happened on his watch, 2) LP% are so low nowadays it's easier to stand out in a crowd, 3) his score is far more incomplete than that of the other managers. I AOWPd two-thirds of all starts from 1876-1969, but only around 10% from the 1990s. Torre's score is partially a sample size bugaboo. (Prior to the Bronx, I only have brief stints by Jerry Koosman and Phil Nierko under him.) 4) An unusually large collection of 102s and 103 AOWP+s. Thus he makes the Peer Number leaderboard while his career LPA of 2.94 would put him among the 15 lowest pre-1960 managers.
The bottom of the list:
Manager No. SP Peer No. 1. Fred Haney 21 49.35 2. Jimmy Collins 20 49.42 3. Harry Craft 24 55 4. Burt Shotton 23 56 5. Ralph Houk 37 63 6. Pinky Higgins 22 64 7. Miller Huggins 57 69 8. Walt Alston 82 71 9. Pat Moran 28 72 10. Jimmy McAleer 37 72.6 11. Patsy Donovan 23 72.8 12. Luke Sewell 28 76 13. Gene Mauch 35 78 14. George Stallings 30 78.8 15. Tris Speaker 20 78.9
Talk about a photo finish. Lots of 1960s managers on the list. That makes sense—that's when leveraging died out. Looking up, you can see which clubs were on the cutting edge of abandoning it. Now here's the biggest Peer Number for those with 10-19 pitchers:
Manager No. SP Peer No. 1. Hank Bauer 17 251 2. Red Rolfe 13 200 3. Pants Rowland 14 193 4. Ossie Vitt 11 177 5. Bill Shettsline 17 164 6. Pie Traynor 15 163 7. Dave Bristol 15 149.4 8. Billy Meyer 10 148.7 9. Bill Donovan 11 147 10. Stan Hack 11 146 11. Roger Bresnahan 11 145 12. Cito Gaston 12 138 13. Eddie Sawyer 11 136.3 14. Bill Killefer 17 136.8 15. Eddie Stanky 15 134
Again, a few 1960s managers. Only these guys didn't last as long as the Alstons and Houks. Apparently, one of the biggest factors determining how long a manager would last back then was how willing he was to abandon leveraging. Guys like Alston and even Mauch did it initially, but moved away. Hank Bauer stuck with it, and was through by 1970. Please note only two men have more impressive scores than Frank Chance. Remember what I just said about Torre? Same general thing explains Gaston on this list, as well as the modern manager on the next list—here it is, lowest Peer Number for the short career men:
Manager No. SP Peer No. 1. Frank Bancroft 10 39 2. Jimmie Wilson 18 49 3. Dave Bancroft 10 56 4. Nap Lajoie 14 59 5. Dan Howley 19 65.6 6. Jimmy Williams 14 65.7 7. Horace Phillips 13 66 8. Lum Harris 13 69 9. Del Baker 15 71 10. Ossie Bluege 17 73 11. Ed Barrow 16 74.18 12. Arthur Irwin 14 74.21 13. Bill Virdon 18 77 14. Kid Gleason 12 78 15. Gus Schmelz 18 79
Mostly the same names.
Well, the next question is what era had the highest LPA. That's the next article.
References and Resources
What the heck is AOWP+?: The stat I invented to judge pitcher leveraging. It's AOWP/TOWP*100. AOWP is Average Opponent Winning Percentage. TOWP is Team's (Average) Opponent Winning Percentage. To figure AOWP for a single season, you take the number of starts a given pitcher had against each opposing team, and multiply that by the team's winning percentage. After doing this for all rival squads, add up the products and divide by the pitcher's total games started. The result is his AOWP. The same logic applies to TOWP, only here you look at how many games the team played against all rivals. If a pitcher's used evenly, his AOWP will be the same as the TOWP, and he'll have an AOWP+ of 100. If he's used more against better teams, he'll have a higher AOWP+. I calculated AOWP+ for 659 pitchers who started 182,000 games, including more than two-thirds of all games from 1876-1969.
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.