Pitcher Platoon Leveraging (SP Lev., Part 5)by Chris Jaffe
April 30, 2007
Welcome back to yet another article here at THT on starting pitcher leveraging, a.k.a. what happens when teams intentionally use their pitchers against specific teams as often as they can to aid the team's success. Though this practice no longer exists in any meaningful way, for almost a century it was a primary way managers utilized their rotations. Previous articles have looked at the best and worst careers and single seasons of all time, and other areas of interest by using a stat I invented called AOWP+. If you're curious, scroll to the bottom of this article for an idea what AOWP+. However, I’m not even going to give a thumbnail description of AOWP+ here, because this article focuses on a different metric.
The most striking finding in the previous articles is the excessive representation of left-handed pitchers on the career and single-season AOWP+ leaderboards. Despite accounting for only one-fourth of the pitchers I AOWPd, 14 of the highest 16 career marks belonged to southpaws. The reason for this was platoon leveraging. Managers often started their southpaws as frequently as possible against teams with the best left-handed hitting such as the Ty Cobb & Sam Crawford's Tigers, Stan Musial and Enos Slaughter's Red Birds, and the Ruth/Gehrig Yankees. Enough of the game's best teams had threatening lefty bats that platoon leveraging became one of the most important forms of starting pitcher leveraging.
Since platoon leveraging is so dang important to the overall story of pitcher leveraging, it's worth focusing on it and trying to quantify it. For this, I need a new metric because that ain't what I designed AOWP+ for.
Enter L#The new metric, called L# (I haven't made up my mind if "L" stands for lefty or leverage. Both I guess) differs significantly from AOWP+. Whereas an AOWP+ revealed how a team used an individual pitcher, L# determines how opposing pitchers were used against a team. Instead of quantifying how the A"s used Eddie Plank in 1914, L# determines how the rest of the AL used lefties against the 1914 A's.
Here’s the concept. Take the 170 or so lefties I AOWPd, and find out how many times they started against each opposing club in every season. That's L#’s numerator. The denominator is how many starts southpaws should have had if they had been used evenly. Divide the numerator by the denominator, multiply by 100, and round to the nearest integer. That's L#.
Let me use an example. In the 1941 NL, I AOWPd 10 lefties: Johnny Vander Meer, Ken Raffensberger, Johnny Schmitz, Howie Pollett, Carl Hubbell, Max Lanier, Larry French, Cliff Melton, Ken Heintzelman, and Dave Koslo. They started 154 games. Thus the Boston Braves should've faced them approximately 22.0 times. (154/7 = 22.0. You divide by seven because these pitchers all had seven opposing teams to start against). In reality, they started 24 games against Boston. Twenty-four divided by 22.0 times 100 is 109. That's their L#.
Or, still using the 1941 NL, there are the Reds. Unlike the Braves, who had none of the AOWPd lefties on their roster, the Reds had Vander Meer. Deduct his 32 starts, and the remaining southpaws started 122 games. 17.4 should have been against Cincy. In reality, the other hurlers started a grand total of three games against the Reds. Yeah, that's right. Three. All came from the Cubs (twice by French and the other by Schmitz). Three divided by 17.4 times 100 gives them an L# of 17. They were the defending world champs, had won the last two pennants, their only left-handed bat (Lonny Frey) wasn't a great hitter, and Crosley Field at that time was much kinder to right handed pitchers.
I figured L# for every team I could through 1969. In the process I learned something - there weren't many left handed pitchers in nineteenth century baseball. Until the mid-1880s, Lee Richmond was the only one worth a fart. When they did emerge, they primarily worked in TheA. As late as 1891, there was only one—count him, one—AOWPd lefty in the entire NL (Duke Esper—yeah, I never heard of him either). By the 1890s there were more lefties, but they still made up a smaller percentage of starts than at any point in the twentieth century. The first decade of the new century was their breakthrough period. I have no idea why, but there it is.
After 1892, there were virtually always a decent number of lefties in all MLB. Well, not the Federal League. Between a lack of lefties and a lower level of competition, I had almost no AOWPd southpaws there, so I'm ignoring that league. The only other leagues since 1892 with under 100 GS by AOWPd lefthanders were the 1922 AL (85 GS) and 1945 NL (67). So I can find out which teams had the most platoon leveraging thrown at them from 1892-1969.
Highest L#s, 1892-1969
One key qualifier here: None of this is based on a full sample size. It's just AOWPd guys, but again, that's 2/3rs of all starts in the years in question. Here's the highest L#s:
Year Team L# 1 1895 PHP 202 2 1909 BOX 185 3 1957 CWS 182 4 1955 STC 176 5 1892 CLV 174 6 1942 STC 173 7 1933 PIT 172 8 1940 STC 168 9t 1945 CHC 167 9t 1908 NYG 167 11 1936 WAS 166 12t 1908 DET 165 12t 1951 CIN 165 14 1915 CHC 164 15 1908 WAS 162 16t 1939 NYY 161 16t 1892 LOU 161 18t 1940 CLE 160 18t 1955 CLE 160 20t 1935 PIT 159 20t 1959 CWS 159
First, platoon leveraging existed in the nineteenth century. The relative dearth of southpaws in the 1890s inflated Philly's leading score. There were only 7 AOWPd lefties in the league that year, and one of them (Kid McGill) pitched for the City of Brotherly Love. The remainder only had 136 starts, for a mere 12.4 expected against them in that 12-team league. Ted Breitenstein had seven of his 51 starts against them. Duke Esper had six of his 25. Heck, Phil Knell had six of his 19 against them. It adds up. Instead of 12.4, they saw 25 AOWPd lefties. The year before Cooperstown-bound lefties Sam Thompson and Billy Hamilton both hit .400. Then again, so did right-handed Ed Delahanty, but apparently he wasn’t the bat opponents feared the most.
This brings up a question: when the heck did platoon leveraging begin? Two clubs on the list are from 1892 , so the notion of platoon leveraging pitchers had already become a common strategy by then. Its origins must go back earlier. As early as 1885-6, the AA had a good number of southpaws. Heck the 1886 Double-A had the highest percentage of games thrown by lefties in any league until the twentieth century. Looking it up, the L#s aren't as stark for the most part as they would later be. Virtually every team has a score between 70 and 130. Most of the extremes came at the end of the AA's existence. That's good. You want to see a gradual rise as the practice proliferated across the league. Better yet, one team has a clear pattern that can be easily explained, the St. Louis Browns. Their L#s:
Year L# Rank 1886 85 Lowest 1887 86 2nd Lowest 1888 88 2nd Lowest 1889 112 Highest (!) 1890 65 Lowest real team 1891 84 3rd Lowest 1892 65 2nd Lowest
(In 1890, the Baltimore Orioles had a lower L#, but they played less than 40 games). The Browns were the best team in the AA. They had an almost entirely right-handed line up. And they saw fewer lefties than about any team. Perfect. Their scores' modesty reveals that the concept of platoon leveraging was just taking root throughout the league. In 1889, one pitcher—Frank Foreman—skewed the data drastically by starting 10 games against them. There weren't as many southpaws as there had been in 1885-6, and he could affect the data that drastically by himself.
Is that the beginning? Well, if you push it back to 1885 St. Louis has a pedestrian L# of 98, which is meaningless. Some teams had considerably high or low L#s that year, but the numbers look random. When there’s barely 100 starts by lefties (most of them by one pitcher, Ed Morris), one team facing a little than they should against them can give them a league leading L#. Brooklyn, for example, had a 137 mark in 1885, but looking at their line up I can’t see why. At this point you have to start looking at how individual teams used their southpaws to see if there’s any signs they platoon leveraged them against the competition.
Sure enough, some early platoon leveraging predated the AA. In Part Two of this series I noticed that in 1879 Harry Wright, the game's first great manager, platoon leveraged his starters. He avoided using ace pitcher Tommy Bond against the Providence club, which had the game's best left handed hitting. Instead, Wright used southpaws like Curry Foley, and Lee Richmond. I’m willing to bet Jim Tyng (who had all his starts against Providence) was also a lefty. This would be a perfect story to say platoon leveraging began then. You’d want a long-lasting practice like this to be started by one of the sharpest of the ancient wise men, and that's Harry Wright.
But facts have this tricky little way of not fitting our narratives. Looking backwards, while there were virtually no lefthanders at all, one man does look like he was leveraged before 1879. Instead of Harry Wright inventing the practice, the honor falls to Cincinnati’s manager Cal McVey, who used southpaw Bobby Mitchell in a revolutionary manner in 1878.
Mitchell only had 9 starts that year. Despite only having 9 starts, he was with the club all year. Such was the lot of a #2 pitcher for a team that played 61 games back in the early stone age. Here’s how many starts came against the five opposing clubs (yea, only 5 other clubs back then), how many lefties were in their starting line-ups, and the OPS+s for those hitters:
Team GS Lefty OPS+ Prov 4 3 158, 144, 82 Ind 2 1 186 Mil 2 2 151, 71 Chi 1 2 152, 110 Bos 0 0 None
Admittedly, Chicago also had two switch hitters, but there’s a clear pattern. He was platoon leveraged, and looking backwards I can’t find any other examples when this happened. It’s possible someone did it in the NA, but it’s impossible to tell. They had no set schedule for Pete’s sake.
L# & the Origins of Offensive PlatooningThe above is mighty nice, there’s a second item on the top-20 list that leaps out at me. Tied for 12th on the all-time leaderboard were the 1908 Tigers. In the Historical Baseball Abstract Bill James wrote about the origins of platooning. Not of pitchers, but offensive platooning. With the help of Jim Baker, James figured that the 1906 Tigers were the first team to have any offensive platooning when they juggled John Warner, Boss Schmidt, and Freddie Payne at the backstop. This forgettable second division squad inventing platooning greatly disappointed James. He wrote:
It makes a terrible story, you see. The first known platoon arrangement was operated by some guy named Bill Armour, who nobody knows anything about, who thought it up and implemented it in his fifth and final season as a major league manager, had a poor year and was fired?
Two years later, that very same franchise scored one of the highest L#s ever. Hmmmm . . . Is that just a coincidence or was something going on? Let’s look closer. Here’s the L#s for those early twentieth century Detroit teams, and where they ranked in the AL:
Year L# Rank 1904 114 2nd Highest in AL 1905 113 2nd Highest 1906 118 2nd Highest 1907 120 4th Highest 1908 165 Highest 1909 150 2nd Highest 1910 154 Highest 1911 141 Highest
The White Sox were the team usually in first place. In 1907, the only year the Tigers weren’t in the top two slots, they were in a three-way photo finish for second place with the Red Sox (123 L#), and Browns (124). At the same time the Tigers invented offensive platooning, they constantly had more lefties thrown at them than almost any other team in the league. Well, that’s interesting.
But remember what I said earlier in the article? The turn-of-the-century witnessed an explosion in the number of southpaws. Also, in the first decade, far more lefties were in the AL than in the NL. It began balancing out as the decade went on, but the AL was where the lefties were. Remember how those 1895 Phillies only had to face 25 lefties to get an L# over 200? The '06 offensive platoon inventing Tigers, with their 118 L#, saw 36 AOWPd lefties. In 1895 lefties started only one out of every seven games. By 1906, they started one-fourth of all games.
The Teddy Roosevelt Era White Sox and Tigers saw southpaws more frequently than any team ever had in baseball history up to that point. If ever a team was going to invent an offensive strategy to counter pitcher platooning, it should have been one of those clubs. And guess what? One did. I have no smoking gun bit proof showing that A happened because of B, but all the circumstantial evidence indicates that not only did pitcher platooning exist, not only did it predate offensive platooning, but the latter developed in response to it.
This also solves one riddle from the “Historic Abstract.” For the life of him, James couldn’t find any contemporary articles discussing platooning’s birth. James theorized, “maybe that’s why platooning has no myth of creativity, like everything else; it just sort of snuck up on everybody, and it was fifteen years old before anybody realized it was here to stay.” Well, sorta. The 1906 creation wasn't so much an invention of platooning as it was an inversion of it. James couldn’t find articles on its origins because it was a reaction more than an action.
Sadly, all this does is replace a terrible story with an even worse one. No, platooning didn’t begin with a manager you never heard of at the end of his career platooning the guy who beat up Ty Cobb a few years before the Tigers became good. Instead it was a man who only managed less than 160 games using a pitcher who tossed under 400 innings in his career for a franchise that collapsed during the Rutherford B. Hayes administration. Yippee.
It would be so much nicer if it had begun with Harry Wright handling Tommy Bond and his lefthanders for the Boston club in 1879. But facts are stubborn things. Platooning began the year before when Cal McVey sport started the forgettable Bobby Mitchell. A generation later offensive platooning developed in response to it, which in turn helped spawn the modern day LOOGY. Jesse Orosco is the bastard grandson of Bobby Mitchell.
I still have the bottom 20 L#s to go over, but this article’s getting a little longer than I’d like. That’ll be in the next article in this series where. The heart of the next work, however, will be a data dump. Right now I’m just sitting on all these L#s that no one has ever calculated before so I may as well share ‘em.
References and Resources
Source: James, Bill. The Bill James Historical Baseball Abstact. New York: Villard Books, 1986. The article on platooning is one of the longest articles in the book. It's on pages 112-123. I pulled the quotes from pages 121 and 122.
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 GS. 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 over 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.