Note: You can listen to Darius discuss this article on Effectively Wild Episode 812 with Ben Lindbergh and Steven Goldman.

 

Roger Kahn’s “The Head Game” is about a series of pitching greats, one of whom is Warren Spahn. The legendary left-hander spent 21 years in the majors, predominantly with the Milwaukee Braves. He pitched over 5,000 innings, winning 363 games, and racking up more than 90 WAR. In his homage to the Hall of Famer, Kahn supplies a steady stream of statistics and anecdotes that highlight Spahn’s quality, but amidst the impressive array of achievements, there is a somewhat strange note, which is as follows:

“He was not flawless. For three years during the 1950s, he did not start at Ebbetts Field. The combination of powerful right-handed Dodger hitters and a short left-field fence was too much for him.” (The Head Game, Roger Kahn, p. 163)

It seems implausible that one of the greatest left-handed pitchers of all time just didn’t pitch away games against the Brooklyn Dodgers for three seasons, regardless of how good their right-handed hitters were. Implausible, and yet Kahn’s statement actually undersells the extent to which Spahn was held out of Brooklyn matchups. Baseball-Reference reveals that not only did Spahn not pitch at Ebbets Field between 1954 and 1957 (when the Dodgers left for Los Angeles), he pitched just 3 2/3 innings against Brooklyn in total over those four seasons. These are not seasons in which Spahn missed time or was ineffective either: his ERA from 1954-7 was 2.96 (best in the league among pitchers with at least 600 innings), his ERA+ was 122 (third), and his FIP was 3.47 (fifth). He never started fewer than 32 games, and appeared in 39 in each of the four seasons. In 1956, he finished third in Cy Young voting, and he was the runaway winner in 1957 – when the Braves also won the World Series – with 15 out of 16 votes. Yet for that entire period, Spahn made just one start, and three appearances in total, against a team that twice won the pennant and finished second and third in the other two years. In short, Spahn didn’t pitch against the best team in the National League between 1954 and 1957. That was in spite of the fact that the Braves and Dodgers faced each other 88 times in that four-year span.

Joe Posnanski was also amazed when he came across this phenomenon a couple of years ago. His first port of call for an answer was Bill James, who suggested that this was pretty much par for the course when it came to pitcher strategy and the approach taken to the Dodgers’ lineup during the 50s:

Managers and pitchers worried a lot more about team match-ups then and a lot less about individual match ups. Bill says the Braves were actually kind of stubborn in the early 1950s about pitching Spahn against the Dodgers — he said most teams had already decided it was pointless and self-defeating to start lefties against that lineup. The Braves finally gave up in 1954 and they saved Spahn for every other team.

So just how far were teams willing to take this strategy? Posnanski cites the number of starts made by lefties against the Dodgers in the mid 50s, but I wanted to take it a bit further and show the whole decade – you’ll see why. Here’s the total number of games the Dodgers had against left-handed starters during the 50s, courtesy of the trusty Play Index Split Finder:

 

YearGames vs LH Starter
195056
195140
195221
195324
195418
195511
195614
19576
195825
195926

 

The Dodgers went from 56 starts against LH starters in 1950 to six in 1957. The rest of the league were more reluctant to start a left-hander against Brooklyn in the mid-50s than they were to pitch to Barry Bonds with a man on base in 2004. It seems that every team really was just as reluctant to use lefties against the Dodgers, employing a no-lefty strategy even if they could start one of the finest left-handers in the game, as the Braves could.

In order to provide a contrast, and to be sure this wasn’t also linked to some league-wide phenomenon, I also wanted to compare the Dodgers’ percentage of plate appearances against southpaws compared to the league average in each season. However, Baseball-Reference does not have entirely complete data for platoon splits during this period; they estimate that 1-2% of plate appearances are missing, and there are some teams with significant numbers of plate appearances that are clearly lacking platoon data, making a percentage fairly irrelevant. Nevertheless, platoon data for the Dodgers is complete – presumably thanks to the efforts of their remarkable statistician Allan Roth and/or the volunteers at retrosheet.org – so we can at least be confident in their stats specifically. I therefore instead have compared the Dodgers’ percentage of PA against lefties to other teams that do have complete data, which is certainly not as comprehensive as desired, but still serves to illustrate the extremity of the split.

 Dodgers vs LHP

There is something of a trend downwards in overall league PA against lefties, but it’s clear that as soon as 1951, Brooklyn had already started to deviate beyond what might be expected from that much more moderate decline, and by 1954 it was getting downright uncommon for any left-hander to face the Dodgers. The huge gaps between Brooklyn and the team with the second-lowest percentage in 1955 and 1956, and the record-low mark of just 5.8% of PA vs LHP in 1957, demonstrates the extremes to which this approach was taken.

Intrigued by the prospect of this as a strategy, whilst acknowledging that it was probably impossible in modern baseball, I sent the question to Ben and Sam, and was lucky to get it answered on an email edition of Effectively Wild. Aside from pointing out that it was generally a pretty bad idea, and later adding (courtesy of BP’s Craig Goldstein) that a rather disgruntled Wei-Yin Chen had actually been skipped in such a fashion to avoid the predominantly right-handed offensive juggernaut that was the Blue Jays once in 2015, they also wondered how the Braves had actually managed Spahn during this stretch, and whether he had actually been moved up to pitch on short rest, rather than simply having him miss starts to avoid the Dodgers.

Despite – or perhaps because of – the ubiquity of this strategy, further information on this fear of the lefty-killing Dodgers and the manner in which managers handled it is somewhat harder to come by, with even searches of newspaper archives from the time bringing up very little. A trawl through the archives of the Brooklyn Eagle eventually did reveal a mention of the incredible fear the Boys of Summer struck into opposing managers, and specifically Spahn’s manager of the time, Charlie Grimm. In the January 26th, 1954 edition, writer Tommy Holmes highlighted some of the most fascinating statistics compiled by Roth, who was preparing for the team’s 1954 Year Book. At this time, Spahn was still being allowed to start against the Dodgers, but Holmes notes that Milwaukee manager Grimm “subscribes to the popular theory that using left handers against the Dodgers is leading with your chin”, and Spahn only took two losses because Grimm only let him start three games total against Brooklyn.

Spahn hadn’t even pitched that badly against the Dodgers in 1952, at least from an ERA standpoint, with a 3.47 mark. He did, however, go 0-5 in his five starts, and that seems to have triggered Grimm’s reluctance to use even the best of left-handers against Brooklyn. In 1953, Spahn had one heroic, eleven-inning effort against Brooklyn, which he might have won in nine had it not been for an error that allowed a run to score in the third, but the other two outings were not pretty. Spahn lasted fewer than three innings in both, giving up a total of ten runs (although only seven were earned, after Milwaukee threw away a lead in a September 2nd start by making three errors in the second inning to let three runs score).

With Spahn 0-7 against the Dodgers in his time as Braves manager, Grimm’s chin had taken enough of a beating: Spahn would not get to take the mound for a start against Brooklyn again until June 5th, 1956, a decision Charlie appears to have quickly regretted. Spahn pitched a scoreless first on that day, and then surrendered two runs in the second when he gave up a single to three-time MVP and future Hall of Famer Roy Campanella, another to right fielder Carl Furillo, walked pitcher Roger Craig and then allowed another single to leadoff hitter Jim Gilliam. Grimm had seen enough and pulled Spahn after the future Hall of Famer had recorded only four outs and before he had to face two more players who ended up in the Hall of Fame, Pee Wee Reese and Jackie Robinson (yes, the Dodgers were really good in the 50s). The Milwaukee manager would be fired just weeks later, although whether it was directly related to the fearsome Dodgers giving his best starting pitcher fits is unclear.

Grimm’s successor, Fred Haney, largely continued with this policy throughout 1958, although Spahn was permitted two starts. It would take until 1961 for Spahn to make as many as five starts against the Dodgers in a season again, a full ten years since he had last done so in 1951.

Sam’s question of whether Spahn was actually moved up to avoid the Dodgers requires some more in-depth digging through game logs. There’s very little consistency in the number of days rest for any of the pitchers during this time period, and even with 22 games a year against Brooklyn, there are a number of series in which Spahn seems to naturally miss Brooklyn games without needing to be pushed up or held back.

It is easier to see that Grimm was practically forced into using Spahn in the aforementioned 1956 start. The Dodgers were in town for a four game set, Ray Crone and Bob Buhl (he of 1-for-100 fame) had both pitched in a Sunday doubleheader against Pittsburgh the day before the series started, with Buhl going nine innings, and Lew Burdette had started the first game on the Monday. Gene Conley had gotten the ball on Saturday, which meant that Spahn, with three days off since his last start, was clearly the most well-rested member of the rotation. Had Grimm skipped him for the whole series, two of his other three regular starters would have been pitching on two days’ rest, with Spahn waiting a full week between starts.

The All-Star break makes the situation a little clearer. In 1954, Spahn pitched two days before the break, on Saturday, July 10th. His next start didn’t come until Sunday, July 18th. The reason? Milwaukee would resume after the break with a five-game set against the Dodgers, in which Grimm seemingly used every other starter available to him, including right-hander Dave Jolly, in what was the first and only start of his major league career. Jolly actually pitched very well, surrendering just one run over ten innings, but the Braves lost in the eleventh when Grimm brought in Burdette, who had already started the first game of the series, and Gil Hodges singled home Reese. Spahn would get 7 days of rest on this occasion as a result.

The year 1955 offers less in the way of clear Brooklyn-dodging, but there was another case of a week between starts for Spahn in September, when he was held out of a two-game set against Brooklyn despite having last started five days before the first of those games. Grimm did bring in Spahn for an inning against the Cubs the game before, presumably knowing that he was never going to utilise Spahn in the subsequent two games. In 1956, following Grimm’s last ill-fated attempt to start the lefty against Brooklyn and subsequent firing, another series immediately following the All-Star break loomed. This time Spahn, having started three days before the break, was given eight days of rest by Haney just so he didn’t have to face the Dodgers. Spahn would have at least six days between starts twice more in 1957 as a result of avoiding the Dodgers, although in the second case Haney, like Grimm in ’55, did bring him in for a brief relief appearance in-between the two starts.

Milwaukee did lose some Spahn starts and innings by adopting this policy, even if it only worked out at one or two starts a season. Spahn pitched over 290 innings in each of the four seasons prior to 1953. He wouldn’t get back to that mark until 1958, following that five-year span that included a season of ‘just’ 32 starts and 245 2/3 innings in 1955, when he was not used at all against the Dodgers, even in relief.

Given the previously-cited start statistics, it’s abundantly clear that the whole league did just throw in the towel when it came to using left-handers against Brooklyn, so Spahn certainly wasn’t alone. Did that make any sense? By OPS+ for this split, the best team against left-handed pitchers in 1950 was actually the Red Sox at 119, but Brooklyn was right behind them at 118. However, Boston absolutely crushed right-handers, to the tune of a ludicrous .864 OPS, while the Dodgers managed a respectable – but hardly remarkable – .783. The Dodgers certainly appeared to have a pronounced advantage against lefties, whereas the Red Sox were going to destroy pitching either way; by wOBA, they actually rank as the best offense since the 1936 Yankees, and the 1948-50 Red Sox have three of the best five wOBA seasons since World War II. This isn’t to say that the Dodgers weren’t a great offense themselves – the 1950-55 teams represent six of the top 40 wOBA seasons in the same period – but their splits were certainly gaudier against left-handers in 1950, right before the trend started.

Then, for each of the next five seasons, the Dodgers were the best team in the league against lefties, and it got more and more extreme. By 1953 they were 20% better than second-place Milwaukee, putting up an .826 OPS against lefties. In 1954 they improved, to .831. In 1955 they sent left-handers screaming back to the bullpen with an incredible .911 team OPS, fifty percent better than league average. A separate search for the best splits against LHP in history (data available in the Split Finder is from 1914 onwards) reveals that only two teams had ever managed a higher OPS+, and upon closer inspection, both of those come from the early 1940s, when platoon data is extremely limited – in other words, the stats provide very little indication of whether the 1941 Giants or the 1942 Red Sox (with war-depleted rosters) were actually good against lefties.

The 1955 Dodgers therefore have a legitimate claim as the greatest lefty-crushing team of all time: no other team with anything approaching complete platoon data is close to their 150 sOPS+, with the ’98 Rangers and ’68 Reds tied for second with 134. It should be noted that against left-handers that year, they did have less than what would be considered a season’s worth of plate appearances for a single player: 484, a fairly small sample size.

The league didn’t immediately adjust when Brooklyn was essentially league-average in 1956 either, as evidenced by the fact that the Dodgers didn’t face a left-hander at all in over two-thirds of their games. They were back to their old tricks in 1957, jumping back up to more than 30% above league average, but that was in a very small sample of just 346 plate appearances. Then, in 1958, the worst of the fear seemed to have gone, as their number of plate appearances against lefties more than doubled and they were once again just league-average. By 1959 there was nothing remarkable about the Dodgers’ splits at all, with three teams facing a lower percentage of lefties.

The reversal of the trend was perhaps at least partly triggered by the fact that Ebbets Field was no longer in play in 1958, as the Brooklyn Dodgers crossed the country to become the Los Angeles Dodgers. While managers retained some of their caution in the early days of the LA team, that soon dissipated.

The greats of the early to mid-50s were also declining, and several of those Hall of Famers were either no longer on the team or were soon to depart. Robinson had retired after the 1956 season at 37, and would have been in a different uniform even had he stayed in the game, after the Dodgers traded him to the Giants that offseason. Reese, at 38, struggled to a .555 OPS in 1957 and fell out of the starting lineup. Campanella, an excellent hitter whether facing left- or right-handers (career .860 OPS, 123 OPS+) but particularly potent against southpaws, had barely got his OPS above .700 for two seasons, and was tragically paralysed in a traffic accident right before spring training in 1958 that left his neck broken and spinal cord irreparably damaged. Hodges, Duke Snider and Furillo were the only above-average regulars at the plate in 1957, and the lefty-swinging Snider never hit left-handed pitching that well, doing a huge proportion of his damage against righties (career .949 OPS, 374 of his 407 home runs vs RHP). As Posnanski notes, Snider should have faced left-handers much more often given his drastic splits, but his right-handed hitting teammates saved him from doing so for the majority of his career.

In total, only 17 different left-handed starters would get to take the mound for a start against the Dodgers in the 616 regular season games they played between 1954 and 1957. Eight of those, including Spahn, would get just one start, and only four would be permitted to start five times or more: Jim Davis and Curt Simmons (five each), Harvey Haddix (six), and Johnny Antonelli (twelve). Davis was never a full-time starter, so it’s not entirely clear why Cubs manager Stan Hack used him specifically in those five starts, particularly when he didn’t use any other lefties to start against Brooklyn between 1955 and 1957. Four of Simmons’ five starts came in 1954, two of which didn’t go beyond the fifth, and he would get just one more in the next three years.

Leo Durocher, at least, did not seem to be quite as frightened of the Dodgers. It was the legendary Giants and Dodgers skipper who gave New York’s Antonelli those 12 starts against his former club over the period in question. Nonetheless, considering that Antonelli made 137 starts in that span, that total was still seven or eight below what might be expected if the starts had been distributed equally amongst the seven National League teams the Giants faced.

What did the Dodgers themselves think of all this? Apparently, they didn’t know just how much of an advantage they had against lefties. In a May 10th, 1955 article for the New York Times, Joseph M. Sheehan examined the reasons behind the platoon split (referred to as the “left-right percentage”). After looking at some of the causes of the split, Sheehan notes that some right-handed players actually claim to prefer facing righties:

The Dodgers, whose batting power is predominantly right-handed and who consequently are fed a steady diet of right-handed pitching, made a particular point of [preferring right-handers]. Jackie Robinson, Roy Campanella, Carl Furillo and Gil Hodges chorused “We like right-handers best.”

Perhaps they were simply so good that the hitters really didn’t notice that they weren’t as good against righties. Perhaps they were hoping that opposing managers would start using lefties against them again after seeing that quote. Perhaps it was just so rare for them to face a left-hander, it was natural that they’d place more value on their performance against righties, which by this point made up over 90% of their at-bats. Whatever the reason, it’s clear that keeping your left-handers away from Brooklyn was a strategy that was so widely accepted, it applied to every single pitcher on every team’s staff, and while it may have been something of an overreaction to prevent a pitcher of Spahn’s quality from facing them, it was also an approach based on some remarkable dominance of all lefties. It seems fitting to end on a quote in that Sheehan piece from Snider, who displayed a touch more awareness than his teammates of the Dodgers’ incredible platoon advantage:

“I don’t dislike southpaws. I just don’t know much about them. And I’m quite willing to let matters stand.”
Next post:
Previous post:

3 Responses to “The Brooklyn Dodgers’ Incredible Platoon Split”

Trackbacks/Pingbacks

  1.  In Defense of Bull Durham | Banished to the Pen

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.