Boone Logan and Javier Lopez signed two of the top six most lucrative free agent deals by relievers this offseason, the former landing $16.5 million from the Rockies and the latter re-signing with the Giants for $13 million.
What do Logan, Lopez and Thatcher have in common? Based on their transactional history, they’re three of the most in-demand lefthanded relievers in the game today. They also have three of the shortest leashes in the game, belonging as they do to a group of a dozen relievers—lefties, all—to appear in at least 20 games last year while averaging fewer than three batters faced per outing.
Such is life for the lefty situational reliever, a job description that seemingly becomes more sought-after and more specialized each season.
Baseball, like all sports, rewards players who can identify an opponent’s weakness and exploit it, and one of the game’s most exploitable weaknesses is an oldie but a goody. Lefthanded batters, as a group, never have been all that productive against lefthanded pitchers, and the left-on-left .645 OPS in 2013 had not dipped lower in more than 40 years, not since registering at .637 in 1971.
|Platoon Split 2011-13||AVG||OBP||SLG||OPS|
|LHP vs. RHB||.262||.329||.419||.748|
|RHP vs. LHB||.260||.331||.413||.744|
|RHP vs. RHB||.249||.308||.389||.697|
|LHP vs. LHB||.235||.297||.352||.649|
At right are the cumulative batting splits for all four left/right platoon-split combinations during the past three seasons, sorted by descending OPS.
The table illustrates why managers aggressively seek the platoon advantage through the use of relievers late in games, and why general managers hoard bullpen inventory as they seek pitchers who can neutralize same-sided batters. Lefty batters sacrifice about 100 points of OPS when a southpaw takes the mound—righties are about 50 points worse in same-sided matchups—and the seven-man bullpens of today can easily accommodate surplus lefty relievers.
Major league teams carried an average of two lefthanded relievers per pitching staff on Opening Day rosters in 2013, and no team went without. In fact, more teams carried three lefty relievers (nine) than carried only one (six), and those tallies do not count Aroldis Chapman, Phil Coke or Glen Perkins, the three lefty closers on Opening Day rosters a year ago.
Furthermore, the sheer quantity of pure relievers—those with zero starts—who throw lefthanded is on the rise. The highest seasonal total all time for lefty pure relievers occurred last year, when 113 made at least one appearance. The next two highest totals? Those happened in 2011 (101) and 2012 (98).
New View On The Platoon Advantage
Teams today accept the proposition that in order to gain the platoon advantage for the pitcher they must cede the corresponding advantage on offense. It’s all about roster construction. In order to have the relievers on hand to match up against the opposition in the late innings, teams regularly devote 12 of 25 roster spots to pitchers, which leaves 13 spots to divvy among position players.
Nine of those spots go to lineup regulars, the ninth being the DH in the American League or the primary pinch-hitter to bat for the pitcher in the National. Once a team has accounted for a backup catcher, a utility infielder and a reserve outfielder, it has only one discretionary roster spot for an extra bat, who might or might not be used in a tactical fashion late in games.
Teams did not always divvy up roster space in this fashion.
Gaining the platoon advantage for one’s offense was a primary concern for managers of the 1980s and early ’90s. In every season from 1984 through 1992, batters faced an opposite-handed pitcher in about 60 percent of all plate appearances, with a high of 61.1 percent in 1990 and a low of 58.9 percent in 1992.
These days, managers prioritize the platoon advantage on the pitcher/defense side of the ball. Batters faced an opposite-handed pitcher about 55 percent of the time in each of the past two years, and that rate has failed to crack 56 percent in each season from 1996 to present. The platoon-advantage rate for batters bottomed out in 2001 (52.7 percent) but has been steadily rising since then, correcting toward the overall average for the Expansion Era (55.4 percent).
The difference between a batter platoon advantage of 52.7 percent (in 2001) and 61.1 percent (in 1990) doesn’t seem like much—8.4 percentage points—but when applied to the 38 batters who appear in a typical game, it translates to an average of three plate appearances per team. In others words, teams received three fewer PAs with the platoon advantage in 2001 than they did in 1990. Even today, they receive about two fewer PAs with the platoon advantage per game when compared with the early ’90s.
What Lefties Have Wrought
Dig into the statistical trends for same-sided matchups and one begins to see why managers, more and more, fixate on gaining the platoon advantage for their pitcher/defense, even if it diminishes their tactical ability on offense. This is particularly true when one pits a lefthanded pitcher against a lefthanded batter. (All data for the following graphs from Baseball-Reference.com.)
First we’ll measure how pitchers have fared in terms of strikeout rate (per plate appearance) in same-sided matchups during the past quarter century. Note that the y-axis, to enhance the display, begins at 14 percent, not zero.
As the graph illustrates, lefthanded pitchers are retiring lefthanded batters via strikeout at an unprecedented rate, which peaked at 23.3 percent in 2012. (In case you’re wondering, the highest value of the Expansion Era prior to 1988 happened in 1968, the Year of the Pitcher, when it was 21 percent.) Given that the right-on-right strikeout rate has historically tracked with the major league average—and actually has regressed toward the average in recent years—we can conclude that left-on-left matchups are more than a little responsible for driving the game’s overall strikeout increase. The cumulative strikeout rates over the last three years:
|Platoon Split, 2011-13||SO%|
|Major League Average||19.4%|
Next, let’s look at how isolated power rates have changed in the same platoon splits during the same time period. Note again that the y-axis does not begin at zero. It begins at .100.
Here we see that lefty pitchers are much more successful at limiting extra-base hits to lefty batters than are pitchers in RoR matchups. Part of the difference is the enhanced strikeout rate previously alluded to, but even when making contact, the home run rate in LoL matchups (2.8 percent) has not been lower in nearly 20 years. (The RoR value of 3.6 percent is almost exactly league average.) Again, we see that the RoR trendline tracks closely with the major league average, which suggests that LoL matchups have unduly influenced the overall recession of total bases in the majors. The cumulative rates for isolated power over the last three years:
|Platoon Split, 2011-13||ISO|
|Major League Average||.146|
Next up we’ll consider batting average, though this evidence is not nearly so convincing as strikeout rate or isolated power because the extreme defensive shifts employed by many teams today have a measurable effect on the hit rate of lefthanded pull hitters. Again, note that the y-axis begins at .230, not zero.
We won’t dwell on this one because of the aforementioned defensive shifts, but note the consistent pattern we see with strikeout rate and isolated power. Pitchers in LoL matchups enjoy a distinct advantage, one that seems to grow each year. The cumulative rates for batting average over the last three years:
|Platoon Split, 2011-13||AVG|
|Major League Average||.254|
We see from this data that a lefthanded or righthanded batter with the platoon advantage, from 2011-13, has hit about .261 with a .154 isolated slugging percentage and a strikeout rate of 18.5 percent. However, if a manager spots a righthanded pitcher against a righty batter, the batter hits .249 AVG with a .140 ISO and a 20.0 SO% (in the aggregate). If he spots a lefthanded pitcher against a lefty batter, the batter hits .235 AVG, with a .118 ISO and a 22.8 SO% (in the aggregate).
Obviously, not every lefty batter will suffer the deleterious effects of the platoon disadvantage to the same extent as the overall LoL average. The best of the best middle-of-the-order lefty bats will be serviceable versus southpaws—but they still suffer a dropoff in production when compared to their split versus righties. All of them. David Ortiz lost 132 points of OPS when spotted against a lefty from 2011-13. Joey Votto lost 94 points. Prince Fielder lost 145 points. Robinson Cano lost 220 points. Even three-time batting champ Joe Mauer lost 145 points.
The fact that lefthanded pitchers tend to neutralize lefthanded batters is not news, of course, but what’s scary for the future of this particular platoon split is that lefty pitchers have shown dramatically improved control versus lefty batters in the past few seasons. Southpaws issued unintentional walks in just 6.7 percent of plate appearances versus lefty batters last season, which is the lowest rate of the Expansion Era. You can imagine the effect this has had on LoL on-base percentage, which bottomed out at .294 in 2012, a depth not previously reached since 1968, the Year of the Pitcher, when it was—gulp—.277.
Perhaps the extent to which the platoon advantage favors the lefty pitcher might encourage more teams to cultivate righthanded batters, who are not punished by same-sided pitchers to nearly the same degree. Right on cue, the 2014 rookie class will feature a number of high-ceiling, righty-hitting prospects, including Red Sox shortstop Xander Bogaerts, White Sox first baseman Jose Abreu, Tigers third baseman Nick Castellanos, Astros outfielder George Springer and possibly Cubs shortstop Javier Baez. Diamondbacks shortstop Chris Owings and White Sox third baseman Matt Davidson have paths to playing time, if not the lofty ceilings of the aforementioned.
The keen-eyed reader will note that the distinct disparity in LoL performance didn’t really begin to develop until about 2002. Prior to that, the rates for batting average and strikeout rate in LoL and RoR matchups tended to run parallel and even criss-cross, while LoL isolated power improved gradually until it practically touched the RoR rate at the turn of the century. For example, in the year 2001 we find not nearly the gulf in same-sided performance we see today.
|Rates in 2001||LoL||RoR||MLB|
Lefthanded pitchers are striking out more lefthanded batters and walking fewer than ever, while allowing fewer hits than they have since the 1960s and early ’70s, so naturally their performance in that platoon split has improved dramatically. The question is: What factors have contributed to this improvement? We’ll dig into some of the leading theories in the next installment.