This afternoon, the Dodgers are starting a four game series against the Rockies in Coors field. Given what usually happens to Dodgers there, that’s pretty terrifying. There’s also a hidden part of the series that might be cause for some mild concern. The following starters are lined up to pitch against the Dodgers: Franklin Morales, TBD, Jorge De La Rosa, and Yohan Flande (?). The names aren’t the scary part, of course. It’s the fact that all three of them are left-handed. Christian Friedrich went after Flande last time around and was originally in line to start on Sunday before he was sent down to clear room for Nolan Arenado.
So far this year, the Dodgers have struggled against southpaws, producing a .230/.310/.368 line against them. The batting average makes the line sound abysmal, but the walks and the slugging make it “only” mediocre; the Dodgers’ 97 wRC+ against left-handed pitchers is 19th in baseball. When you compare that line to what the Dodgers do against right-handed pitchers (.275/.343/.439, 121 wRC+, best in baseball), it sticks out even more.
There are a lot of reasons for this, of course. Andre Ethier, Carl Crawford, and Dee Gordon (who is leading off against a lefty) famously have bad platoon issues. Scott Van Slyke offsets that a bit, but there are two culprits who are getting most of the attention in this regard this year. Matt Kemp is hitting .214/.267/.333 against lefties in 2014, as opposed to .288/.349/.492 against right-handed pitchers. Adrian Gonzalez has shown a mild aversion to left-handed pitching in the past, but nothing like this year’s Ethier-esque .178/.242/.244 batting line.
Those splits will be cited to criticize batting orders or to generate discussion, but here’s the problem: statistically, those splits mean almost nothing. It takes 1000 plate appearances for a left-handed hitter’s platoon splits to stabilize against a left-handed pitcher. This year, Gonzalez has had 99 plate appearances against left-handed pitching. It takes 2200 plate appearances for a right-handed hitter’s platoon splits to stabilize against a left-handed pitcher. Kemp has only 90 plate appearances against lefties this year. His career number of plate appearances are about halfway to the stabilization threshold. Platoon splits are wildly variable and take a very long time to stabilize.
Before the season started, I used regression methods and Steamer projections to determine the best second base platoon configuration. That spreadsheet still exists, so it’s easy enough to use the same method to forecast Kemp and Gonzalez’ platoon splits going forward. The forecasts use each player’s Rest-of-Season Steamer projection as the overall mean performance, then the career splits (and, in Kemp’s case, regression) are used to determine the forecast lefty split for the rest of the year.
|wOBA against LHP, 2014||wOBA against LHP, forecast|
The table shows that each player is expected to improve their split against left-handed pitching by about .100 points of wOBA, which is an extremely significant amount. Gonzalez’ projected split is a hair above the league-average wOBA of .317. Kemp’s projected performance going forward is about 40% above league-average.
While saying these two have been bad against left-handed pitching so far this year is factually correct, the cause can be attributed to random fluctuations in extremely volatile platoon split numbers. There’s no reason for concern going forward, which is good news for the upcoming series.
|Wed 8/20||Thurs 8/21||Fri 8/22||Sat 8/23||Sun 8/24||Mon 8/25||Tues 8/26|