How left/right splits determine value

Josh Donaldson is struggling against righties this year, but he may be due for an improvement soon. Kyle Terada/USA TODAY Sports

Josh Donaldson's recent struggles have been difficult to ignore. He's a .174/.211/.290 hitter in his past 35 games.

Much of that has to do with his struggles against right-handed pitching, a skills concern many of his owners have shared. A .285/.371/.442 hitter against righties in 2013, his resulting .358 weighted on-base average ranking 16th out of 75 qualified right-handed hitters versus righties, Donaldson has slumped to .231/.296/.386 triple-slash rates and a .299 wOBA against his "weak side" this season. He has just 19 hits versus righties in his past 110 at-bats.

As a 28-year-old, in-the-heart-of-his-prime slugger, Donaldson's regression is somewhat puzzling. One might think he would have had a couple more seasons of statistics like his 2013 before declining, but perhaps that's a mistaken assumption. Is there something to radical shifts in righty-lefty splits that portend bad things going forward? Conversely, when a player steps up his game against his weak side -- as Anthony Rizzo has done this year -- should we assume it a fluke or the player adding a new skill?

Let's use history, specifically the past 10 seasons (2004-13) for our study, and players' wOBA splits only. We're looking for left-handed hitters with at least 50 plate appearances against lefties or right-handed hitters with at least 125 PAs against righties in a given season whose wOBA against that side either increased or decreased by at least 20 points and did so while improving a wOBA that was beneath the typical league average of .315 in one year to above that number the next (or vice versa, going from above to below it). What did these players do in their follow-up seasons?

It turns out that a player's age has plenty to say about his improvement/regression prospects and that there's plenty of annual fluctuation in righty-lefty splits overall. That's no shock; selecting splits does narrow the sample size to the point that inconsistent results are expected. Here are the follow-up results of the players in this study:

Ages 20-24: 24% continued to improve, 34% maintained their gains, 23% regressed or actually declined comparative to pre-breakout year
Ages 25-27: 30% continued, 26% maintained, 21% regressed
Ages 28-30: 20% continued, 28% maintained, 27% regressed
Ages 31-plus: 25% continued, 35% maintained, 22% regressed

Splitting up the groups into lefties and righties showed a more marked trend, as left-handed hitters appear to have a wider development curve. Take a look:

Left-handed hitters versus left-handed pitchers:
Ages 20-24: 27% continued, 38% maintained, 19% regressed
Ages 25-27: 36% continued, 20% maintained, 20% regressed
Ages 28-30: 17% continued, 21% maintained, 38% regressed
Ages 31-plus: 21% continued, 32% maintained, 20% regressed

Right-handed hitters versus right-handed pitchers:
Ages 20-24: 22% continued, 30% maintained, 26% regressed
Ages 25-27: 26% continued, 30% maintained, 22% regressed
Ages 28-30: 22% continued, 31% maintained, 22% regressed
Ages 31-plus: 28% continued, 37% maintained, 23% regressed

Again, none of these percentages reveal extremely great odds, which is why each individual player warrants his own examination. Still, there's enough of a difference there to suggest that left-handed hitters who show growth against their weak side at a young age have the best odds of maintaining it over several seasons, while right-handed hitters tend to be pretty consistent from year to year.

These are the left-handed hitters who have exhibited the greatest increase in wOBA against left-handed pitchers this season (minimum 40 PAs):

Brandon Crawford: +.153, age 27
Anthony Rizzo: +.143, age 24
Josh Hamilton: +.136, age 33
Bryce Harper: +.116, age 21
Christian Yelich: +.112, age 22
Jon Jay: +.110, age 29
Chris Parmelee: +.107, age 26
Brian McCann: +.106, age 30
Brandon Moss: +.101, age 30
Jacoby Ellsbury: +.086, age 30

Now, the lefties who have declined in wOBA versus lefties the most:

Alejandro De Aza: -.218, age 30
Jason Heyward: -.141, age 24
Joe Mauer: -.137, age 31
Will Venable: -.128, age 31
A.J. Pierzynski: -.108, age 37
Adrian Gonzalez: -.104, age 32
Jason Kipnis: -.102, age 27
Josh Reddick: -.097, age 27
Carlos Gonzalez: -.091, age 28
Brad Miller: -.087, age 24

Crawford, Rizzo, Harper and Yelich are the most notable names on the former list, as younger lefties with room yet to grow. Rizzo, Harper and Yelich have long been hailed as future stars; Harper's advances against his weaker side are a major plus in a disappointing season for him after he struggled against them earlier in his career. But it's Crawford whose inclusion is curious, and it shows a clear skills bump, as does his career-high 11.0 percent walk rate. Normally known as a "glove man," Crawford might yet be a handy middle infielder in NL-only and deeper mixed (12-plus teams) leagues the rest of the year.

Though they aren't as likely as the lefties -- at least if history is any indication -- to keep up their success going forward, these are the right-handed hitters who have exhibited the greatest increase in wOBA against right-handed pitchers this season (minimum 75 PAs):

J.D. Martinez: +.168, age 26
Devin Mesoraco: +.164, age 26
Derek Norris: +.130, age 25
A.J. Pollock: +.108, age 26
Brandon Barnes: +.094, age 28
Ruben Tejada: +.083, age 24
Michael Morse: +.082, age 32
Kurt Suzuki: +.067, age 30
Justin Ruggiano: +.063, age 32
Russell Martin: +.062, age 31

And the righties who have declined in wOBA versus righties the most:

Jonny Gomes: -.098, age 33
Allen Craig:-.097, age 30
Jedd Gyorko: -.089, age 25
Michael Cuddyer: -.088, age 35
Carlos Quentin: -.085, age 31
Jose Molina: -.079, age 39
Dan Uggla: -.079, age 34
Mark Ellis: -.076, age 37
Alfonso Soriano: -.072, age 38
Wilin Rosario: -.072, age 25

Donaldson (minus-.056, age 28), interestingly, is only 19th on the latter list, though he's sixth if you only include players who have yet to turn 30. That right-handed hitters in his age group tended to be least likely to continue their downward trends at least is a positive; Donaldson also has a .259 BABIP against righties (his career number against them is .299) that offers additional hope. This could perhaps represent the buy-low point on his 2014 season curve.