Out of all the insanity that occurred in the American League this year — the Royals dominating, the Red Sox and Athletics bottoming out, the Twins somehow competing — the Mariners had one of the more quietly disappointing seasons. After a winning season in 2014, Seattle looked like a club on the rise for 2015, one that could contend for a postseason slot. That didn't happen, as several key contributors underwhelmed and the team finished fourth in the AL West. One of those players, second-year starting pitcher Taijuan Walker, did so in a somewhat unusual manner.
Since the Mariners picked Walker in the first round of the 2010 draft, everyone has placed high expectations on him. He placed in the top 20 on all three main prospect lists (Baseball America, Baseball Prospectus, and MLB.com) in 2012, 2013, and 2014; the latter year saw him cruise to a 2.61 ERA in 38.0 innings as a major-leaguer. Heading into 2015, the Mariners hoped he'd give them another solid rotation option behind Felix Hernandez and Hisashi Iwakuma.
As the team came up short, so too did Walker, who worked 169.0 innings of 4.56-ERA ball — a level of play 18 percent worse than average. His peripherals, however, told a different story: Behind a 22.2 percent strikeout rate and 5.7 percent walk rate, he posted a 4.06 FIP, only three percent behind the league mark. Walker didn't fall victim to ball-in-play misfortune, with a solid .291 BABIP. Rather, an unsightly strand rate of 67.8 percent, which only beat six other qualifiers, accounted for his underperformance.
Diving into Walker's extreme situational splits shows us why Walker allowed so many runners to score. With the bases empty, Walker held hitters to a .264 wOBA — the same as Noah Syndergaard. Once men had reached, though, that spiked to .395, around the level of Aaron Harang. Syndergaard and his 3.24 ERA have a solid shot at the NL Rookie of the Year, while Harang's 4.86 ERA could make him retire, so I'd deem this a significant difference.
Let's break this down further, by examining some more splits. The manner in which Walker declined gives a clue as to the cause:
|Men on Base||18.8%||3.9%||.346||5.9%|
Whereas most hurlers will hand out more bases on balls with runners on, Walker managed to cut down in that regard. Simultaneously, hitters struck out less often and inflicted more damage when putting the ball in play.
What does this mean? Over at our sister site Lookout Landing, Brendan Gawlowski noted this trend back in July, and came up with a convincing theory — that "[Walker]'s throwing a lot of bad and hittable strikes with runners on". Based on the evidence here, I'd have to agree with that hypothesis.
Baseball Savant's PITCHf/x Search can give data for any imaginable split, including the base state. Via that tool, we can see how Walker changed his pitch usage with runners on base:
The cutter, curveball, and sinker rates each shift a bit, but not to a significant degree. The important change comes from the two primary pitches in Walker's arsenal: the four-seam fastball and the splitter. Walker used the latter as an out pitch this year, as pitchers often do with a secondary offering that formidable, so on some level it makes sense that he'd increase its usage when in trouble. He's not the only player on the Mariners to change his approach in that manner — Hernandez has always relied on his changeup more often with men on base.
But there's a big difference between Hernandez's changeup and Walker's splitter. The former bests every other cambio in the game, while the latter cost its owner 4.7 runs in 2015. If opposing hitters put up a .301/.331/.409 line against one of your pitches — as they did to Walker's splitter this year — you probably don't want to throw it in the most integral situations. By contrast, the four-seam fastball yielded a .240/.286/.430 triple-slash, which made it worth 8.0 runs. Eschewing the heater for the splitter seems like a good idea in a vacuum; for Walker, though, it led to disaster.
The problems with the splitter go even deeper: It actually got significantly worse with runners on — its batting line inflated from .245/.268/.319 to .385/.420/.538. As Gawlowski posited, Walker left the pitch up over the plate in such scenarios:
We can see the heavier concentration as shown by the redder areas, which clearly expand with the opposition aboard. Walker still placed the splitter in the same general location, but he put it over the plate on far too many occasions, and batters made him pay.
With a split as massive as Walker's — remember, Syndergaard versus Harang — poor luck obviously plays a role. Walker only faced 706 batters total this year, so the sample is pretty minuscule overall, to say nothing of the samples by base state. Still, he does seem to have a different approach based on the circumstances, and that could stick around in future seasons. Some pitchers consistently perform worse with runners on; Walker could join the likes of Ricky Nolasco and Derek Lowe if he doesn't progress.
Back in spring training, the Mariners wanted to see how Walker would react under pressure. He dominated overall in the Cactus League, but that didn't continue into the regular season; looking back now, we can understand what Seattle feared. Under a new manager and general manager, the team will try to bounce back in 2016, and to do so, they'll need Walker to strand a whole lot more runners.
. . .