When a pitcher like Justin Verlander has a start like he did on Saturday against the Cleveland -- Verlander allowed four runs (three earned) on six hits and five walks in five innings -- there are going to be questions about what went wrong. It's easy to chalk one start up to any number of factors. It was cold on Saturday. Verlander didn't have his fastball and, contrary to popular opinion, is actually human. The Indians have a good offense.
That said, there have been concerns about Verlander's velocity all season long. He has shown us a fastball in the high 90s on occasion, but hasn't sustained his velocity like in previous seasons. It was hard to argue with his approach, given that he got off to the best start of his career. Even after last night's game, Verlander is still 4-3 with a 1.93 ERA and a strike rate of 10 batters per nine innings.
His full season numbers show that he is walking 3.2 batters per nine innings, but that's largely due to last night's walk-a-thon. In his first seven starts, Verlander was walking a perfectly acceptable 2.5 batters per nine innings.
One of the prevailing thoughts about Verlander's drop in velocity is that it is causing opposing teams to foul off more pitches, especially with two strikes. According to PitchFX, this isn't true. Opposing hitters are fouling off a near-identical percentage of pitches with two strikes this year compared to his first eight starts of 2012.
The top graph is his pitch chart with two strikes for his first eight starts of 2012, while the bottom chart is 2013. One thing I noticed at first glance isn't the number of foul balls, but rather the number of pitches outside the strike zone. With two strikes, Verlander is throwing 33 percent of his pitches for balls this year compared to just 27 percent at this point last season. Even if we take last night's start out of the equation, the percentages hardly budge.
While this could be due to the small samples that we are looking at -- I would have pulled the entire 2012 season, but the graphs were too cluttered -- I decided to take a closer look at the rest of the PitchFX data. The verdict? Something is up.
Once again, the top graph is 2012 and the bottom is 2013. As you see, Verlander's fastball had a lot more running action to it in 2012 (the further away from the midline the points are, the more horizontal movement the pitch has). While the velocity might not mean much in the grand scheme of things, the movement (or lack of) is a problem. His fastball has been only 0.61 runs above average per 100 pitches this year compared to 0.83 last year, according to PitchFX.
As for his other pitches? Verlander's slider (the middle cluster) looks to be consistent in both seasons. His changeup is also lacking some fading action, but this hasn't resulted in a drop-off in 2013. In 2012, Verlander's changeup was only 0.44 runs above average per 100 pitches after sitting at or above 1.40 runs above average for the previous three seasons. In 2013, it looks like his changeup is back to this pre-2012 level despite the lack of run. Given that the success of the pitch depends more on arm action and setup (i.e. the hitter is guessing fastball and doesn't get one) than movement.
However, his curveball also doesn't look right. Sure, he has broken off a few beauties in 2013, but the big 12-6 curveball that we know and love hasn't been the same in 2013. Part of that may be due to the weather he has pitched in. Part of that may be due to the thumb issue he suffered against the Kansas City Royals earlier this season. Part of it may be due to the same workload that kept him out of the World Baseball Classic. This makes sense, given that most pitchers say that their breaking ball is the last thing they are able to command during Spring Training. Regardless, the lack of consistent movement on the curveball is showing in the numbers. In 2012, the pitch was 2.04 runs above average per 100 pitches. This year, that number is down to a paltry 0.17.
Now that we know that Verlander is throwing a fastball and curveball that are both well below his normal standards, the data from the first graphs makes more sense. It's possible that Verlander is trying to get hitters to chase more often, knowing that his fastball and curveball aren't as effective as they have been in years past. It's also possible that Verlander has no idea, or that this is a product of a small sample size. Whatever the root of this drop-off may be, I hope he can solve it soon before we see more outings like we did on Saturday.