There's always a discussion of whether hitting is more of an "art" or more of a "science." I haven't heard the same discussion around pitching nearly as often, however, and while pitchers have their mechanics, they seem to be considered more "artful" in general. When you see a pitcher in his groove, it really is like watching a hairy, sweaty, muscular ballerina perform poetry in motion.
So while I'd contend that pitching is a artform all it's own, it's important that we acknowledge that not all art it good art. Stuart Wallace has spent considerable time showing us the best pitches of baseball over the winter, but I'd like to shine the light on the other end of the spectrum. So the question must be asked: what were the worst pitches in baseball in 2013?
To find said pitches, I looked at starters who threw over 120 innings last season. Yes, relievers can throw some terrible pitches, but given their limited time on the mound, the impact can only be so big. Instead, I looked for guys who threw roughly a half a season's worth of pitches. Next, I threw out any pitches that weren't thrown at least 10% of the time by a given pitcher. PITCHf/x can misread pitches from time to time and I felt like it wasn't fair to highlight a pitcher who was just "experimenting" with a new pitch and handful of times over the course of the year. Instead, I sought out pitches that were a deliberate part of a given pitcher's repertoire.
The last bit of filtering I did sought to answer another question: how much better could these pitchers have been if they had quit throwing this terrible pitch? Because starters generally need three functional pitches to remain in a rotation, I only selected pitchers who threw at least four different pitch types at a minimum frequency of 10% each. If we want this pitcher to eliminate his "bad" pitch(es) altogether, he's going to need a remaining three pitches to get by as a starter, at least according to conventional wisdom. This left me with 74 pitchers who met the above criteria in 2013.
With these 74 pitchers in mind, I identified poor pitches by looking at the pitch type linear weights for each pitch they threw at least 10% of the time. Rather than looking at each pitch's total value over the course of the season, pitch type linear weights essentially tells us how many runs better or worse than average each pitch was per 100 pitches thrown. You can read more about it here at FanGraphs, but it's like grading a hitter's strikeouts with K% (preferred) rather than total number of strikeouts. In sabermetrics, we like things normalized and pitch type linear weights gives us the pitch value on a normalized basis for comparison.
So who threw these terrible pitches? Take a look at the table below to see who had the worst pitches of 2013.
|Pitcher||Pitch||Frequency||Run Value per 100 Pitches||Volatility Factor|
|James Shields||Two-Seam FB||11.8%||-2.23||26.31|
|Dylan Axelrod||Two-Seam FB||12.8%||-2.23||28.54|
|Kevin Correia||Four-Seam FB||18.2%||-2.07||37.67|
|Jason Vargas||Two-Seam FB||19.6%||-2.11||41.36|
|John Danks||Four-Seam FB||27.0%||-2.17||58.59|
|Dylan Axelrod||Four-Seam FB||34.1%||-2.83||96.50|
*Note: the "Volatility Factor" is a blend of the pitch value and frequency with which it was thrown, calculated as: (Run Value per 100 Pitches*Frequency Thrown)*-100
After glancing at the table, you can clearly see that all of these pitches were horrendous in 2013. So why did these pitchers continue to throw these pitches? Let's discuss that aspect.
The one name above that probably surprised you is James Shields. He did not fare well when using his two-seamer last year, but he didn't use it all that often and more than made up for it with his other pitches. Could he get by without using that fastball? Probably, but it may throw off some sequencing. My gut tells me that James Shields is the kind of pitcher who could eliminate a bad pitch and be just fine, though.
Some of the others names above, not so much. Jason Vargas, for example, got lit up when using his two-seamer and while it would be easy to say that he should just scrap it in favor of his four-seam fastball, that's not easy to do considering that his four-seamer was nearly as bad. The other thought is that he relies heavily on his changeup, which is actually a pretty good pitch. His two-seamer has similar action to the change but with a little more velocity. He needs the fastballs to "speed up" the bats of hitters so he can get to his changeup, the best weapon in his arsenal. Therefore, the two-seamer has a purpose.
Kevin Correia is in a similar situation where he uses a poor four-seam fastball to set up his above average slider. The difference with him is that his two-seam fastball graded out relatively well, but the movement of the pitch likely doesn't set up the slider as well as the four-seamer does. Like Vargas, to get to his best pitch, he has to use his worst one. At least there's a logical reason to use his worst pitch.
When we get to names like Joe Saunders, Jordan Lyles, Dylan Axlerod and the recently unemployed (but not retired) Barry Zito, there's a lot of head scratching to do as the vast majority of their pitches grade out far below average. These guys simply lack the "stuff" to fool hitters across the board. In the case of Saunders, Lyles, and Axlerod, they were back-end starters on staffs that were merely trying to fill a void. Zito was just finishing out a very lucrative contract and had to pitch to justify the salary he was receiving. These guys threw their worst pitches because they're not very good but had to throw something up there to keep their rotation spot.
Jeremy Hefner is an interesting case. His changeup was the worst per 100 pitches of any of the starters identified. The rest of his stuff isn't necessarily inspiring, but it's not atrocious. It would be easy to say that he should just scrap the changeup, but looking at his 2012 numbers, his change was his best pitch. Perhaps it suddenly became his worst as he tried to pitch through an injury that ultimately led to Tommy John surgery. We'll have to wait and see if and how he returns before we can say that the injury led to the decline in this particular pitch. Oddly, his slider and curve were better in 2013 than in 2012, so it's not like his entire repertoire suffered immensely.
Looking at the entire group, there's a clear lack of power arms. Velocity isn't everything, but it certainly helps and only two of the pitchers above have fastballs that average more than 91mph (Shields and Lyles). There's also an inability to miss bats, which I guess is how they ended up on this list in the first place. Shields, unsurprisingly, has the lowest contact rate and is the only pitcher on this list with a below average one. Everyone else is above the median for qualified starters, and it's not even close for Lyles, Saunders and Correia as they get hit and hit a lot. So while we may not expect James Shields to be on this list again next winter, the same can't be said for the others here.
Pitching is an art form and a complex one at that. While we've identified some terrible pitches of 2013, they can't necessarily be eliminated from each pitcher's repertoire. Sequencing and a team's game plan against individual hitters likely has a large impact here. How much? We can't really know without conducting a bunch of interviews, but my guess is that if these guys could just throw away these pitches and be better off, they'd have done it before now. Still, there's always room to improve and I'll be tracking to see if these pitches improve in 2014.
. . .
Jeff Wiser is a contributor to Beyond the Box Score and co-author of Inside the 'Zona, an analytical look at the Arizona Diamondbacks. He occasionally blogs about craft beer at BeerGraphs and you can follow him on Twitter @OutfieldGrass24.