The Mariners signed Joe Saunders last week to a 1-year $7 million dollar deal.
For his career, Saunders has a 98 ERA – and 108 FIP – (107 xFIP – and 4.66 SIERA). His 72.4 LOB % for his career is about league average. His BABIP is also about average. His OPS on balls in play is very slightly below league average, while his OPS on balls not in play (HR/BB/K) is much worse than league average (thanks to his high HR rate and below average K/BB).
In 2012, Saunders gave up an average of 262.579 feet per batted ball, which isn’t bad, but it isn’t particularly good either. This is slightly off his career averages, as he has given up 259.098 average feet per batted ball since 2007. Compare this to the pitchers he is trying to replace in the Mariners rotation:
Millwood 2012: 246.211
Vargas 2012: 261.519
Saunders throws mainly moving fastballs (or sinkers, depending on which system you ask) and in 2012 averaged just under 90 MPH. Even as a left-hander, this is slightly below average velocity, though it is still harder than the pitcher he is obviously replacing, Jason Vargas.
He may thrown an occasional slider, but it is not a big part of his repertoire, and there is evidence that he doesn’t throw it anymore (more on that below). His main breaking pitches are his changeup and his curveball. He uses them traditionally, with curveballs over a quarter of the time, especially when he is ahead, to lefties (when he has the platoon advantage), and changeups nearly a quarter of the time, especially when behind, against righties (when he doesn’t have the platoon advantage). He very rarely throws the change against lefties. It breaks horizontally like two other pitchers that had been connected to the Mariners in rumors, Daisuke Matsuzaka and Chris Capuano (Saunder’s change is close in velocity to Dice-K’s, but about 4 MPH harder than Capuano’s). Vertically, it breaks like Andy Sonnanstine’s and Josh Outman’s. Sonnanstine’s change is known as a pitch that moves well vertically, which is the comparison for Saunders, but not horizontally. Dice-K didn’t use the change much in Japan, but it became a part of his repertoire in the Majors, especially in 2012. Here is a video of it, focus on the horizontal movement:
The change is below average when it comes to swings and misses, and does a little better job at getting grounders, but is still below average. His 258.156 average feet per batted ball on the pitch is unimpressive. This is most likely why he has large platoon splits. He relies heavily on a below average fastball/sinker and a change that is clearly not getting the job done.
The curveball is better, averaging 249.999 per batted ball. According to whiff/swing %, it isn’t an elite pitch, but it is pretty good (though it isn’t very good at getting grounders comparatively). It isn’t a slow curve, even though his skill set (lefty with a sub 90 MPH fastball) suggests that perhaps he it might be. Sitting around 76 MPH, James McDonald and Felix Doubront have similar velocity curves. Horizontally, the curve moves like Travis Blackley’s and John Lannan’s, while vertically, Mark Hendrickson (pre-sidearm) and Jose Quintana provide the best comparisons. While it was positive in 2012, his career curveball run value is actually in the negative. For his career, he has a .324 BABIP on his curve, but because he doesn’t give up a lot of power with it, it still has a good TAV. What really interested me was the difference between the average horizontal release point of his curveball versus the average horizontal release point of his 4-seam fastball. The average is a .26 feet, or about 3 inches, which doesn’t seem like a lot, but seemed to me somewhat significant when it comes to where a pitcher releases the ball. We have seen elsewhere on this site that pitchers tend to release their curveballs higher than their other pitches, and it doesn’t seem to be a problem. So what about horizontal movement? To test this, I looked at the top 10 whiff/swing curveball pitchers and looked at their horizontal release point for their fastballs and subtracted it from their horizontal release point for their curveballs.
Kyle Drabek: .36
Patrick Corbin: -.17
Mitchell Boggs: -.71
A.J. Burnett: -.06
Billy Buckner: -.08
Roy Halladay: .14
Manny Parra: -.27
David Pauley: .11
Mike Minor: .18
John Smoltz: -.19
So 6 pitchers released their curveballs further out horizontally than their fastballs, and 4 released their fastballs further out. The average of the top 10 released their fastballs about .227 feet different than the curveballs. The difference is smaller, but somewhat insignificant compared to Saunders, as they are extremely similar. So I wanted to look at the worst 10 curves to see if we could get a better idea.
Darrell Rasner: 0
Jeff Weaver: .02
Paul Byrd: .13
Steve Trachsel: -.10
Rodrigo Lopez: -.05
Mike Mussina: -.65
Barry Enright: .06
Blake Beavan: -.16
Wade Leblanc: -.04
Kyle Lohse: -.13
We see here that 6 of the “worst” curveballs are released further out than the fastballs. The average difference was .134, below Saunders average, and below the average of the best 10 whiff/swing curves. This is bizarre, as the more inconsistent release points (though not as inconsistent as Saunders) were the more successful pitchers, or at least have the more successful curves. My guess is that this just isn’t predictive (rather than consistent release points for all pitchers not being a good thing). What about Saunders’ overall release point? We have seen that inconsistent or changed release points can lead to injuries or ineffectiveness, so where does Saunders stand in this regard? When just looking at season by season charts at Brooks Baseball, there seemed to be some inconsistencies. I looked at FanGraphs’ release point data and 2011 and 2012 looked a little different:
While the 2011 chart had more random stray fastballs, he released a lot more moving fastballs (FT) closer to his body, and more curveballs were released higher. It also seemed that, unless I was going crazy, he was releasing more pitches over 6 feet as well. This turns out to be the case, as he released the average pitch 6.006 feet vertically in 2011, and 6.0725 feet vertically in 2012. Of course, this doesn’t seem to be a big difference. Statistically, Saunders was actually better in 2012, so perhaps the change was actually good for Saunders. For his career, Saunders has thrown strikes 61.27 % of the time. In 2011, he threw strikes 61.6 % of the time, and in 2012 he threw strikes 61.9 % of the time. So he seems to be improving in that regard. With all this said, Saunders seems to be pretty inconsistent from pitch to pitch during games when it comes to release point data (it should be noted that even though Saunders is improving in the strike throwing business, he is below average, as MLB starters averaged a 63.6 strike percentage in 2012). Here is Saunders’ last two regular season starts (pretty much all the starts over the past couple years are basically the same, I checked):
Here is Hisashi Iwakuma’s (whose release point and control is somewhat inconsistent, at least according to the eye test) last 2 starts:
For more comparison, here is the data from James Paxton’s outing in the AFL:
Of course, many are questioning whether or not Paxton will remain a starter. To give you comparison of what a good release point in a game looks like, here is a start from one of the premier strike throwers in baseball, and former Mariner, Cliff Lee:
A lot of Saunder’s platoon splits have to do with pulled balls. Right-handers have a much higher OPS on pulled balls against Saunders, while they are actually a little worse at going to the opposite field against Saunders. This is despite the fact that he does a good job of keeping the ball away from right-handed hitters.
This is especially true when it comes to his “off-speed” pitches”:
When you just look at the maps sorted by home runs against right-handers, almost all of them come in the middle/middle section, the low/middle section, or the middle inside section. So while he has a plan of staying arm-side and away from right-handers, when he does leak inside, it hurts him. As we have seen above, with both his strike percentage and release point consistency, he does make mistakes, and doesn’t always have control of where the ball is going (compared to other starters).
According to the Edge % leaderboards, he threw less pitches on the edge of the plate in 2012 than in 2011, and is mediocre at best at doing so anyway. So even though you think of Saunders as a pitchability lefty, he isn’t a great artistic pitcher at hitting the corners. The good news is that he should be able to eat up some innings for the Mariners, as there isn’t much of a reason to think he has a great chance to be injured, and he has been pretty consistent late in games.
Back to the slider Saunders’ has evidently quit throwing. For his career, he threw it almost exclusively to left-handed hitters, but didn’t really have a specific count he threw it in. By 2011, he was throwing the pitch that averaged about 81.6 MPH nearly a 1/3rd of the time against lefties. For his career, he has a 3.91 kwERA against lefties, with a .625 OPS against. In 2011, when he was throwing a ton of sliders against lefties, he had a 3.15 kwERA against lefties, with a .581 OPS. Of course, this was a small sample size and in 2012, he had a 3.03 kwERA against lefties, with a .451 OPS. So he seemed to improve against lefties when he dropped the pitch (he got worse against righties, but that doesn’t seem to have anything to do with the slider). According to pitch values and linear weights, the slider wasn’t a bad pitch, and was better than his curveball. The slider is the pitch he has thrown the least in his career, but according to FanGraphs, it has the best wRC + against out of all his pitches. In velocity, it is a below average slider, and has one of the lowest horizontal movement out of qualified sliders. His vertical movement is about league average for starting pitchers. In whiff/swing and GB %, it was an above average slider, so while there is some bias since he only threw it when he had the platoon advantage, it did not seem to be a bad pitch at all. Again, there isn’t a ton of data when it comes to the pitch, but he usually likes to throw it away from lefties, usually low, which isn’t surprising. Inside Edge actually listed the slider as one of his strengths and noted that he didn’t throw many in the strike zone.
So what has Saunders used to replace the slider? The sinker and curveball. He also seems to be throwing more sinkers instead of fastballs as well. We have already discussed his curveball, which I have my doubts about, but let’s look at his sinker. Obviously Saunder’s sinker velocity is not impressive, but since 2007, it has been just above the median. It moves horizontally like Trevor Cahill’s and former Mariner (and free agent) David Pauley, and is in the top 40 (out of over 300) since the Pitch F/X era began. Pauley’s is a little below Saunders’ in velocity, but we have seen firsthand the amount of grounders it can get when it is on. Vertically, Saunders’ is less impressive, but very similar to the one new Mariner Jeremy Bonderman threw when he was still pitching. Unfortunately, Bonderman’s sinker was absolutely clobbered, giving up a .300 average with a 137 wRC + in his last 3 years. Compared to other sinkers, he gets a lot of swings, not many whiffs, a decent amount of foul balls, and perhaps most importantly, is in the top 100 of GB %. When comparing his sinker locations to other left-handed sinkers, you notice that he is pretty normal other than he locates his sinker to the glove side and low more often than most. He has had problems with it when it stays in the middle inside part of the plate against righties. Predictably, the most often place he has given up homers with his sinker is right down the middle. It isn’t a bad pitch, but it isn’t great either.
Certainly, when one thinks of Joe Saunders, one thinks of a pitchability lefty, with not many strikeouts, not many walks, quite a bit of homers, but a package that makes sense at the back of a rotation. This is not necessarily wrong, but it is misleading. He doesn’t have advanced control, neither the heat maps, the release point data, nor just looking at strike percentage would say this, in fact, it seems to say almost the opposite. If I were the Mariners, I would see if he would be willing to mix in the slider again, because right now, he is a very one dimensional pitcher, throwing a bunch of unimpressive sinkers mixed with some mediocre curveballs. The Mariners desperately need arms, and if everything goes right with Erasmo Ramirez and Hisashi Iwakuma, Saunders could slot in as their 4th best starter, and there wouldn’t even be much pressure on him to be even average. If one or both of the above fail, then Saunders needs to be really good, which, with his current pitch selection, he isn’t.
Since Saunders has such aggressive platoon splits (4.58 kwERA against righties in 2012, not counting giving up homers 3.7 % of the time), the Mariners should be very aggressive in pulling him in high leverage situations when right-handers come up. Especially with the deep bullpen the team has (so deep that they jettisoned Shawn Kelley), there is no reason for him to face righties in close games past the 5th inning.