Goooood Morning VEB!
Continuing the recent trend of lurkers introducing themselves via the fanpost, I decided to put together my own little piece and share something I thought to be interesting about Fernando Salas, in hopes of contributing to the high quality content and discussion I have been enjoying to read, and read only, on this site for quite a while now. Since Salas just got called up again, this seemed like an opportune time to do it. I don't want to get too much into his earlier 2012 performance and results, but a quick recap to provide some context nonetheless:
During his first stint in 2012, Salas had assembled a 6.32 ERA and earned himself a subsequent demotion to AAA, which was not at all based on SSS-results, but clearly because the Cardinals saw something in his mechanics that they wanted him to correct at a place that is not the major leagues. Or so I tell myself. His K/9 (12.06) and BB/9 (5.17) were up, but for the most part due to a higher number of batters faced per inning. His K% was 26.6%, slightly higher than last years 25.4% and way above league average for relievers (22%). Salas' BB% stood at 11.4%, which was clearly worse than his own 2011 mark (7.1%). However, in this grand sample, it would have taken exactly two less walks to push his rate below league average (8.9% vs. 9.7%). The obvious glitch in his stats comes courtesy of a .458 BABIP. To mention that a .458 BABIP is unsustainable feels redundant, but here ya go: A BABIP of .458 is unsustainable, even when you're allowing line drives at a rate of nearly 30%.
On a side note: Are BIP categorized by means of an objective measurement or the result of a subjective classification? Out of 47 balls that have been put into the field of play against Salas this year, Fangraphs calls 18 a "Fliner". 12 of those then go on to be categorized as line drives in his BIP data. Again, it would have taken 4 of those to be classified as flyballs to give Salas a GB/FB/LD line of 34%/44.7%/21.3% (career: 33.9%/48.6%/17.4%). I realize this is cherry picking, but I guess this is also what reliever volatility looks like when you break it down to single events.
After moving from results to performance, I felt like I have to go deeper (/obligatory inception pun) and take a look at Salas' PITCHf/x data to reduce meaningless SSS even further: (I also promise that by the end of this, I will have managed to work my way down to two individual pitches.) The differences between '11 and '12 Salas is minimal at best. His fastball and changeup are nearly identical to last year in terms of velocity, movement, strike% and whiff%, with the changeup yielding slightly better results (strike%/whiff%) than last year. His curveball, while thrown at the same speed, has more spin to it this year (1206rpm/826rpm), resulting in more movement (5.31/3.33 horizontal, -44.65/-41.37 vertical) and, counterintuitively, more flyballs as well as a lower strike%. Then again, it's a sample of 65 pitches. The pitch selection shows slightly less fastballs, more changeups and breaking balls, which could be a shift in approach. Or it could be the result of individual matchups and game situations necessitating more off-speed pitches. Really, not much to see here. Except for that one little thing which was supposed to be the actual subject of all this (sorry for the oversized foreplay).
This is officially the point where this post seizes to be about pre-2012 Salas vs. 2012 Salas and concerns itself with something more general about his pitching: Fernando Salas' fourth pitch, which I intentionally embezzled in the last paragraph. If memory, a quick google search and this trustworthy looking site serves me correctly, Fernando Salas does not possess a fourth pitch. However, PITCHf/x tells me that he also throws a sinker/two-seam fastball (FT). I'm aware that pitches are occasionally misclassified, or else I'd have to believe that out of more than 10.000 career pitches, Adam Wainwright has thrown exactly 2 sliders on purpose. The thing about Salas' FT is the fact that the he is delivering it quite frequently. To be specific, about 4% (brooksbaseball), 6,5% (fangraphs) or 6.8% (texasleaguers). My presumption is that Salas is not throwing this thing on intent, but rather as a result of his fastball getting away from him at times. That proposition would be supported by
a) the absence of any report (I could find) about Salas mentioning a FT
b) him consistently throwing the FT only once in every twenty pitches (which would be uncharacteristic for any pitcher)
c) the fact that Yadi is putting down the same sign before either the FT or FA
d) looking at the following graph:
I personally believe it to be obvious that Salas is not throwing two different fastballs, but that the classification of FT is the result of the large spread in movement with his regular fastball. On average, those fastballs classified as FT move as much horizontally (-8.08) as his changeup (-7.76), nearly twice as much as his FA (-4.28) and have a little bit more downward movement as well (FT 8.37/FA 10.17). In a vacuum, the pitch itself wouldn't even be considered bad (poor choice of words, it would actually be quite impressive if he could do that to a baseball in a vacuum). In fact, in regards to movement, velocity, spin axis and rotation, it would almost be identical to Adam Wainwright's SI/FT.
The problem with throwing a pitch that moves without intending to, is the likelihood of missing your spot with it. The results of those pitches classified as FT would back this thesis as well: He misses the zone with it more often than with his other fastballs (54.9%/67.5% texasl | 55.81%/66.85% brooks) and generates less swings and misses (3.9%/9.2% texasl | 2.33%/10.17% brooks). Even when accounting for the lower strike%, the FT induces a whiff on only 10.53% of all swings against it vs. 21.44% for the FA (brooks, couldn't find this on texasleaguers). A look at two examples of said pitches, which I obviously selected completely biased and solely for the purpose of proving my point, illustrates this argument. Both pitches were thrown in his outing against the Pirates on 5/1/2012. I decided to use this particular game because it contains a couple of pitches that were considered FT (4 brooks, 3 texasl) and because it was played at Busch Stadium, featuring the (awesome) camera angle from straight away CF.
First pitch fastball to Andrew McCutchen
1-0 pitch to Neil Walker
The pitch to McCutchen illustrates what I believe to be the reason why Salas seemed to have "gotten away" with pitches in the past. His regular fastball has what some broadcasters would call "giddy-up", "rise" etc. Although not thrown at an elite velocity, it does not move as much, especially on the vertical plane, as your brain would expect it to as a batter. In fact, the movement of Salas' straight fastball (10.17v/-4.28h), or rather its lack thereof, is much closer to Jason Motte's (10.12v/-4.02h) than, say, Chris Carpenter's (6.31v/-9.08h), which would explain the high percentage of swings and misses he generates with it (again, Salas 9.2%/Motte 11.8%/Carp 3.9%). His FT on the other hand, moves more (8.37h/-8.08v) and induced less whiffs (3.9%). As mentioned, throwing a pitch without meaning to generate this kind of movement could lead you to either miss the zone more often when trying to hit the inside corner (to a RHB), exemplified by the pitch to Walker, or out over the plate when targeting the opposite corner. This in turn could explain the lower percentage of strikes thrown, as well as the higher rate of contact. Although to be fair, a 3,9% whiff rate on a FT is not out of the ordinary. However, my point remains that he is not throwing a FT on purpose, but rather that his fastball is simply getting away from him more often than you'd like to see, resulting in a much less effective pitch. The mechanical reason for this could be a drop in arm angle or a slight tilt of his hand, which would explain why the spin axis of his FT (224) is somewhere between his regular fastball (203) and his changeup (243).
I wrote most of this before Salas got called back up and before he pitched in yesterdays game. But since I crossed the line from statistically meaningful sample to pure subjective observation a long, long time ago, I decided I might as well take a look at those 13 pitches. This is an overlay of his release point yesterday (front) vs. the above cited outing against PIT (back):
His release point has changed a little bit, while I couldn't make out any difference in his positioning on the rubber, indicating that he is throwing more over the top now (or at least he did yesterday). This shows itself in the behavior of his pitches as well. Same drill, horizontal vs. vertical movement:
The movement on those fastballs is even less than on his FA before, although the same holds true for his changeup, which I'm not sure is a good thing, but would be the logical result of him throwing less from the side and more over the top. A look at the spin axis indicates the same.
What's the point of all his? I don't know. Hopefully it's you now knowing a little bit more about Fernando Salas than you did 10 minutes ago. Or it's just me hoping the organization detected something similar in his pitching and had him work on on it, resulting in an all improved, even more awesome Salas v2.0, throwing lots of non-moving fastballs by opposing batters. Again, this isn't about Salas' result-slump of early 2012, but rather something that has been going on for the last three years. It will at least be interesting to see if yesterday was just an outlier or the start of a trend.
Other than that, I now leave you with the freedom to discuss, agree with, debunk or build on this theory. Also, Matt Hollilegobobbleheadday.
/Back to Lurk's Landing.