Ian Snell, Shaun Marcum and their elbow injuries

In the past few weeks, two pitchers were sent to the disabled list with right elbow injuries. One was Shaun Marcum, Toronto’s gem found in the rough; the other was Ian Snell, who is regarded as Pittsburgh’s ace. The injuries were different—Snell had a muscular strain of his elbow and forearm, while Marcum had ligament sprains (tears) about the inner aspect of his elbow. The common factor is the location—the inner elbow.

I was thinking about what could have caused these injuries to similar locations on the elbow. Obviously, the more heavily used a pitcher is, the more likely he is to become injured. Some form of pitch count obviously would alleviate the injury effects to these pitchers, but this is not likely to happen. Will some teams move to a six-man rotation in the near future, trying to shave off hundreds of pitches of wear and tear on the pitchers in a given season? I heard something about the Dodgers perhaps trying this with their staff. I think this is a good idea, especially with some pitchers having two-start weeks during their teams’ six-game weeks.

With all the cool things that are being looked at and discovered with PITCHf/x data, this is certainly something that we should take advantage of. It would be extremely valuable if the data translated to injury susceptibility or injury patterns. I have not worked with PITCHf/x, because (1) I am generally terrible with computer programs, spread sheets and graphs and (2) Josh Kalk and Mike Fast, among others, are doing such a great job in this regard. So before I move forward, I want to thank Josh Kalk for guiding me through the process of PITCHf/x—obtaining the data, understanding it, graphing. He has done a tremendous job. I also want to thank Derek Carty for helping me as well with the data—graphing, the statistical aspects and giving me pointers on what to look for, what to ignore and what to chart.

For all analysis, I used the data from Kalk’s player cards, and the graphs are his as well. I want to make clear that the data set from Fan Graphs and Kalk’s player cards are different, because as Carty informed me, Kalk’s are from PITCHf/x while Fan Graphs is from BIS.

Shaun Marcum, RHP, Toronto Blue Jays
Diagnosis: Right elbow sprain (ligament), currently on 15-day DL

The Statistics

The 2007 and 2008 numbers of Shaun Marcum are remarkably different in many ways, (from FanGraphs):

    	    ERA     WHIP    K/9     K/BB    BB/9    BAA     FIP    GB/FB  BABIP   HR/FB
    2007    4.13    1.25    6.91    2.49    2.77   0.249    4.93    0.95   0.272  13.70%
    2008    2.65    1.00    7.84    3.19    2.46   0.206    3.66    1.17   0.241  10.20%

As you can see, Marcum’s 2008 production is far superior to what he did in 2007. He has shown greater ability to strike batters out, and his walk rate is lower. He is inducing ground balls at a much better rate than in ’07, which also has translated into giving up fewer home runs. Though the .206 BAA is fantastic, this could rise, given his .241 BABIP. So, why is there such a drastic improvement? What is he doing differently in ’08?

In looking at Marcum’s player card, the differences are obvious:
{exp:list_maker}He is throwing his slider nearly three times less than he did in ’07 (5.31 percent vs. 15.01 percent)
He is relying more on his changeup (24 percent vs. 18 percent)
He is throwing his change-up and curveball significantly more to right-handed batters than in ‘07
His curveball is being thrown harder than it was in ’07 (decreasing the batters’ ability to detect the spin of the pitches, even though the movement is less on these pitches in ‘08)
He has changed his release point {/exp:list_maker}

Release Point

Here is a representation of Marcum’s release point in 2007 and 2008. Notice how consistent his release point is for all of his pitches this year. He was more erratic with this in 2007 (though not terribly so). This newfound consistency has likely made it more difficult for batters to distinguish his pitches. Additionally, he has actually raised his release point this season. Usually a pitcher’s release point will become lower with an injury if it is shoulder-related, but since this is an elbow injury, that point may not be valid.

Without looking at slow, frame-by-frame video comparison of the ’07 and ’08 seasons, it would be difficult to see how this higher release point is coming about. It is possible that he is either side-bending his torso more toward first base at release (hence “tilting” his scapula and shoulder more vertically—not necessarily a bad thing if he maintains a good acromial line), or he is simply raising his pitching arm higher in relation to his scapula (less desirable due to possible impingement at the shoulder).

Here is a plot of his 2007 release points:


Notice how the plots are spread out more in a horizontal fashion, and the vertical release points are just below the seven-foot mark, while many are below the six-foot mark.

And here is a plot of his 2008 release points:


In stark contrast, the release is more consistent among his pitches, with less horizontal translation. Another big difference is that almost none of the plots fall below the six-foot mark, indicating a consistently higher release point. Also, you can see that the curveballs (blue) are being released higher than they were in ’07. With a more consistent release point, Marcum is less likely to tip his pitches to the batter.

Curveball velocity and movement

Even though he is throwing his curveball with a greater average velocity and less horizontal movement, he is getting batters out with greater ease than in ’07, perhaps because the hitters are less able to distinguish his pitches. Of course, some of this can be chalked up to some luck given his low BABIP, but it is interesting nonetheless. Without looking at video to compare his mechanics, it is hard to say if this apparent increase of velocity with off-speed pitches is due to mechanics, or by design on the part of Marcum.

Here is the horizontal movement versus velocity graphs for 2007:


And here is the same graph, but for 2008:


By comparing the two graphs, you can easily see how he is throwing his curveball (blue) with greater velocity and less horizontal movement than he was in 2007. Remember, he also is throwing all his pitches from a higher, more consistent release point this season, making them more difficult for batters to identify.

Two-strike counts

Marcum also is relying less on fastballs and sliders in two-strike counts this year, but a marked increase in curveball usage also may indicate that he has increased confidence in his control of it. Pretty much across the board, he is throwing more curveballs in two-strike counts than he did in ’07. Anyone who has pitched knows that in two-strike counts, the pitcher smells blood and tends to throw with more intent (and hence, velocity), which could lead to increased strain on the arm.

 Count   FB 07   FB 08   CU 07   CU 08   SL 07   SL 08
 0-2   26.53%  21.31%  18.37%  14.75%  18.37%  8.20%
 1-2   20.59%  27.38%  9.56%   26.19%  25.74%  4.76%
 2-2   28.57%  24.68%  11.04%  16.88%  14.94%  6.49%
 3-2   44.12%  40.91%  1.96%   4.55%   5.88%   2.27%

Overall, it looks like Marcum is changing the way he pitches. He is attacking the batter more often with hard curveballs that have slightly faster velocity but less horizontal movement, while his fastball actually has slowed some. This leads me to believe that he is perhaps using more intent with his pitches this season (hard to say without seeing 2007 video). What I do know is that when I watch Marcum throw on MLB.tv this season, he seems to “whip” the forearm as he releases, with a fairly still lower body, and a more active upper body and pitching arm.

His entire approach has been different this season: He is not necessarily abandoning his fastball (which was not too fast to begin with), but he is throwing it less often. Perhaps the most interesting trend is that according to Kalk’s player cards, he is using his slider nearly three times less frequently than he did in 2007. Perhaps he can’t locate it; perhaps he can’t throw it without hurting—who knows? Maybe he is worried that he was throwing the slider too much last year and throwing it less often in 2008 by design. We just cannot know for sure.

Ian Snell, RHP, Pittsburgh Pirates
Diagnosis: Right elbow/forearm strain (muscle); 15-day DL

The statistics

While Marcum saw his production increase this season before his injury, Snell has seen his production get drastically worse:

            ERA     WHIP    K/9     K/BB    BB/9    BAA     FIP     GB/FB   BABIP   HR/FB
    2007    3.76    1.33    7.66     2.6    2.94   0.263    4.01    1.23   0.313   9.60%
    2008    5.99    1.93    6.72    1.31    5.15   0.324    4.33    1.17    0.38   6.80%

It is apparent that Snell is getting hit around, but some of this can be attributed to some bad luck, as his .380 BABIP indicates. Despite that, his control has been awful: He is walking batters at a significantly higher rate this season. His strikeout rates have dropped, and as a result he has been getting tagged for a lot more runs. If he is healthy, his numbers should improve when he returns, but I would not expect an ERA under 4.50 or a WHIP under 1.50 for the season.

Release point

His release point has been less concise than it was in ’07, especially when looking at his horizontal release point. Below is his 2007 release point. Notice how tight the plots are.


Now look at the 2008 release points, and see how spread out they are, especially with the horizontal aspect:


Reliance on the fastball

Snell is using his fastball much more this season, especially on 0-2 and 3-2 counts, possibly indicating that he is trying to use it as his “out” pitch more often. On 1-2 counts, his usage pattern seems pretty much the same, and on 2-2 he is using it less (he almost exclusively prefers to throw sliders on 2-2).

Fastball  0-2     1-2     2-2     3-2
  2007   29.31%  31.94%  32.08%  42.42%
  2008   53.73%  31.03%  22.78%  58.00%
So, what is going on with Snell?

It is hard to say, but you could conclude that his haphazard release point is to blame. More difficult to figure out is why the release point is changing.

Typically, a pitcher lowers his release point in response to an injury—usually a shoulder or rotator cuff injury. This could be conscious or subconscious. Our bodies can develop abnormal movement patterns in response to pain or dysfunction. This could be the case with Snell, but there is no way to tell definitively. I have not seen any news of him having shoulder problems. He also has demonstrated decreased velocity with all his pitches, especially his fastball, further evidence that an injury could be present or developing.

What could all this mean?

Marcum and Snell are two entirely different pitchers. They have different injuries to the same anatomical region (right medial elbow), and their pitching presentation appears to be dissimilar, especially when comparing efficacy and approach. What I wanted to know through all of this was what, if any, commonalities exist between them that could have led to these injuries, or could have been a result of their injuries.

Similarities between Marcum and Snell

The pitchers displayed the following similar characteristics:
{exp:list_maker}An altered release point from 2007
Decreased velocity on their fastball
Decreased use of slider
Increased use of changeup
Increasing workloads dating back three seasons {/exp:list_maker}

There is no way to be certain if the injuries were caused by these similarities, simply by looking at the stats. A game-by-game account of the respective data sets might show a better picture of what (if any) changes were going on in their release points, velocity or pitch selection. This could be looked at in more depth as I familiarize myself with PITCHf/x. Possibly, each pitcher could have been hurt regardless of any of these data or any of these trends.

Other factors are involved here that we as writers cannot gauge, such as the pitcher’s overall health, conditioning, body structure, range of motion, joint mobility, strength, flexibility, nutrition, stress, pitching/usage history dating back to youth, etc. Perhaps this is where someone in the medical field—a doctor, physical therapist or athletic trainer—who is well-versed in baseball, mechanics and PITCHf/x could come into play. Some also may question whether certain players are taking steroids or other performance-enhancing drugs.

With PITCHf/x, we are at least able to start looking at baseball injuries from a fresh perspective, particularly on the pitching side of the game. Certainly, much more stringent means of data analysis is needed, probably on a game-by-game basis. I would go as far as having a spring training and offseason PITCHf/x data set for each pitcher so that we can truly see a full spectrum of information across the entire year.

Print Friendly
« Previous: The 2008 Von Hayes All-Stars
Next: Beyond Moneyball: Player development, Part 5 »

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *