John Maine, Shoulder Fatigue, and His Release Point
By Bloomberg Sports //
Here are the facts: John Maine threw five pitches to Nyjer Morgan in his last start for the Mets and was pulled. Maine was then placed on the disabled list with chronic shoulder fatigue.
Which leads to the question: is it possible that Maine’s shoulder fatigue showed up in the pitching data? With the Bloomberg Sports’ Pro tool we decided to try and find out, focusing specifically on his release point.
The following graphic shows John Maine’s release points for the Morgan at-bat, along with the corresponding strike zone.
Quadrant A is of particular note – that’s the spot from which two of his pitches originated. The left side of the graphic shows the release points, while the right side is the strike zone. Highlighting a particular zone results in those specific pitches being highlighted in the strike zone.
Both pitches were balls, and neither reached 84 miles per hour. Or to put it another way, Maine released 40% of his pitches from Quadrant A, and 100% were balls. The three other pitches Maine threw were right on the edge of Quadrant A.
Five pitches is an insignificant sample size, so let’s look at all the pitches Maine has thrown this season and how many originated from Quadrant A.
Maine has thrown 768 total pitches, of which 31 came from Quadrant A. You can also see that Maine throws the majority of his pitches from the quadrant directly below A.
The following table provides a pitch results breakdown, including how many of the pitches were greater than or equal to 85 miles per hour.
What we see is that 4% of his pitches came from Quadrant A, more than 50% were balls, and just about half were greater than or equal to 85 miles per hour.
To get a clearer picture, here are John Maine’s pitches from 2008-2010 in an aggregate form.
The number in Quadrant A is a little difficult to read – it says 103.
The following table shows how many of Maine’s pitches over the years were thrown from Quadrant A, along with the results.
The data provide a clearer picture of what happened, including:
1) This season, Maine has already thrown nearly 1/3 of the total pitches he has ever
thrown from Quadrant A (103 for his career, with 31 this season)
2) Percentage wise, Maine has nearly doubled the number of pitches released from
Quadrant A (4% versus 2.21)
3) In this particular game, the velocity from the release point in Quadrant A was lower
than usual (41.74% for his career versus 0 for the game)
What exactly is wrong with Maine is best left up to the medical experts, and the data do not point directly to shoulder fatigue. What is evident is that this season, statistically speaking, he was releasing the ball from a higher point with more frequency, and was pitching differently.
Bloomberg Sports 
A few comments:
1. This spread does not seperate Maine’s fastball from offspeed pitches. It would have been interesting to see if his velocity was affected seperated by pitch type. For instance, rather than measuring what percentage of overall pitches were over 85mph, maybe what percentage of fastballs were over 85 mph?
2. Maine had injuries during each of the previous three years, and thus it might have been useful to compare the 2008-2010 results with 2007, his most illustrious and healthy year, rather than comparing numbers across three injury years. It is difficult to make claims about whether shoulder fatigue affected him in comparison to 2009 and 2008 when he had shoulder issues all three years.
3. If we are trying to figure out what happens when Maine releases the ball from quadrant A, it would be useful to see the pitch types released from that quadrant. If, for instance, he only releases his curveball from that slot, then it might be difficult to make any assertions. However, if I remember correctly, the Nyjer Morgan incident was only fastballs, so maybe the chart should have just included release point for fastballs?
Thanks for the comments. Always appreciated.
1. I chose an overall percentage for speed as the data only shows the result of a pitch type – not the intent. If Maine thought he was throwing a fastball and it was released at 80 miles an hour then the system most likely shows that as a changeup. The data doesn’t care what Maine thought he was throwing – it cares what he threw.
To answer the inevitable follow up question, the data it taggede to distinguish that different pitchers have different fastball speeds. R.A. Dickey has a fastball that shows up at 84 mph.
2. Good point. I’ll add that into the tables and charts in the near future. I was using three seasons as a starting point, though adding the fourth will make the sample that much stronger.
3. The intention was to show he was releasing more pitches from Quadrant A. However here is the requested data of pitch type for from Quadrant A:
2010 – Fastballs 14/31 Sliders 8/31 Changes ups 7/31 Curves 2/31 Total 31/31
2009 – Fastballs 3/12 Sliders 6/12 Changes up 1/12 Curves 2/12 Total 12/12
2008 – Fastballs 16/60 Sliders 16/60 Change ups 16/60 Curves 12/60 Total 60/60
2008-2010 – Fastballs 33/103 Sliders 40/103 Change ups 24/103 Curves 18/103 Total 103/103
In Maine v Morgan
From Quardant A he threw two change ups (from just below Quadrant A he threw three fastballs).
Maine very well may have intended to throw five fastballs. Though again, I remind people of intent v actual execution.