I was supposed to be writing about the Mariners when I started writing this, but frankly, writing about the Mariners can be boring sometimes (look, Justin Smoak just chased a breaking ball) and I got distracted. Instead I started wondering about the sudden fall of Tim Lincecum and Jair Jurrjens in relation to the absolute Mariner free agent disasters of Hong-Chih Kuo and George Sherrill.
When browsing Brooks Baseball (an invaluable tool if you haven’t figured it out already), I stumbled upon (I also discovered something similar when I wrote an article on King Felix Hernandez) something that seemed to be a problem that plagued them all, a change in release point. For most pitching coaches and baseball experts, this isn’t a drastic or novel discovery. I haven’t invented the moon or swiss cheese (or a moon full of swiss cheese), but it is of ample importance for fans, fantasy owners, and real GMs.
First, we can look at Joe Nathan, a guy coming off Tommy John surgery that seems to have regressed. Here was Joe Nathan’s highest win probability game in 2008, when Nathan was really on his game, earning a 2.54 SIERA:
Notice that Nathan is all over the place, but the scouting reports have always said that he puts a ton of effort in his pitches and seems very mechanically unorthodox. Here is Joe Nathan as a Ranger, where I think (in watching his outings) he has looked absolutely awful. The metrics vary on this, as his good walk to strikeout ratio makes FIP and SIERA really like him. He has mainly missed in the strike zone, which is why he has given up so many line drives:
His release points are actually more consistent then what we saw in his prime, but notice that it is a lot lower, nothing over 7 feet. With such a small sample size, it seems unwise to classify him as “worse”, but he is definitely different. He is not the same Joe Nathan as we are used to seeing, he is a different pitcher.
The most egregious release point regression that I could find was Hong-Chih Kuo, which I detailed here (yes, I finished my Mariner article). However, I wonder, since we don’t have data on it we can’t really know, what pitch f/x release point data would have said about Dontrelle Willis. Willis was a dominant pitcher in 03-06, and then turned terrible but keeps getting jobs (minor league ones now, and that maybe over with this Baltimore debacle). Willis was terrible in spring training and cut pretty early on by the Phillies, but he did all his pitching in parks that didn’t have pitch f/x. So unfortunately, it seems that lack of data hurts us here. On the other hand, J.P. Howell is a very good example of someone who was effective, then had arm problems, and looked awful the next year. Here is when Howell was good in ’09:
It is not an extremely consistent release point, but it is a whole lot better than what he looked like in 2011:
There are exceptions of course, and nothing should be construed as a hard and fast rule, just look at Scott Feldman:
One would think the 2010 start was better, but that is not the case. Feldman was not bad in 2009, but was pretty awful in 2010 (and the 2009 start shown above was his best start according to win probability).
Cliff Lee is a starter that has a pretty inconsistent release point for someone who is a top of the rotation starter. However, consistency within a start is still predictive for Lee:
The first start was a better start than the 2nd one, so it would seem that a more consistent release point is helpful to Cliff Lee, one of the league’s best pitchers.
George Sherrill is a curious one when it comes to his release points. As you will see, it is something he has really struggled with, and he has also had tons of arm problems. Of course, the two seem very related. George Sherrill was pretty bad in 2010 and his release points were even worse:
In 2011, Sherrill had a bounce back year, and it seems more consistency with his release point was one of the reasons:
The Mariners signed him for the 2012 season, and at a game I was at (actually Yu Darvish’s U.S. regular season debut), he pitched absolutely terrible and then went on the DL:
Amazingly, his release point was more consistent. It did seem to drop in height a little bit again like in 2010, but it is hard to call it really drastic. Of course his velocity sat at 86 MPH during the game, and I actually thought he was throwing breaking balls at the time (probably not good if your fastball is slow enough to get mistaken for a breaking ball).
Jair Jurrjens was once a really nice starter for the Braves, but has been absolutely terrible in 2012 (7.77 FIP) and was demoted to AAA. Notice how Jurrjens is actually releasing the ball higher now.
This was Tim Lincecum in Game 5 of the World Series in 2010, where he was brilliant:
Here was his horrible opening day start in 2012 against the Diamondbacks:
His release point drops even further on his start against the Mets on Monday, a start that provided good results, even though some scouts had some negative things to say about it:
Much ink has also been spilled on how Roy Halladay is different this season (there was even that story of a scout in spring training saying that Halladay was done. If I were a GM or President of a team, I would want to know who the scout was so I would make sure he is not on my payroll). So I looked at a 4 start stretch in 2011:
So pretty consistent as you can see. What about the “new Halladay”?:
It doesn’t see there is much of a difference. So the idea of a “new Halladay” doesn’t really seem to fly, at least compared to the drastic changes we have seen in Lincecum or Jurrjens.
Release point changes may actually be more telling than velocity changes. The next market inefficiency (this is most likely already the case with the best pitching coaches in the game, but I just never seem to hear much about release points) may be a guy in the dugout (the manager being able to see it would be preferred so he can yank the pitcher with a bad release point) who can see release points and know when a pitcher is cooked. It also is important for GMs to evaluate when discussing contracts with pitchers. For fantasy owners, it is a tool to see whether a pitcher has actually changed or just going through a bad streak (although I guess a pitcher could have a bad streak with his release point, a further study to see whether pitcher’s can recover and re-discover there release points should be done).
One should always consult other resources as well, such as advanced statistics, velocity, and scouts, but I believe I have shown that release point changes or inconsistencies can be a determination as to whether to keep a pitcher or not.