Monday, April 20, 2015

Early Season Velocity Issues



After pitching like the ace of the Nationals' staff last year, Jordan Zimmermann has gotten off to a rocky start in 2015. Through three starts, he sports a 6.14 ERA and a 4.36 FIP with a reduced strikeout rate and an increased walk rate. In other words, almost everything that could get worse for Zimmermann, has. Underneath those numbers is an inflated BABIP allowed of .321, much higher than league average and a good 20 points higher than Zimmermann’s mark last year, so he has been on the losing end of some bad luck. Much of Zimmermann’s struggles stem from his nightmare of an outing against the Red Sox where he lasted only 2.1 innings. He has pitched on 14.2 innings so far in the young 2015 season, so I’m not going to extrapolate these numbers out for the year. However, there is one troubling indicator that may be a telling sign of things to come.

Zimmermann relies heavily on his fastball, more so than your average starting pitcher. He has thrown his fastball over 60% of the time over the span of his career and he upped his reliance on it last year to 70.9%, a career high. Between his stellar command, the movement on his fastball, and his ability to throw the fastball at 94 MPH (this velocity all velocity readings going forward are from Pitchf/x provided data), it is a tough pitch for hitters to make solid contact with. After averaging right around 94 MPH on that fastball over the last 3 seasons, Zimmermann is slinging it at only 92.1 MPH this year. Given his heavy reliance on the fastball, this could spell trouble for the most reliable member of the Nationals’ pitching staff.

If this loss in velocity is real, there is a two pronged problem Zimmermann will run into. First of all, the decrease in velocity means hitters have more time to read and react to the pitch and will be able to make better contact. Secondly, they likely won’t chase fastballs out of the zone as much since they have will have more time to discern whether the pitch is a ball or strike. What you end up with, then, is a vicious cycle. Zimmermann can’t fool batters with pitches out of the zone, so he has to throw in the zone more. Since he is in the zone more with a slower fastball, hitters will be getting better pitches to hit.

Granted, we are only looking at 14 innings of work here. While a pitcher’s velocity stabilizes quickly (meaning that our measurements early on become a reliable data point more quickly), velocity also has a tendency to increase during the season as the weather warms up and pitchers build up to their optimal arm strength. So perhaps all we are seeing is Zimmermann building up arm strength at the start of season. Taking a look at last year, Zimmermann’s first 3 starts were also below his average for the season, sitting around 93.2 MPH. That is still a full 1 MPH faster than what we have seen this year and only 0.5 MPH off his 2014 average velocity of 93.7 MPH. While I would expect Zimmermann to up his velocity somewhat over the course of the season, his history tells us it won’t be a significant increase of 1-2 MPH.

The other concern when working with Pitchf/x data is the reliability of the measurements. The people who run Pitchf/x openly admit that their early season numbers for velocity can occasionally be fast or slow by a few ticks as they calibrate their system for the new season. Comparing the velocity readings of multiple pitchers can help to cancel out some of the noise that may be inherent in the velocity data due to this calibration time period. Zimmermann has three starts in 2015, two at home in Nats Park and one on the road at Fenway. In game 1 of the year, Zimmermann averaged 92 MPH according to the Pitchf/x gun in DC. In that game, he faced off against Mets phenom Jacob deGrom who clocked in at 94 MPH. In 2014, deGrom averaged 93.5 MPH on his fastball. It appears that, if anything, the readings from this first game were slightly hot (i.e. higher than true velocity), a bad sign for Zimmermann. In game two against the Red Sox, Zimmermann averaged 91.7 MPH while his opponent Rick Porcello came in at 91.2 MPH compared to his average in 2014 of 90.5 MPH. Again, if we were to make any manual adjustments to the velocity readings, we would have to decrease them. Finally, in Zimmermann’s most recent outing against the Phillies, he was up to 92.5 MPH while Aaron Harang threw at an average speed of 88 MPH versus his 2014 average of 88.8 MPH. In this case, we might add a couple of ticks to the register, although Harang is on the part of the career performance curve where you would expect to see velocity loss each year, given his age.

Taken together, you have 2 games in which you might dock a little off of Zimmermann’s readings and one game where you might add slightly to the readings. Overall, nothing looks totally out of line and any change you made to the data wouldn’t change the fact that Zimmermann’s velocity has not measured up to what it used to be. As we already covered, there are several valid reasons that this velocity loss is not something to be concerned with: early season arm strength building, cold weather, Pitchf/x systems slightly out of tune. On the other hand, pitchers are fragile players and drop off in performance or, even worse, injury are always lurking around the corner and a decrease in velocity is usually one of the first signs of trouble. I’m not standing outside Nationals Park with my “the end is coming (for Zimmermann)” sign. I’m just raising the red flag.

No comments:

Post a Comment