When to Believe Velocity Gain

Last week, I wrote about some findings regarding in-season fastball velocity loss and how experiencing a loss in different months affects a pitcher’s chances of finishing a season with diminished pitch speed. The general takeaway was that June and July were the most telling months.

But what about velocity gain? We know that, generally speaking, pitchers lose velocity more than they gain it. So while velocity loss isn’t good, it’s to be expected — and starting pitchers seem to be able to deal with that loss better than relievers. Pitchers who can stave off velocity loss (year-over-year change between +/- .5 mph) perform even better. Moreover, if a pitcher gains at least 1 mph on their fastball in a season they are twice as likely to maintain some or all of that gain the following year.

Gaining velocity, while not a guarantee of better performance, is certainly a boon to a pitcher and his organization. But given that velocity varies for all sorts of reasons, when can a team have confidence that the increase they’re seeing is real and sustainable?

As with the velocity loss study, I looked at fastball velocities for individual pitchers in each month since 2002. I used the same methodology to compute a pitcher’s odds of finishing the season up at least 1 mph when their velocity was up at least 1 mph in a given month from the same time the previous year. (For example, if April 2012’s velocity is at least 1 mph higher than April 2011, what are the increased odds of finishing 2012 up at least 1 mph?) Only pitchers who fulfilled the same role in consecutive years — in this case, starters in both years — and pitched at least 25 innings in each month were included.

Here are the results:

April through June exhibits essentially the same patter that we saw with velocity decline. In fact, the odds ratios for increased velocity lasting for the entire season are extremely similar to those for decreased velocity. Once we hit July, though, there appears to be a qualitative difference in how reliable a predictor a velocity gain is than a velocity loss in that same month.

Starting pitchers who gained at least 1 mph in July — relative to the same time the previous year — were 25.4 times more likely to finish the season up a at least a full mile-per-hour than pitchers who were not up at least 1 mph in July. Compare that with 13.7 for velocity loss.

% with 1 mph increase in month and full season % without 1 mph increase in month but did increase full season
April 35% 6%
May 43% 5%
June 45% 3%
July 53% 2%
August 49% 2%

Of course, this raises the question as to why July (and August, for that matter) should be better indicators of velocity gain than loss? I don’t have a sound answer at the moment, but my guess is that there’s something qualitatively different about a velocity gain. We know it’s harder to gain velocity and to sustain that gain — and the reasons for the gain are probably fewer than for a loss, especially once you approach the final few months of the season.

A pitcher’s velocity can be down for all sorts of reasons — mechanics, fatigue, injury, weather, park conditions, etc. But it might be that the reasons behind velocity gains (year-over-year, at least) are less fleeting and variable. In other words, once a pitcher has lasted into July with increased velocity, that velocity isn’t likely to decline enough during the last few months to end the season lower than the year before. Obviously, there are lots of research opportunities here to further flesh out this initial data.

As before, here is a list of those (few) starting pitchers whose average fastball velocity is up at least 1 mph this July versus last year:

Name Team July 2012 Fastball Velocity (mph) Difference from July 2011 (mph)
James Shields Rays 92.1 1.3
David Price Rays 95.9 1.1
Max Scherzer Tigers 94.8 1.1
Francisco Liriano Twins 93.1 1.1
Jeremy Hellickson Rays 91.9 1
Jake Westbrook Cardinals 91.1 1

Let’s take a look at Max Scherzer, who has been an interesting pitcher this year. So far in 2012, Scherzer’s average is fastball is about 1 mph faster than in 2011 and 2010. He’s induced 3% more swinging strikes with that pitch than in each of the previous two seasons. Additionally, Scherzer’s overall strikeout rate has jumped almost 8% compared to 2011. While his FIP has improved (4.14 vs. 3.83), his ERA has actually jumped from 4.43 to 4.62. There seems to be two reasons for this.

First, Scherzer’s BABIP against is .339–probably due in no small part to the Tigers lackluster defense. Second, Scherzer is both inducing more fly balls and more of those fly balls are leaving the park (HR/FB 14.4%)–this also applies to his fastball. Getting bit by the home run bug is probably a function of being more aggressive with his fastball and not locating well at key points in an at bat, but the increase in strikeout rate is also pretty clearly a function of his increased velocity. Increasing velocity is not a silver bullet–like any other pitch, you still need to locate the ball. However, the fact that Scherzer is throwing the hardest he has since his rookie season–and that the gain seems to be real–is good news for the Tigers. Scherzer is only 27 years old, so if he can make further refinements to his control and combine that with a more explosive fastball the Tigers will have a very nice asset in their rotation who is just entering his first year of arbitration.

 





Bill leads Predictive Modeling and Data Science consulting at Gallup. In his free time, he writes for The Hardball Times, speaks about baseball research and analytics, has consulted for a Major League Baseball team, and has appeared on MLB Network's Clubhouse Confidential as well as several MLB-produced documentaries. He is also the creator of the baseballr package for the R programming language. Along with Jeff Zimmerman, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Twitter @BillPetti.

10 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Chris
11 years ago

Why is the metric the end of the season and not 1-year out? You take great pains to ensure that everything is on the same metric month-wise throughout, except for there, where the difference between April or May until the end of the season is much different than July/August? “Starting pitchers who gained at least 1 mph in July — relative to the same time the previous year — were 25.4 times more likely to finish the season [why isn’t this the following July?] up a at least a full mile-per-hour than pitchers who were not up at least 1 mph in July.” Maybe I am just confused because the graphic above is quite challenging.

Chris
11 years ago
Reply to  Bill Petti

To my eye, you add a temporal confound to what I see as the broader question – how much does a velocity change from (t – x) to (t) predict a velocity from t to (t + y). Fixing the time frame for x, but not for y seems arbitrary to me, and I think would confuse any attempt to interpret the results. I don’t know whether it is necessary for x=y, but I think the very reason why you try to compare apples to apples on the front end of the change (May 11 vs. May 12, for example) means you should be making the same comparison on the other end.
While I agree the comparison can be made between velocity loss and velocity gain, I would prefer to see that comparison made in with a fixed time difference. Regardless, this is fascinating stuff to me and I am only commenting because I would love to see more. Cheers.