Sixty Feet from Home

Throwing Smoke

A few days ago I read an article by Dave Cameron at FanGraphs about fastball velocity across MLB. The article makes a few interesting arguments about how average velocity changes over the course of a season. According to the Cameron, April is the lowest velocity month and fastball speed gradually rises throughout the summer and peaks in September. He also points out that pitching velocity stabilizes early on in the season and is a good predictor of pitcher performance.

At the time, I speculated that Mike Leake was throwing a little slower than previous years (the gun showed him at 89 most of the night). It made me curious to do a deeper investigation into the subject, not only for Leake, but for the rest of the rotation. Here is what the underlying data shows so far for the Reds’ starting pitchers:

Table1[FBv13 is average fastball velocity in 2013. FBv14 is average fastball velocity so far in 2014. BABIP is batting average on balls into play, both for 2014 and career. K% and BB% are pretty straightforward.]

The first number that jumps out of this table is that Alfredo Simon, despite transitioning from a reliever to a starter, has not lost any speed on his fastball. He leads the rotation in fastball velocity.

Despite April being the slowest month, Mike Leake is the only one of the starting five that has lost speed on his fastball compared to the average for 2013. This is a strong positive sign for the Reds’ starting five. Johnny Cueto has added almost a full MPH to his fastball, and we can expect that to increase as the season progresses. I’m working on another post about the relationship between fastball velocity and FIP, but the early evidence indicates that small changes in fastball velocity can play a significant role in a pitcher’s performance.

Cueto’s pitching line so far is ridiculous and almost certainly unsustainable. A .191 BABIP (especially compared to his career number) indicates that he will undergo a regression as the season continues. His K% will come back to earth eventually. Despite this, both his xFIP and SIERA project him to have a great season.

Homer Bailey’s line is exhibit #1 for why we should not trust small sample sizes. His ERA is inflated by an unreal .429 BABIP and his FIP is suffering from that time he threw batting practice to the Pirates his last start against the Pirates. Both of these are unsustainable, and his 2.80 xFIP demonstrates how unlucky Bailey has been with HR/FB ratio over his first few starts. I won’t write much more about this because Justin already has written convincingly about it.

Tony Cingrani, while possessing a weapons-grade fastball, is also working on his other pitches. As that article points out, Cingrani is excellent at hiding the ball on his delivery and has made a huge improvement on his secondary pitches. Combine all that and you can see why he has struck out 29% of the batters he’s faced this season.

Here’s a chart that shows various bottom-line measurements of pitcher performance, in this case for the Reds’ rotation based on their current numbers from 2014:

Table2Looking at the ERA column in the above chart, you can see why evaluating a pitcher on the number of runs he has given up, not on strikeouts, walks and home runs can be extremely misleading. Analyzing their current BABIP (and the fact that a pitcher has little control over his BABIP is backed up by how similar the career BABIP is for all five pitchers), you can see how “lucky” four of the five Reds’ pitchers have been through three games.

Each of them (save Homer Bailey) has a much, much lower BABIP than the league average (~.300) or their respective career averages. As their luck evens out and that number moves toward their usual number, we can expect their ERA to rise.

A few weeks ago we discussed how using walks, strikeouts, and home runs are a better way to predict future performance than ERA. The scatter plot I posted indicated that only 9% of the variation in year-to-year ERA could be explained by the previous year’s ERA.

At the end of the year, which do you think will be closer to Simon’s 2014 ERA: his current ERA (1.20) or his xFIP (3.92)? Or Cueto, will he throw a sub-2 ERA or be closer to his SIERA projected 2.52? If you answered xFIP and SIERA to those two questions, then let’s extend the courtesy to Homer and realize that his xFIP (2.80) and SIERA (3.16) are much closer to where he will end the season than his current ERA (8.16).

I feel bad for our other authors who have to write about the Reds hitting, because so far we have an embarrassment of riches to report from sixty feet from home (knock on wood). With 4 of the 5 at or above their 2013 fastball velocities, we have early evidence that the wave of good pitching will continue into the future.

11 thoughts on “Throwing Smoke

  1. Leake’s velocity is down, however, it seems to me that he has more motion on his fastball this year. The movement seems somehow sharper and later in the pitch. That could be a great sign that he’s really coming into his own and might push his season ERA lowere than his career numbers suggest. I’m quite hopeful after seeing his first few outings this year.

    • Leake has looked good, but he’s also been fortunate in his home plate umpires: they’ve been guys willing to call strikes on the edges, which Leake has to have to be effective. I fear that Leake is likely to have his ERA go up as soon as he has a start or two with umpires that have a tighter zone.

      Great article, BTW. Interesting information, presented clearly.

      • I agree that Leake needs competent umpiring to be successful (so did Maddox). And, while I don’t doubt Michael’s data of current average velocity, I have seen Leake hit 92 and even 93 a number of times. That is Cingrani territory, and I suspect that, as a contrast to Leake’s heavy use of off-speed pitches, his fastball is pretty effective.

        • Leake has been able to crank it up form time to time — his max speed this season (From FangGraphs PitchFX) is 92.3. FBv is his average fastball speed. Given the small number of games played, pitch misclassification is a bigger problem than later in the season (sometimes PitchFX will incorrectly record a two seam fastball a 4 seam. its not often, but it happens).

  2. Very interesting. While I fully support Homer and his contract I do think we have to be fair to his detractors and acknowledge that his ERA might also be inflated due to giving up 6 HRs in 3 games. That other logic and math stuff you state is comforting, at least to me, though.

  3. Remember, there are several things that go into making a good pitcher. One can be being able to throw heat. But, there is also location, fooling the hitters, and differences in pitching velocities. That was what made Arroyo good, when his fastball and his breaking stuff had a “significant” enough difference in velocities. And, when Arroyo was bad, it was either later in the games or when his fastballs and breaking balls didn’t have enough difference in velocities.

  4. The bright note on the offensive side of the ball is that half the starters have an early BABIP significantly below .300.

    Cozart .132
    Bruce . 156
    Frazier .220
    Hamilton . 222

    • I wonder about this and using the .300 average BABIP as “standard” as more teams start to use defensive shifts and positioning to play against a hitter’s tendencies. Will we see this number go down across the league over time? I realize much of baseball is still luck and chance, but if you play the odds to your favor I could see this hurting the average player and making the elite players really stand out that can hit well to both sides of the field. Just a random thought.

  5. I’m not sure I buy that BABIP for a pitcher is something you can look at and say there is an “average” across the league that everyone needs to regress to.

    For example, let’s say you have a pitcher out there whose fastball tops out at 85MPH with no movement. After hitters see a few of those pitches and realize the ball is coming in at that speed straight as an arrow, I’m willing to bet most major league hitters could make contact in such a way that would not result in an out. In fact, probably the ONLY outs would come from guys swinging for the fences who just missed.

    Changing speed and movement is the name of the game as far preventing batters from putting balls in play. If a pitcher is not doing this effectively, you can bet people are going to hit him, regardless of what the league “average” is. Put a lifer AAA pitcher out there, and I’m pretty confident his BABIP will be quite a bit higher than the league average.

    • Yes, but that is why it is called the league average. BABIP has averaged between .293 and .302 across all pitchers for the last 12 years (I didn’t bother looking further back).

      You are correct in your thought process, but when the career BABIP numbers are taken into consideration (all 5 are from .292-.299), it could and should be expected to average out.

      I have long preferred the advanced metrics for pitchers because the take into consideration things like HR/FB Ratio, BABIP, and hit sequencing.

    • It depends on the pitcher. If you have a pitcher with a history of success and an established baseline BABIP, then I think you can look at his current year BABIP and extrapolate some measure of future performance.

      For the AAA guy in your example, you are correct. BABIP varies by type of batted ball with line drives having the highest BABIP. So, if a pitcher isn’t fooling anyone with his stuff or his speed, then the hitters are going to have more line drives. In that case, you couldn’t expect the pitcher’s BABIP to decline to a “league average” unless he was also able to make adjustments to his pitching.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s