This article is part of our The Saber's Edge series.
This week, I will give some simple baseline rules to follow when looking at the StatCast data. Additionally, I will look at some of the players with the biggest changes from 2015 to 2016.
The basic information that is newly available has three variables: velocity, launch angle (flyball vs. groundball) and launch direction (where on the field). All three can be combined to do some outstanding work like Bill Petti did recently for The Hardball Times where he found the likely chances a batted ball can turn into a hit. The possible studies are endless with all the new information. I will start with just average batted velocity and see how it affects certain stats used in fantasy baseball. By using just velocity, it will be easy to look up a hitter's values at MLB's StatCast website, BaseballSavant.
One quick disclaimer on StatCast data. Some of the data is missing, but the collection of this data is getting better. Recently, I talked to some of the people collecting
This week, I will give some simple baseline rules to follow when looking at the StatCast data. Additionally, I will look at some of the players with the biggest changes from 2015 to 2016.
The basic information that is newly available has three variables: velocity, launch angle (flyball vs. groundball) and launch direction (where on the field). All three can be combined to do some outstanding work like Bill Petti did recently for The Hardball Times where he found the likely chances a batted ball can turn into a hit. The possible studies are endless with all the new information. I will start with just average batted velocity and see how it affects certain stats used in fantasy baseball. By using just velocity, it will be easy to look up a hitter's values at MLB's StatCast website, BaseballSavant.
One quick disclaimer on StatCast data. Some of the data is missing, but the collection of this data is getting better. Recently, I talked to some of the people collecting the data and they mentioned the missing information was extreme batted balls -- balls hit straight up and straight down. Also, hard line drives right down the foul lines sometimes don't get picked up. Users of this data need to understand the data will always be changing for the better and someone in the future should and will take a hard look at the data again to get some more accurate measures.
For this study, I compared how a few stats changed as the average batted-ball velocity changed from the first half of 2015 to the second half. The hitters had at least 100 batted balls in each half. It will be nice after this season to find how stats compare from season to season instead of comparing half seasons. I decided to concentrate on some of the most common stats. If you want more examined, please let me know.
To start the work, I will start with comparing how batting average (AVG) compares to the average exit velocity.
Note: r-squared measures how much two values are correlated. A value of one is perfect correlation and zero is no correlation.
Exit Velocity correlated with AVG
Change: +1 mph = +.005 AVG
R-squared = .10
Batting average is the most common fantasy stat I will look at today, but the problem with it is the noise involved. Batting average contains strikeouts, which have nothing to do with exit velocity.
To limit the discussion's scope, I will move onto batting average for balls in play (or BABIP).
Exit Velocity correlated with BABIP
Change: +1 mph = +.0065 AVG
R-squared = .05
The change value is similar with a 1 mph change being just over five points in batting average change, but the correlation gets worse. Historically, both AVG and BABIP take a while to stabilize, so it seems that batted-ball velocity has some overall effect, but not enough to make any significant claims.
Now, BABIP doesn't include all batted balls. It is missing home runs. Here are the numbers for home runs per batted ball.
Exit Velocity correlated with HR/BIP
Change: +1 mph = +.004 HR/BIP
R-squared = .08
Again, very bad correlation, so any increase in velocity most likely will increase the number of home runs, but it is not guaranteed. Now, the .004 increase in HR/BIP may not seem like a bunch, but over a season's course, it can add up. Usually, a full time hitter puts 400 balls in play in a season. A 1 mph increase means the hitter likely will have 1.5 more home runs that season.
Now, home runs are not the only aspect of a player's power, so I will now look at how the hitter's isolated power (ISO)
Exit Velocity correlated with ISO
Change: +1 mph = +.012 ISO
R-squared = .13
Well, the r-squared value is stronger, but not at all significant. The average increase in ISO is decent. Players who see their batted-ball velocity change should look to changes in their power profile.
With the preceding information, I have two conclusions.
1. We just have not gotten enough batted-ball data to really conclude anything when looking batted balls in play.
2. If a person feels compelled they have to use StatCast data to draw any conclusions, look at the power numbers (HR and ISO) instead of (AVG and BABIP).
Now, let me put some of the preceding data into context. Here are the biggest batted velocity movers from 2015 to 2016 and how their stats stack up.
Biggest Increase in Batted-Ball Velocity 2015-2016
PLAYER | 2015-16 CHANGE | 2015 VELO | 2016 VELO | 2015 AVG | 2016 AVG | 2015 BABIP | 2016 BABIP | 2015 HR/BIP | 2016 HR/BIP | 2015 ISO | 2016 ISO |
Jose Altuve | 3.8 | 85.1 | 88.9 | .313 | .344 | .329 | .350 | .026 | .043 | .146 | .207 |
Nick Markakis | 3.5 | 87.3 | 90.8 | .296 | .239 | .338 | .280 | .006 | .005 | .080 | .083 |
Eric Hosmer | 3.5 | 88.9 | 92.4 | .297 | .318 | .336 | .354 | .036 | .063 | .162 | .212 |
Buster Posey | 3.3 | 88.4 | 91.7 | .318 | .257 | .320 | .259 | .037 | .044 | .153 | .189 |
Chris Carter | 3.1 | 90.5 | 93.6 | .199 | .225 | .244 | .254 | .098 | .113 | .228 | .286 |
Ryan Zimmerman | 2.8 | 90.9 | 93.7 | .249 | .233 | .268 | .253 | .058 | .055 | .217 | .194 |
Ben Revere | 2.6 | 81.1 | 83.7 | .306 | .211 | .338 | .229 | .004 | .008 | .071 | .075 |
Victor Martinez | 2.6 | 87.7 | 90.3 | .245 | .333 | .253 | .352 | .028 | .048 | .120 | .185 |
Danny Espinosa | 2.5 | 85.5 | 88.1 | .240 | .223 | .299 | .246 | .049 | .073 | .169 | .188 |
Starlin Castro | 2.4 | 84.8 | 87.3 | .265 | .253 | .298 | .274 | .024 | .046 | .110 | .165 |
Biggest Decrease in Batted-Ball Velocity 2015-2016
PLAYER | 2015-16 CHANGE | 2015 VELO | 2016 VELO | 2015 AVG | 2016 AVG | 2015 BABIP | 2016 BABIP | 2015 HR/BIP | 2016 HR/BIP | 2015 ISO | 2016 ISO |
Brandon Belt | -3.4 | 88.6 | 85.3 | .280 | .300 | .363 | .337 | .052 | .046 | .197 | .211 |
Paul Goldschmidt | -3.5 | 92.3 | 88.8 | .321 | .282 | .382 | .342 | .078 | .073 | .249 | .216 |
Jason Heyward | -3.6 | 89.6 | 86.1 | .293 | .236 | .329 | .297 | .028 | .025 | .146 | .104 |
David Peralta | -3.8 | 90.7 | 86.9 | .312 | .259 | .368 | .320 | .047 | .037 | .210 | .189 |
Lucas Duda | -3.8 | 92.1 | 88.2 | .244 | .231 | .285 | .240 | .080 | .068 | .242 | .200 |
Andrelton Simmons | -4.0 | 85.6 | 81.6 | .265 | .219 | .285 | .224 | .008 | .009 | .073 | .061 |
Mark Teixeira | -4.0 | 89.5 | 85.5 | .255 | .180 | .246 | .235 | .099 | .025 | .293 | .084 |
Jed Lowrie | -4.4 | 88.3 | 83.9 | .222 | .303 | .233 | .364 | .048 | .007 | .178 | .063 |
John Jaso | -4.4 | 89.4 | 84.9 | .286 | .287 | .336 | .321 | .034 | .018 | .173 | .123 |
Giancarlo Stanton | -6.9 | 97.7 | 90.8 | .265 | .192 | .294 | .243 | .144 | .103 | .341 | .223 |
Notes on a Few Players
Jose Altuve , Giancarlo Stanton: The two players on the extremes have seen the excepted results. Altuve has seen all four of the stats increase a decent amount while Stanton has seen all four of his decline. Instead of the crazy near 98 mph level, Stanton's exit velocity is similar to Victor Martinez and Nick Markakis. For Stanton to be Stanton, he needs to start making more solid contact.
Nick Markakis: Markakis has seen the second largest jump in exit velocity, but the increase is not showing up in his results at all. He may be a candidate for positive improvements in his other stats.
Brandon Belt: Belt is having a great season, but a person wouldn't know it looking at his exit velocity. The key to Belt's turnaround is completely related to his improved plate discipline. His BB% has increased from 10 to 16 percent and his K% has dropped from 26 to 16 percent. Belt is a perfect example of not concentrating on one stat when examining a player, but looking at the player's full profile.
Hitter exit velocity, along with the direction and angle, likely will be an area with great advances in the next few seasons. Now, though, most of the breakthroughs are limited because of the limited amount of data. As hitters see their exit velocity increase or decrease, obviously their production likely will change. I found the average change expected from some common fantasy stats to give fantasy owners some perspective. Power related numbers are the most likely to change as exit velocity changes. I am done on this subject for today, but let me know in the comments if you have other aspects of exit velocity you want explored.