This article is part of our The Saber's Edge series.
Last week, I examined how individual batted-ball profiles inflate pitcher ERAs. I'll now dive into how team defense and handedness splits get involved. The key to higher-than-expected ERAs is high BABIPs. As I showed last week, some of the differences can be attributed to the pitcher's overall profile. This week, I'll show how much of the difference can be attributed to team defense and handedness.
Team Defense
I've valued Robbie Ray quite a bit higher than just about everyone else in the industry; in fact, I recently ranked him as my No. 21 starting pitcher while none of my peers had him any higher than No. 40 overall.
Ray repels owners after posting a 4.90 ERA and 1.47 WHIP last season. A .352 BABIP caused those high numbers. While Ray's line drive profile leads to more hits, he doesn't get all the blame.
As a team, the Diamondbacks allowed a .320 BABIP leading to a team ERA (5.09) noticeably higher than its FIP (4.50). That's the fourth-highest BABIP any team has allowed since 2002. (Thanks, Yasmany Tomas.) The Cubs were on the other end of the spectrum; their .255 BABIP was the best mark over that time frame, and they posted a 3.15 ERA while their FIP stood at 3.77.
So how much of a team's BABIP tendency trickles down to the pitcher? Common sense says it should be near 1:1, but as I showed in my last article, the pitcher's profile matters. By taking all pitchers who threw
Last week, I examined how individual batted-ball profiles inflate pitcher ERAs. I'll now dive into how team defense and handedness splits get involved. The key to higher-than-expected ERAs is high BABIPs. As I showed last week, some of the differences can be attributed to the pitcher's overall profile. This week, I'll show how much of the difference can be attributed to team defense and handedness.
Team Defense
I've valued Robbie Ray quite a bit higher than just about everyone else in the industry; in fact, I recently ranked him as my No. 21 starting pitcher while none of my peers had him any higher than No. 40 overall.
Ray repels owners after posting a 4.90 ERA and 1.47 WHIP last season. A .352 BABIP caused those high numbers. While Ray's line drive profile leads to more hits, he doesn't get all the blame.
As a team, the Diamondbacks allowed a .320 BABIP leading to a team ERA (5.09) noticeably higher than its FIP (4.50). That's the fourth-highest BABIP any team has allowed since 2002. (Thanks, Yasmany Tomas.) The Cubs were on the other end of the spectrum; their .255 BABIP was the best mark over that time frame, and they posted a 3.15 ERA while their FIP stood at 3.77.
So how much of a team's BABIP tendency trickles down to the pitcher? Common sense says it should be near 1:1, but as I showed in my last article, the pitcher's profile matters. By taking all pitchers who threw 100 innings for the same team in back-to-back seasons, I've determined how much a pitcher's BABIP follows his team's BABIP.
The results were as expected, with the average pitcher changing .00086 BABIP for every .001 team change. The correlation was not strong, with a .12 R-squared, but the trend was noticeable. Considering the noise involved in BABIP, I was surprised there was any correlation. Pitchers playing for teams on the extreme ends of the spectrum, like the Cubs and Diamondbacks, could see significant changes as their defenses move toward the league norm.
The Diamondbacks' BABIP jumped 26 points from 2015 to 2016 after they lost A.J. Pollock for the season and were using Tomas every day, and that team defense helped push Ray's BABIP up 41 points over the same time frame.
On the other end of the spectrum, the Cubs' BABIP allowed went from .287 to .255 – a 32-point drop. Their starters followed the trend, with Jon Lester seeing his BABIP drop 47 points, Kyle Hendricks 46 points, and Jake Arrieta down 5 points (the average of the three: 33 points).
Now comes the hard part: projecting how team BABIP changes from season to season. Of course, projecting an individual's defense is close to impossible; combining all those players together for a team value seems like extreme fuzzy math. To see how teams will perform, some past references can be used.
Historically, team pitcher BABIP has regressed back to the mean, but it doesn't make huge jumps because the team contains many of the same players. The 10 teams with the highest season BABIP values saw their BABIPs drop by an average of 17.5 points. So if the Diamondbacks follow the average, they would post a team BABIP near .302 and Ray should expect his own BABIP to drop 15 points (17.5 *.86).
On the other hand, the Cubs intrigue me more.
Low-BABIP teams tend to maintain their low BABIPs a little better; the teams with the 10 lowest BABIPs saw an average jump of 13.8 points the following season. Even if the Cubs' .255 BABIP jumps to .268, that's still amazing. FanGraphs (Steamer) projects the team pitching staff to post a .296 BABIP. or a 44-point increase. Using historical samples from 2002 to 2016, the largest jump was 33 points by the Indians from 2005 to 2006. Since no team has posted a BABIP as low as the Cubs did, I think a 25-point jump is reasonable to .280 (especially with Kyle Schwarber in the outfield). That equates to a 21.5-point jump for each pitcher, but their projections reflect higher jumps of 35 to 50 points. Thus I'm a bit more optimistic about them than the projection system is.
To help find how much WHIP and ERA will move with a BABIP change, I compared pitcher BABIP from Season 1 to Season 2 for all starters (min. 100 IP). For a .001 change in BABIP, the pitcher's ERA will move by 0.019 in the same direction (r-square of .32) and WHIP will change by .0044 (r-squared of .58).
Going back to the 25-point jump in the Cubs' BABIP allowed, that would have their ERAs jumping by 0.48 and their WHIPs by 0.11. Most projections have the Cubs regressing by about one run of ERA with a 0.15 jump in WHIP. The WHIP is similar, but the extra half run of ERA is significant.
The Diamondbacks and Cubs weren't the only teams with extreme BABIPs in 2016. The Twins (.319) and Rockies (.317 – they'll always be inflated) had abnormally high BABIPs, while the Blue Jays had a notably lower-than-normal BABIP (.282). The rest of the teams were in a 21-point band from .287 to .308. While the Twins don't have many fantasy-relevant pitchers, owners should expect the Blue Jays' starters to regress to the mean.
It's key to remember with high-BABIP pitchers that the defense behind them is directly responsible for their ERAs being higher or lower than expected. These BABIP outliers will eventually regress toward the mean, and they'll bring their pitchers with them.
With team defense out of the way, it's time to move on to the effects of handedness splits.
Pitcher Handedness Splits
This concept evolves from the idea that a pitcher gets hit around by one batter handedness type, and that impacts his BABIP. For instance, a pitcher's high strikeout rate against lefties in combination with his struggles against righties may result in a high BABIP.
After looking over the data in too many different ways, I found no correlation. None.
For every pitcher who displayed the traits behind the theory, there was another with exact opposite traits. I wonder whether there's some confirmation bias going on with this theory. When the split exists, it gets pointed out. When it doesn't, people just ignore it. It seems like a pitcher's struggles against one handedness type is normally distributed to both FIP and ERA. So the key point is that a pitcher with extreme handedness splits shouldn't be expected to have an ERA higher than his FIP.
That's it for now. As always, let me know if you have any questions and I will gladly answer them.
