If double plays are a pitcher’s best friend, home runs are their mortal enemy. ERA varies from year to year, with the number of homers allowed being the chief driving force. Today, we’re going to look at the level of control a pitcher exerts on their home run total.
There are two main factors governing homers:
- Fly ball percent (FB%)
- Home Runs per Fly Ball (HR/FB)
The math extends beyond the scope of this discussion, but it can be demonstrated that a pitcher has a lot of control over his FB%, but very little influence on his HR/FB. It is this HR/FB variance driving the range of a pitcher’s ERA.
Before focusing on just homers, the previous statement does not mean to imply limiting fly balls (or inducing ground balls) is a skill. It’s a trait. Fly balls leaving the yard are bad, but those staying in the park are caught more than nine times out of 10. An out ensues on two of three grounders, so a fly ball pitcher tends to yield fewer hits than their ground ball counterparts. A fly ball pitcher with a high strikeout rate and good control, especially if they have a power-suppressing home venue, sport better all-around numbers than a ground ball pitcher.
Tying this column in with the series on expected stats, there are a few measures of expected homers (xHR). They will come up frequently in the impending starting pitcher profiles, which are part of the Fantasy Index Baseball Draft Kit.
Statcast xHR
Here is the link for those interested in learning more about Statcast. The nuts-and-bolts version is Statcast is run by MLB, and is state-of-the-art tracking technology, quantifying much of the action on the field.
Statcast measures the exit velocity, launch angle and direction of every batted ball, and records the outcome. The data is often normalized to account for weather, atmospheric conditions, etc.
Every batted ball is compared to those with the most similar characteristics. If that specific contact resulted in a home run 75% of the time, the batted ball is recorded as .75 HR (with the other 25% of the outcomes also recorded). This is carried out on all the contact yielded by the pitcher, and the sum of the HR is pitcher’s xHR.
If the actual HR surrendered is more than the xHR, the pitcher experienced some bad luck. If the actual HR mark is below xHR, he enjoyed some good fortune.
Average Fly Ball Distance
Several years ago, Mike Podhorzer from Fangraphs discussed how well average fly ball distance correlates with HR/FB. I have always been fascinated by the simplicity and power of this finding. Podhorzer has since refined his version of xHR (just search Podhozer Fangraphs average fly ball distance).
The process avails a different version of xHR which can be used in concert with the Statcast xHR, or perhaps as a means of identifying the difference from expected. I have also found it to be a useful in-season tool to identify pitchers (and hitters) ripe for regression.
Barrel%
Barrels are another Statcast metric. Those curious can learn more about its definition. In short, “Barrel is a batted ball with similar hit types in exit velocity and a launch angle that has led to a minimum. 500 batting average and 1.500 slugging percentage. The batted ball requires an exit velocity of 98 miles per hour to be barreled.”
Barrel% is Barrels per batted ball event (BBE). Barrel% correlates reasonably well with HR/PA and HR/BBE. This can be used to come up with another measure of xHR. Personally, I find the other estimations better, so I use it in more of a comparison sense, perhaps noting a certain pitcher’s Barrel% is trending upwards.
Application
Much of the utility of expected stats is in more of a descriptive than predictive sense. That said, those of us in the business of prognosticating player performance need something to plug into our little black box. If I am aiming to project a pitcher, I care more about what outcomes his skills indicate he should have achieved than those he actually generated. However, none of the expected stats are perfect.
The way I account for the delta between expected and actual is regressing actual towards expected. The usual starting point is 50%. That is, if a pitcher yielded 20 homers, but his xHR was 24, I’ll plug 22 into the little black box.
Anything between 20 and 24 can be justified (though I will argue strongly against 20, the actual number). My spreadsheet is designed to easily override the regression. Sometimes, I’ll use the xHR from the above as a guide.
Next week, left on base rate (LOB%) will be reviewed.
Todd Zola is an award-winning fantasy baseball writer and 2020 inductee into the Fantasy Sports Writers Hall of Fame. He's the content provider for the 2024 Fantasy Baseball Index Draft Kit, available now. To purchase, click HERE.