+

Going Deep: Barrels and Ballpark Factors Pt. II

Dan Richards updates his park factors to add 2017-18 data and include directionality.

Recently, the writers here at Pitcher List have endeavored to bring you park factors you can apply to fantasy baseball. Typical park factors are based simply on historical run-scoring in the park relative to the league average. Because those fail to account for the new tools we have at our disposal, two months ago I decided to update park factors for fantasy baseball and make them highly relevant for evaluating home run hitting.

The premise was simple. Looking at total home runs hit in a venue in a given year and saying, alright, the one with the most home runs last year is the best park in which to hit home runs, is flawed. It fails to control for the quality of hitting and pitching in the park. If the home team had an excellent lineup, that would artificially inflate the home run park factor and belie the venue’s true effect on home run hitting.

Enter barrels. A barrel is a ball hit between 26 and 30 degrees at least 98 mph. For each additional mph above 98, the launch angle band expands. Considering home runs as a percentage of total barrels hit in a venue controls for competition across ballparks. In other words, it helps determine how many home runs came out of these extremely well-hit balls, regardless of how many of those hits occurred in the park, who hit them, or who gave them up. Thus, I created park factors based on the rate at which barrels were converted into home runs in certain venues in 2019.

Then, Alex Fast added to this research by considering whether hitters needed to barrel their pulled fly balls in order to hit home runs in certain places. He also researched the difference between expected and actual outcomes at different stadiums by direction.

This piqued my interest.  After Alex’s developments, I wanted to update my park factors to capture “directionality”. You may have been surprised to learn in my last iteration of park factors that Yankee Stadium wasn’t one of the top parks to convert barrels into home runs. Yet, it is the best venue for lefties. That nuance is necessary to properly evaluate hitters. Further, I wanted to expand my sample to reach back to 2017 in order to more accurately capture the effects of the different parks.

Let’s begin with a primer before discussing these new park factors.

 

Background

 

Examining the barrel-to-home-run conversion rate makes sense considering the enormous import of barrels. For example, in 2019, Brl/BBE% had an R2 to HR/BBE% of 0.775. Put differently, hitters’ barrel rates explained 77.5% of their home run rates last year.

This relationship is borne out in how frequently barrels and home runs occur simultaneously. In 2019, there were 9,290 total barrels. Of those, 5,536 resulted in home runs. Accordingly, hitting a barrel resulted in a home run 59.6% of the time. Moreover, there were 6,776 total home runs. That means 81.7% (5,536/6,776) of all home runs in 2019 were barreled and there were only 1,240 home runs produced on non-barreled balls.

These figures speak volumes. Most (59.6%) barrels generate home runs. And most (81.7%) home runs were the product of barrels as well. Unsurprisingly, as explained above, barrels are highly predictive of home run hitting.

 

Required Reading

 

I have a few disclaimers that I want you to read before you jump to the table below.

  1. First, I have not created park factors in the literal sense. After a discussion with Todd Zola in connection with my first iteration of park factors, I wanted to clarify this point upfront. Park factors traditionally measure the net run-scoring environment, whereas these park factors only measure the effect of the park on hitting home runs. As he said, “[a] venue can boost homers but suppress runs, and vice versa.”
  2. Second, certain hitters may outperform their home park factors by hitting non-barreled home runs. As an example, a righty like Alex Bregman or teammate Yuli Gurriel is more likely to eke out a home run in Minute Maid Park than in many other venues by pulling a non-barreled ball down the left-field line and into the Crawford Boxes. It’s worth keeping in mind, but the park factors as they appear below already reflect certain assumptions we have about non-barreled home runs (e.g., Yankee Stadium is extraordinarily favorable for lefties, Minute Maid Park for righties, etc.). In other words, it shouldn’t matter too much.
  3. Third, and finally, these park factors are not particularly useful for evaluating pitchers. We would use them for hitters because home runs are a counting statistic in most fantasy baseball leagues, but home runs allowed are not. Put differently, this third point relates to the first. A venue may boost home runs and suppress run-scoring based on a number of factors, including the size of the outfield, for instance. The venue may hurt a pitcher by causing him to cede more home runs according to these factors, but the venue may compensate for that in terms of balls in play that these park factors fail to capture.

And with that out of the way, let’s have a look at my new park factors.

 

Barrels and Ballparks

 

Home Run Park Factors 2017-19

 

There’s a lot to digest in this table, which I encourage you to sort in different ways. Each ballpark is broken down into three directions: (1) left-handed pull-side (right field), (2) center, and (3) right-handed pull-side (left field). Therefore, each park has a HR/Brl conversion rate and a corresponding park factor going in all three directions.

The second column, for example, shows the rate at which barrels were converted into home runs by lefties when they pulled the ball at each venue. The next column reflects corresponding park factors for lefties based on those HR/Brl conversion rates. These are z-scores. A z-score shows the relationship to the mean of a group of values, measured in terms of standard deviations—degrees of spread—from the mean. Where a z-score is 0, the value is equivalent to the mean in the sample. Where a z-score is 1.0, the value is one standard deviation greater than the mean.

Z-scores are useful because, in a vacuum, the percentages above are meaningless. They only have meaning in relation to one another. By way of example, the RHH HR/Brl% at Oriole Park of 85.1% was extremely high and the best in baseball. Put differently, when righties pulled barrels there, they resulted in home runs 85.1% of the time. That was beneficial to a hitter like Manny Machado, until his departure for San Diego. You might now look for Petco Park and think, he’ll be fine, Petco Park has a favorable RHH park factor as well, ranking sixth with an 80.7 HR/Brl%.

By just looking at the parks’ ordinal rankings or raw HR/Brl rates, however, you’d miss how much more favorable Oriole Park was than Petco Park. Indeed, the z-scores’ difference between the two parks (1.97 – 1.04) was nearly a full standard deviation in terms of spread. That means the difference for hitters between Oriole Park and Petco Park–though only six spots on the list and under five percentage points of HR/Brl%–is the same size as the benefit hitters get at Petco Park relative to the mean.

And that brings me to the mean HR/Brl% in each direction:

 

Direction Mean HR/Brl%
LHH Pull 73.6%
Center 42.8%
RHH Pull 75.8%

 

You should have these in mind when interpreting the HR/Brl rates above. The HR/Brl rate of 63.4% to centerfield at Dodger Stadium is a far greater boon to a hitter than the same rate to left field. Or, you could just use the z-scores, rather than the raw HR/Brl rates, which already account for the varying means.

 

Discussion

 

To be sure, these park factors are more fun as applied to hitters than in the abstract. With that in mind, there are some outlier parks worth discussing.

The first and most obvious is Oracle Park for left-handed hitters. The -3.09 z-score is the largest in the table. While, on average, left-handers’ pulled barrels result in home runs 73.6% of the time, Oracle Park’s right field is so difficult to barrel a home run that the rate plummeted to 48.7%–over three standard deviations below the mean. It really is a tough break for lefty Brandon Belt, who pulled seven barrels at home last season, with only one leaving the park. However, including directionality adds nuance to these park factors’ applicability. In the last iteration of this article, I mentioned that Buster Posey and Mike Yastrzemski may similarly struggle in 2020 to hit homers. Yet, my new park factors indicate that, while still unfavorable for righties, the z-score for them at Oracle Park is less than one standard deviation worse than the mean. The RHH HR/Brl conversion rate of 71.5% is therefore much closer to the RHH mean (75.8%) than the 48.7% LHH conversion rate is to the LHH mean (73.6%).

Likewise, as alluded to earlier, Yankee Stadium plays up far more for lefties than righties. Hence, Didi Gregorious, Aaron Hicks, and Brett Gardner had 95.2%, 94.7%, and 100% HR/Brl rates, respectively, on their pulled barrels from 2017-19. Whereas the park factor for righties actually plays in the opposite direction, being nearly a standard deviation below the mean. Accordingly, Aaron Judge, Giancarlo Stanton, and Luke Voit had 60.6%, 81.8%, 36.4% HR/Brl rates on pulled balls, respectively.

After Oracle Park, Comerica Park had the most extreme park factor to any direction. From 2017-19, there were 404 barrels hit to center in Detroit, the most hit to center field of any venue in that span. Yet, only 49 actually became home runs. That 12.1 HR/Brl% was nearly three standard deviations below the 42.8%. Think about the devastating effect of that on certain hitters for a moment. While they barreled the ball to center there more than anywhere else, they were rewarded at the worst rate.

In light of this, let’s examine Miguel Cabrera’s waning fantasy relevance. From 2017-19, he hit one barreled home run to center field at his home ballpark. Yet, he barreled the ball to center 18 times. That means his HR/Brl rate was 5.6% during that span, which is, of course, far worse than the 42.8 mean HR/Brl%. Similarly, Nick Castellanos hit 27 barrels straightaway at Comerica in that time. Only two resulted in home runs. Meanwhile, Castellanos barreled two home runs to center field at Wrigley Field in half a season last year. Those rates are a far cry from Corey Seager’s 11 home runs on 16 barrels to center field at Dodger Stadium during the same stretch.

Other notable parks include Fenway, which had the lowest HR/Brl% for righties, the second-lowest for lefties, and the fifth-lowest to center. This should end at least part of the debate for which direction is least favorable in Fenway. As it turns out, all fields there suppress home runs. It is no surprise, then, that Andrew Benintendi has been so underwhelming. Perhaps the reason is, over the last three seasons, he’s produced 10 home runs on 19 pulled barrels (52.6%) at Fenway. Similarly, in the other direction, Xander Bogaerts only managed 21 home runs on 31 pulled barrels (67.7%). Both rates are below the LHH and RHH means.

On a more positive note, Minute Maid Park appears extremely favorable to both right and left fields. Both were nearly two standard deviations better than their respective means for converting barrels into home runs. It doesn’t seem to matter whether you’re considering righties like Bregman and Gurriel, or lefty Michael Brantley, as none will have little in the way of resistance from their home park.

You may also be wondering how various offseason moves have impacted different players. In terms of home run hitting, I posit the following:

  • Mike Moustakas moving to Cincinnati makes little difference because Miller Park was also a home run hitting haven to right and center fields.
  • Anthony Rendon likewise nets out because, while Nationals Park is a better place to hit homers to left field, Angel Stadium is significantly more beneficial to center.
  • Yasmani Grandal, a switch hitter, gains value to right and left at Guaranteed Rate Field, but loses to center field relative to Miller Park.
  • Edwin Encarnación gains significant value both to left and center field moving from Yankee Stadium to Guaranteed Rate Field. The market may be underestimating the impact of his move.
  • Didi Gregorious, all of whose home runs in his career have gone to right field, loses substantially because Yankee Stadium is over a full standard deviation better to right than Citizens Bank Park.
  • Josh Donaldson loses some, but not much, value to left field going from SunTrust Park to Target Field, and loses even more to center.

Of course, these will vary based on the degree to which each hitter pulls his barrels relative to hitting them to center field. But I tried to account for that with my predictions. And you should consider other factors, like these players’ new lineups, spots in the order, divisions, etc.

 

Conclusion

 

I plan to update these park factors annually. Ballparks change, after all. The Rangers, for example, are moving to an indoor stadium, which should mitigate the effect of the heat and suppress home runs to a degree.

In addition, I encourage you to use this data to draw your own conclusions when evaluating players in drafts and not to worry too much about mine. And beyond this, I truly hope you find this data useful.

 

Featured image by Justin Paradis (@freshmeatcomm)

Dan Richards

Dan is a lifelong New York Yankees and Giants fan. A practicing attorney, Dan is better known for aggressively bothering his leaguemates about trades. You can follow him on Twitter @Fantasy_Esquire or by clicking the button above.

2 responses to “Going Deep: Barrels and Ballpark Factors Pt. II”

  1. Mark says:

    I may have missed it but did you take into consideration that the Giants are bringing in the fences for 2020 to generate more offense?

    • Dan Richards says:

      That’s a good question. I think the best response is in my conclusion, “I plan to update these park factors annually.” Given the Giants hadn’t moved the fences in before 2020 and my park factors are historical (2017-19), that change won’t be captured in them. To the extent it makes a difference, it will show up in the next iteration of these park factors.

Leave a Reply

Your email address will not be published. Required fields are marked *

Account / Login