Park-Factor Update: Summer, Heat and Fly Balls

Way back in May, I wrote a piece here on early-season park factors. Based on 26,650 balls-in-play struck through May 11, park factors were calculated based on granular exit speed/angle data. Yes, the sample was fairly small, but some interesting data was generated. One conclusion reached was that weather seemed to be playing a fairly significant role: the upper Midwest and Northeast corridor clubs with open-air stadiums endured cool, wet springs which had a clear run-suppressing effect.

The article wrapped up by indicating that we’d check back in a couple months to see what effect the higher temperatures of summer would have on those park effects. And here we are. This week and next, we’ll update these park factors through late August. Today, we’ll focus on fly-ball park factors, and next week we’ll take separate looks at AL and NL overall park factors.

First, here’s the quick-and-dirty on the method. Through August 21, 106,962 balls were put into play during MLB regular-season contests. They resulted in an overall batting average of .328 and slugging percentage of .537, up from .323 AVG-.523 SLG on May 11. Fly balls generated a .328 AVG and .895 SLG, way up from .314 AVG-.839 SLG on 5/11. Line drives generated a .661 AVG and .872 SLG, fractionally down from .662 AVG-.874 SLG on 5/11, and ground balls a .237 AVG and .258 SLG, up from .227 AVG-.249 SLG on 5/11. (Oh, and pop ups have generated a .018 AVG and .028 SLG.) Each BIP type was split into “buckets” separated by 5-mph increments. The top fly-ball bucket begins at 105 mph, and the top liner and grounder buckets begin at 110 mph.

For each ballpark, the actual production derived from that park’s actual BIP mix was compared to the projected production, assuming that each BIP bucket generated MLB average production for that BIP type/exit-speed combination. Convert everything to run values, and voila, park factors, both overall and by BIP type.

Fly-ball production sure has spiked upward from 5/11 to 8/21. Let’s take a quick look at the production by BIP type for the interim period:

We were clearly in a different world with regard to fly-ball production once the weather got warmer throughout the country. Let’s drill down a little bit to see where, exit speed-wise, that production increase is coming from. If you hit a fly ball at 105 mph or higher, you’re golden; through 5/11, such fly balls yielded a massive .920 AVG-3.424 SLG. Through 8/21, that’s up a tad to .937 AVG-3.528 SLG. From 100-105 mph, we’ve gone up from .631 AVG-2.122 SLG through 5/11 to .681 AVG-2.345 SLG through 8/21. Pretty significant increase, but hitters were in a pretty good place in both spring and summer on such fly balls.

Fly balls at 95-100 mph is where the action is. Through May 11, hitters batted an adequate .260 AVG-.778 SLG on such batted balls, with 12.0% of them going over the wall. Through August 21, they are batting a cumulative .321 AVG-.967 SLG on 95-100 mph fly balls, with a 15.5% homer rate. That’s a pretty huge production increase, and it stands out more starkly when you look at production on such fly balls in the interim period from 5/11-8/21: .341 AVG-1.028 SLG, with 16.6% of such fly balls going over the wall.

This is a pretty big deal. Fly balls at 95-100 mph are the ones that leave the park in summer — and tend to stay inside of it in spring — in most ball parks. Some players (Brian Dozier and Ian Kinsler, I’m looking at you) hit a ton of fly balls in this range. And some ball parks treat such fly balls much more kindly. Individual fly-ball park factors, as well as actual fly-ball production levels and homer percentages, for all 30 parks are listed below:

As I hypothesized in my May article, the fly-ball park factors of the upper Midwest and Northeast clubs have almost universally skyrocketed since May. Check out Pittsburgh (up from 101.5 to 124.1), Cleveland (72.2 to 119.3), Chicago White Sox (87.9 to 110.8), Minnesota (91.6 to 105.5), NY Mets (78.1 to 101.9), Washington (73.1 to 89.9) and Chicago Cubs (64.8 to 89.0) in particular. Also, as expected, the park factors for indoor and indoor/outdoor stadia such as Milwaukee (down from 133.4 to 115.7) and Tampa Bay (107.9 to 83.1) have declined fairly sharply in relation to their weather-affected peers. This is a zero-sum process after all: raw fly-ball production barely moved in Tampa, but in relation to league average it took a big hit as the summer progressed.

Lastly, let’s take a look at that key 95-100 mph fly-ball bucket, and how production on such flies in all 30 venues was impacted by the summer warm-up:

Again, the biggest differences in production came in a handful of ball parks in the upper Midwest and Northeast. Cleveland is simply astonishing: .220 AVG-.585 SLG on 95-100 MPH fly balls through May 11, .419 AVG-1.206 SLG through 8/21. The homer rate on such fly balls in Cleveland nearly tripled from 7.3% to 18.8% by late summer. In Washington, the homer rate on 95-100 mph fly balls nearly quintupled from 3.1% to 14.7%, with production on such flies spiking from .125 AVG-.344 SLG to .309 AVG-.919 SLG. The production increases and homer rates on such fly balls also increased very sharply in Boston, Baltimore, Minnesota, Pittsburgh and both Chicago ballparks, among others.

What does all of this tell us? It simply reinforces the importance of context. Nolan Arenado is a great baseball player, but he is not an elite offensive player. He is helped greatly by Coors. Miguel Cabrera is an even greater hitter than he seems: Comerica Park deflates his actual production somewhat considerably. All players’ production is part true-talent, part context. It is the job of talent evaluators to separate the two.

While I would certainly not attempt to debunk the very credible claims regarding a potentially juiced baseball, I would remind everyone that weather has had a great impact on the increased run production we have seen in 2016. Now, let’s strap in and enjoy the stretch run, during which the temperatures will (eventually) cool and likely drain a little offense from the game.