More on Cheney Stadium's Fences

I follow Jeff Sullivan on Twitter because he's the main writer for Lookout Landing, and a mostly neat guy. I mean, sometimes he goes too long between trimming either his face hair or head hair and he really needs to wash out those pint glasses more frequently because sometimes they get a little beer ring on the bottom, but overall, he's well above average for neatness. I do not follow every semi-neat guy on Twitter since I myself am a semi-neat guy and I do not follow my identical twin. In fact, I follow very few actual people. I get just enough Mariners baseball news in order to skim from the work of others and the rest is a bunch of streams that betray my nerdy interests. Also, @EmergencyPuppy. Still, I do follow every quasi-neat guy who is also a writer for this website that I am writing these words for.

Sullivan has this blog and on it he wrote a piece that linked to Mike Curto's blog, who I also follow yet cannot faithfully comment on his level of tidiness. The subject was the number of home runs hit that cleared Cheney Stadium's new nine-feet-tall walls that Curto believes would've bounced off the old 17-foot height. I read both entries and let them settle into my mental digestive tract, not expecting to see them again for about 24 hours when I'd grab a beer to stop the realization of everything's eventual mortality from sending me into an existential crisis. Instead, the entries nagged at me. Curto concluded that 47 extra balls cleared the fences last year due to the lower walls. Forty-seven extra home runs seems like a lot. It is a lot, but how much is that really? Well, yeah, 47, I know, but how much is 47, relatively? Eventually I gave up trying to ignore those questions and began constructing the numbers.

There were 72 games played at Cheney Stadium in 2011, which, averaged out, means about two out of every three games featured a dinger that pre-renovations Cheney would have kept in play. That is quite frequent. You would notice an additional 53 home runs hit at Safeco Field should you be the sort of person who still attends, watches, listens to or reads about Seattle Mariners games. The Rainiers and their opponents combined to blast 161 home runs in Tacoma so 47 represents 29% of that total, a remarkably high percentage.

Offered alone, all that is merely descriptive. It's some context surrounding the 47 number. Measuring 2011 against the most recent three-year sample beforehand gets us comparative notes and those are potentially illuminating. Curto points out that the period from 2008-10 at Cheney saw an average home run rate of 2.03. In 2011, if those 47 home runs were wiped away from the ledger, and assuming that change causes no additional ones,* then the home run rate drops from 2.24 per game to 1.58. That would be the lowest rate since 2006. Such a massive drop is fishy and propelled me to dig further.

*That is, assume the home run being turned into a single or double instead does not alter the chain of events to make a future one more or less likely. Testing that assumption would make for an interesting read. For now it is beyond my level of interest and availability of time.

The first possibility I thought of is that the home run hitting prowess of the batters turning through Tacoma was severely docked, but masked by the lower fences. The data does not support that though. The home run rate for the league as a whole actually went up a bit, even if you toss out the games at Cheney. If the league's homer-happiness didn't dramatically change, you'd expect the home run rate to stay in the ballpark of constant outside of changes to the park and yet Cheney's increase from 2.03 to 2.24 jacks per game approximates fifteen (72 games times 0.21 additional HRs per game) homers.

For a more detailed comparison, I turned to the method that I use for computing park factors. In those, I compare all the resulting events at one park with all the resulting events involving that park's team's away games.* The theory is that formula mostly balances out the individual hitters and pitchers, leaving mostly just park residue behind. Those numbers agree that the relative home run rate increased in Tacoma in 2011, but based on the balls in play, suggests that should the park factors have stayed the same**, only ten fewer dingdongers would have been hit.

*A fuller description available here.
**

This is where it gets murky. Curto makes an assertion which requires data I don't own to refute. It looks like I might, but home run rates and park factor models don't contradict a disproportionate increase in wall-scraping homers. Curto actually watched the home runs, which is the direct sort of proof we'd want for a declaration of how many extra there were. If hit tracker data was available that would be a more suitable model to test his figures because it attempts to answer the same question. My methods deal with a holistic system of park factors though.

Three options came to my mind to reconcile the discrepancy between Curto's in-person though subjective finding and my interpretation of what the overall numbers indicate. The first possibility is that Curto is wrong. In the past, I might have simply concluded that and left it there. I know Curto though and I believe him when he writes that he took a conservative approach to marking new homers and I trust his ability to make that judgment based on his years at Cheney. I like to believe that I am amenable to alternative views now, but the truth is that the source on the other end is probably what's backing me into that corner unwillfully, which, good because it forces me to delve into the nuance here.

Of course, there exists the non-zero possibility that I am wrong either in my model or in my interpretation. I do not think that's a high probability (duh). I came at it several different ways (including ones not written here) and all led me to a bump worth around 10-15 roundtrippers. Still, it's possible. I've been wrong before. Once, back in '49 when Jules asked me if the river ahead was the Reno, I thought it was, but it turned out to be the Rubicon. Whoops! Boy was there egg on my face.

If neither of us is wrong about our particular area of the discussion then that leaves an unsatisfying mess still present. Can Curto be right that 47 balls cleared Cheney's fences that would not have in 2010 and I be right that the home run rates in Tacoma and the rest of the PCL say that number should be closer to 10? Yes; the two are not perfect contradictions.

For instance, suppose the lower fences at Cheney changed the interior wind patterns or strength, knocking down balls in the air. Projectiles that, protected from the wind, might have sailed easily over the fences are now exposed to additional drag and perhaps barely clear the new fence line. That imaginary ball is a home run in either park and yet not looked at that way because a secondary factor, wind, was involved.

That's merely one hypothetical explanation. I could perhaps think of others until I got sidetracked building out a rudimentary physics model to see if the above was plausible at all. That was a waste of my time because complications like coefficient of drag, Magnus force, air resistance and such are difficult and important. If we ever play baseball in a frictionless vacuum, I got it covered and in which case whoa, wind can matter. Also, the players are all dead from lack of oxygen.

It would be great to end with an answer or something approximating an answer. I don't possess one though. The reality is that the renovation could cause many more fringe home runs and not many more overall home runs from occurring. It's weird to think that shortening the fences could have such an impact but real life systems are complex. I hope that refrain strikes you as familiar because it's basically the caveat for every problem. Actually simple things are rare and actually complicated things are worth acknowledging.