As noted here and elsewhere this past week, Rangers pitcher Derek Holland drew some attention with a quote regarding the role of the August heat and humidity in his last outing at The Ballpark. While it's been rare for players to comment on the heat as a factor, I think this theory holds a spot in the top five of the Black Helicopter Theories Of Baseball Park Effects. Lately I've not been able to listen to Rangers Replay or visit the comments sections at the Dallas Morning News' Rangers blog without hearing or reading several comments about how the Rangers can't win as long as they don't have an air conditioned dome because it's too hot in Texas in the 2nd half of the season.

Like any good scientist (or in my case, recreational social scientist), I wanted to poke this theory with some evidence and see what it would do. The boiled-down gist of the "It's Too Hot To Win" Theory seems to be that the Rangers lose in the second half of the season (another popular phrase is "second half swoon") because the temperatures in Arlington are closer to the third ring of Hell than they are to most of the rest of North America for the majority of 2nd half home games.

If the theory is correct in the grand scheme of things, the Rangers ought to consistently have losing records in the second half of any given season at home. I did a quick count of the Rangers' second half home record from 2002-2009 (my numbers may be off a couple games here and there...it put the "quick" in quick count) from ESPN's team summary page.

In three out of the last seven seasons the Rangers have a losing record at home after the All Star Break. While that isn't great, it's also clearly not a consistent record of losing in the second half. For the two extreme seasons of second half home wins, I also looked at the first half home record to see if there was some kind of significant boost or drop off from the 1st half home record to the 2nd half home record. The anecdotal answer is mixed. The 2004 Rangers were a pretty consistent team at home across both halves of the season, indicating that they were a pretty solid squad. The 2008 Rangers did experience a 12 games drop off from their first half to their second half. They were treading water at home in the 1st half and were simply awful at home in the second half. Strength of schedule? Injuries? Heat? The won-loss record doesn't tell us those things.

Since this is just a descriptive look, we can't convincingly give a thumbs up or thumbs down to the "It's Too Hot To Win" Theory based on these results. I think, though, we can get an idea that no, it's not simply a matter of it being too hot to win in Arlington in the second half of the season. I doubt that this is really news to most folks here at LSB.

This first look is purely descriptive and that to properly address this is going to take some serious time and effort to collect data on pitchers, game time temperature and opposing offenses and to specify, test, run and interpret a multilevel model with random effects. As with most social science, the modeling is the easiest part; the data collection is by far the most difficult (panel data on every home starter in the time period). In the case of this question, the data is there, but I don't have the time to grab it in short order. I am working on collecting this data and modeling it and when I can finish this project, I'll share it with y'all. If anyone knows where I could get most of this data already in some kind of spreadsheet or spreadsheet-portable format (like CSV), I'd be most appreciative.

I'm going to keep chipping away at this because I think it's an annoying but interesting question and a question where there's the appropriate data to analyze it with an appropriate method.