September 26, 2016

It Is Time For Stormy Weather

By Bill Barnwell, Football Outsiders – special to BSMW Patriots Game Day

It Is Time For Stormy Weather

Oh, if only the Patriots were hosting the Chargers this weekend instead of heading to sunny San Diego. 62° is the high for Sunday in San Diego; Foxboro, a measly … 60°? El Nino aside, Foxboro’s expecting rain on Sunday; it’ll be bright and clear in San Diego.

I say “…if only…” because of the stories that would inevitably pop up if the Chargers were, in fact, traveling to a blustery Foxboro as opposed to the comfort out West. You know, the ones that would insinuate that Philip Rivers’ fingers would freeze off and that he’d be left crying in a ball on the sideline, quality control assistants cuddling with him, since he’d never been exposed to cold weather before? They always struck me as kinda ridiculous. It’s football! Sure, I don’t like playing football in the extreme cold very much, but I’m not a professional! They can figure it out, right!

What was given as proof of these statements, in the times where they were one, were stats about how the Buccaneers had never won a playoff game above 32°, or that the Packers had never lost one below it, or similar numbers. The figures were obviously a small enough sample and subject to enough bias that they were easily questionable.

So then, when I looked at the Patriots trip to San Diego, I wondered if the opposite was true; if warm-weather teams actually suffered when heading to colder ground, would cold-weather teams suffer when warmed up? Or, alternately, would they play better?

I took every playoff game from 1985 through 2005, outside of Super Bowls, and compiled the results of all 216 games. From there, I added the average January dry bulb and wet bulb temperatures for each city, home and away. In the case that a team played in a dome, I instead replaced the dry bulb and wet bulb temperatures with a dome constant that we’ve found at Football Outsiders to be the most accurate. This was done because teams who play in a dome, after all, would not be used to playing in the cold weather even if their city was cold outside of the dome (see: Peyton Manning).

I then separated the games into three groups. Cold games involved a road team traveling to play in a city where the average wet bulb temperature in January was 25 degrees colder than that of its home. Warm games were the opposite; they involved a road team traveling to play in a city where the average January temperature was 25 degrees warmer than that of its home — for example, the Patriots’ trip to San Diego. All other playoff games were adjudged to be Neutral weather games.

What I found? For one, that riding the winds to success has shown some validity for the playoffs:

weather_wins.jpg

The data clearly suggests that cold-weather teams hosting its warm-weathered brethren perform better than the average.

While the sample is too small to draw a strong conclusion, the data also suggests that the opposite isn’t true: in fact, warm-weather teams seem to be at a disadvantage when facing cold-weather teams!

There is always the possibility that what’s being produced in the data isn’t necessarily because of the separations being made — instead, the results above could have been produced by the cold-weather teams simply being more successful than warm-weather teams. Is that the case?

To check that, I’ll rely on each team’s Pythagorean Winning Percentage. (For more information, please click here.) The median dry bulb temperature for each of the 216 teams was 43°. I used that as the point of delineation in separating the teams into Warmer and Colder buckets. Their performance?

pythag_wins.jpg

As you can see, the warmer teams performed equally as well as the colder teams did, which would eliminate some fears about selection bias in producing the playoff performance results up above.

Is the sample large enough to be able to make strong statements about the weather and how it affects playoff performance? Probably not in the case of cold weather teams visiting warm areas, although it seems pretty apparent that the home teams in those matchups don’t enjoy any advantage by staying warm. When it comes to warm weather teams heading North and East, though, the last twenty years show that they’re most likely to be packing up their lockers Tuesday morning.

Leave a Reply