The title says it all. With Sunday's night's fumble fest in the books, I thought I'd take a quick and dirty look at how cold temperature affects fumbling. It was 19 degrees in Foxboro when I checked the weather there in the 4th quarter.
I looked at all plays from 2000 through the 2012 regular season, excluding kneel downs and spikes. I counted all fumbles, not just fumbles lost. Keep in mind the sample sizes greatly diminish at the temperature extremes.
Here is the breakdown:
It appears that very cold temperatures may have an effect on fumble rates. The average fumble rate was 1.79%. For extremely cold temperatures, the fumble rate was 2.42%, a 35% increase from the norm. There were 909 plays in the 2-11 degree bin, which means that the p-value was 0.08. That's considered marginally significant. But there are 11 bins of temperate, giving the results 11 chances to produce a type I error.
I looked at all plays from 2000 through the 2012 regular season, excluding kneel downs and spikes. I counted all fumbles, not just fumbles lost. Keep in mind the sample sizes greatly diminish at the temperature extremes.
Here is the breakdown:
It appears that very cold temperatures may have an effect on fumble rates. The average fumble rate was 1.79%. For extremely cold temperatures, the fumble rate was 2.42%, a 35% increase from the norm. There were 909 plays in the 2-11 degree bin, which means that the p-value was 0.08. That's considered marginally significant. But there are 11 bins of temperate, giving the results 11 chances to produce a type I error.
