Quote:
Originally Posted by cdcox
These numbers aren't perfect, but they can give us a quick look. I generated them doing a quick spreadsheet analysis of NFL play-by-play data.
Out of 6465 drives in 2012, the number of penalties per drive (against both teams, both accepted and declined) breaks down like this:
0 penalties: 4199
1 penalty: 1595
2 penalties: 490
3 penalties: 139
4 penalties: 32
5 penalties: 6
6 penalties: 4
So Zeke's case should be possible, but rare.
We can also do a rough estimate of what the distribution of penalties per drive should look like if the NFL made all their penalty calls randomly. Let's take the average number of NFL plays per drive 6.57 and translate that into an assumption that all NFL drives have exactly 6 plays. We'll also assume that the chance of a penalty on any given play is about 0.075. If all penalties occurred randomly we would expect the following frequency of penalties per 6 play drive:
0 penalties: 0.638
1 penalty: 0.304
2 penalties: 0.061
3 penalties: 0.006
Compared to actual 2012 NFL frequencies
0 penalties: 0.649
1 penalty: 0.246
2 penalties: 0.076
3 penalties: 0.021
So NFL refs tend to bunch penalties on a given drive more than one would expect if penalties were completely at random. This might be related to sloppiness of a particular team or make-up penalties. Of course the relatively small difference between NFL rates and my ideal rates could be tied to the assumptions I made in the analysis.
Probably only Rain Man and I will find this all fascinating.
|
I find it very interesting as well. Also, the tendency to not call penalties on end-of-game drives would add to the grouping of penalties.