Not true:
Also, just because it doesn't use common statistical tests as part of the analysis doesn't mean it isn't an accurate analysis.
Also not true:
There's more about opponent penalties but I figured you could go read it yourself.
I was talking about statistical tests to show what he was trying to prove, and he didn’t do any at all. The one you’re quoting was a correlation between time of year and penalties and he said himself it was part of descriptive statistics. It had nothing at all to do with the difference between Nebraska and the other teams. It was a baseline to show a trend in penalties over time in a season. It was to set up the actual topic he was getting to. I don’t consider correlation between time of year and # of penalties to be a statistical test for the difference in calls between Nebraska and other teams or bias against Nebraska, and it wasn’t supposed to be.
No, he did not analyze opp. penalties against Nebraska vs opp. penalties compared to other teams. He admits himself he didn’t look at it in a comment. The average differential doesn’t tell what I’m talking about. It doesn’t tell whether teams are getting penalized against Nebraska less than they are against other teams. You need to look at the games to see that.
A team that gets few penalties could be more biased against and this measure wouldn’t tell it. A team could have a -3 when they should have a -6 based on what they and their opponents normally do. A team with a -3 penalty differential could hypothetically have less opponent penalties called than a team with a +3 penalty differential. The opp. penalties vs Nebraska needs to be compared to opp. penalties vs other teams. The mere avg difference doesn’t give all of the information. I may have worded my statement poorly but that’s what I’m talking about.
To use an exaggerated example, if Nebraska gets penalized 20 times per game and their opponents average 5 penalites per game against them, they have a 15 penalty differential. But maybe all of those teams against each other are getting between 3 and 7 penalties. They are still getting called for about the same number of penalties as they were against Nebraska. To show what I’m talking about you’d need to show that the 5 is too small. Obviously the 20 is too big but it could be the team sucks a**.
Then you could have Penn State averaging 2 penalties per game and maybe against them teams get 3 penalties per game. But normally the opponents average 5 against other teams. Penn State is -1 but they’re possibly still getting screwed because they should be -3 if opponents were called for their avg # of penalties against PSU. (Although the difference btw -1 and -3 might not be statistically significant).
I never said the analysis wasn’t accurate. I said it’s not a statistical analysis, because it’s not. I even said it didn’t show further analysis wasn’t appropriate. It shows that it probably is. I might look at the opp penalties vs Nebraska compared to against other opponents myself to see if there is a big difference. Just another example, our difference in penalties was big vs NW and IIRC they had 1 penalty but they averaged 3.2 at the time so it may not have been a significant difference.
