[Husker] NU opponent penalties or 'something rotten in the state of Nebraska'

Thu Nov 18 09:57:32 CST 2010

Two other ways to look at the numbers would be to do a t-test to determine
if there is a difference between the average number of penalties called
against opponents when games are in conference (conference officials) vs.
out of conference (non-conference officials). And an A-B-A design (Do teams
have significantly fewer penalties called against them than their average on
the games they play Nebraska.)

Scott

On Thu, Nov 18, 2010 at 9:48 AM, Steve Schmadeke <husker at schmadeke.com>wrote:

>
> "To what else could it be attributable?"
>
> There are some other possibilities besides the obvious.  Nebraska runs a
> different scheme on both sides of the ball than the typical team.  We are,
> by and large, not a pressure defense and our defensive lineman have been
> schooled in different techniques for gap control than the typical defense,
> so we may not be drawing as many holds as the typical team -- though that
> was a whopper the refs missed on Steinkuhler last Saturday night against
> Kansas.  On offense, we are primarily a running team and therefore not as
> likely to draw as many interference and illegal contact penalties -- or even
> roughing calls.  Our rush-oriented offense combined with a
> bend-but-don't-break defense may be resulting in fewer snaps per game,
> reducing the amount of opportunities for refs to throw flags.
>
> I don't know if any of these explanations are true.  To analyze those
> possibilities would require breaking down the penalties by type and
> situation and to normalize them according to the numbers of snaps in each
> game.  Breaking down the analysis into smaller data sets might also drop us
> down too far into the statistical noise.  While Paul's analysis shows a
> clear, statistically significant bias (in the mathematical, not
> conspiratorial sense) in the total number of calls made against Nebraska's
> opponents, there may not be enough data to draw significant conclusions
> about, say, non-procedural penalties called per offensive pass play.
>
> On Nov 18, 2010, at 5:40 AM, Paul Dalen wrote:
>
> > I'm a member of HuskerMax.  There was a thread yesterday about how NU had
> > the lowest average opponent penalties in the B12.  I did some number
> > crunching on it.  Here's my post to that topic.  I can't post pics to
> this
> > list, so you'll have to click the links to see the graphs.  I did the
> work
> > pretty quickly last night, and I may have takes some *ahem* liberties
> with
> > assumptions.  But in general, I think my analysis is sound.
> >
> > BLUF:  There are some statistical anomalies when you compare NU’s
> penalties
> > to the average of the rest of the B12’s opponents’ penalties, and when
> you
> > compare NU’s opponent penalties to an average of the rest of the B12
> > penalties.
> >
> >
> >
> > Note:  None of the averages I used in analysis include NU’s numbers.
> All
> > stats are from 2010.  I'm assuming that the data used conform to a normal
> > distribution, which may or may not be true.  The data I used are found
> here:
> >
> >
> http://www.teamrankings.com/college-football/stat/opponent-penalties-per-game
> >
> >
> > First, I compared NU’s per-game team penalties to an average of the rest
> of
> > the Big 12 opponent per-game penalties.  NU’s average per game penalties
> are
> > 7.1, and the rest of the Big 12’s average opponent per game penalties are
> > 6.69 with a standard deviation of .92 (n=11).
> >
> >
> > NU’s average of 7.1 converts to a z-score of .44, which equals a
> probability
> > that any random team would have a penalty average of no more than this of
> > .67.  To me, this says that there is no evidence at all that the refs are
> > calling more penalties on NU than they are on the rest of the B12 in
> > aggregate.
> >
> >
> >
> > The graph below illustrates this probability.  The top of the bell curve
> is
> > the average number of Big 12 opponent penalties, and the line between the
> > blue and white areas is NU’s per game average penalties.  The closer to
> the
> > middle of the graph the line falls, the closer to ‘normal’ the average
> is.
> >
> >
> >
> >
> http://lh4.ggpht.com/_elioP_aq1UY/TOSfdmk1WMI/AAAAAAAADP8/4d70P0MpyqI/s800/part%201%20penalties%20analysis.JPG
> >
> >
> > The numbers become more eyebrow-raising when you compare NU’s  average
> > opponent penalties per game to the average per game penalties of the rest
> of
> > the Big 12.  NU’s opponents’ average per game penalties are 4.3, and the
> > rest of the Big 12’s average per game penalties are 6.59 with a standard
> > deviation of 1.21 (n=11).
> >
> >
> > NU’s average of 4.3 converts to a z-score of -1.90, which equals a
> > probability that any random team would have a penalty average of no more
> > than this of .029.  To me, this says that there is evidence that the refs
> > are calling significantly fewer penalties on NU’s opponents.
> >
> >
> > The graph below highlights this.  The shaded blue area is the probability
> > that a random team’s opponents would be assessed an average of 4.3
> penalties
> > per game.    As fan and as a stats nerd, it takes significant suspension
> of
> > disbelief to accept that this average is attributable solely to chance.
> >
> >
> >
> http://lh4.ggpht.com/_elioP_aq1UY/TOSfdmk1WMI/AAAAAAAADP8/4d70P0MpyqI/s800/part%201%20penalties%20analysis.JPG
> >
> >
> > There is one other thing to consider.   The Big 12 averages include four
> > teams that the Huskers have not faced yet, and these teams include the
> three
> > teams (CU, BU, A&M) with the highest per game average penalties.   If we
> > perform the same calculations but exclude the teams NU has not played
> yet,
> > the numbers become even more eye opening.
> >
> >
> > Comparing NU’s per-game team penalties to an average of the rest of its
> Big
> > 12 opponents' per-game penalties, NU’s average per game penalties are
> 7.1,
> > and the average of the rest its Big 12 opponents' per-game penalties are
> > 6.69 with a standard deviation of 1.03 (n=7).
> >
> >
> > NU’s average of 7.1 converts to a z-score of .40, which equals a
> probability
> > that any random team would have a penalty average of no more than this of
> > .66.  Again, there is no difference in the number of penalties called on
> NU
> > even when you only consider the teams NU has played.
> >
> >
> >
> > However, when we compare NU’s  average opponent penalties per game to the
> > average per game penalties of the rest of its Big 12 opponents we see
> that
> > NU’s opponents’ average per game penalties are 4.3, and the rest of the
> Big
> > 12’s average per game penalties are 6.04 with a standard deviation of .61
> > (n=7).   NU’s average of 4.3 converts to a z-score of -2.85, which equals
> a
> > probability that any random team would have a penalty average of no more
> > than this is .002!
> >
> >
> >
> http://lh5.ggpht.com/_elioP_aq1UY/TOSftvXGtPI/AAAAAAAADQc/JxJ4c1fDv08/s800/part%203%20penalties%20analysis.JPG
> >
> >
> > While the number of penalties called on NU’s opponents seems hard to
> accept
> > as coincidence when compared to the entire Big 12, it’s almost 10 times
> more
> > likely to happen than when you look at only NU’s opponents!  In fact, the
> > probability is so small as to be almost invisible on the graph.  Trust me
> > though, there is a very small blue shaded portion above.
> >
> >
> >
> > What’s this all mean?   To me, it’s clear that the per game average
> number
> > of penalties called on NU’s opponents is very difficult to attribute
> solely
> > to chance…which begs the question of “to what else could it be
> > attributable?”  I leave it to each of you to draw your own conclusions
> about
> > that…as we’ve left the realms of science and entered the more treacherous
> > realms of conspiracy theories.
> >
> >
> > Full data and computations are here.
> >
> >
> >
> https://docs.google.com/leaf?id=0By3lZgFCOqcjNzZkNmViZDctMTgxZS00ZDk4LTg3ZGMtNWZkMmNiZDcwMTU4&authkey=CMz19e4C&hl=en
> >
> >
> > Thread on HuskerMax is here:
> >
> >
> >
> http://www.huskermax.com/vbbs/showthread.php?10074-NU-s-oppoents-least-penalized-in-Big-12
> .
> > _______________________________________________
> > husker site list
> > husker at tssi.com
> > http://romaine.tssi.com/mailman/listinfo/husker
>
>
> _______________________________________________
> husker site list
> husker at tssi.com
> http://romaine.tssi.com/mailman/listinfo/husker
>