Why the Irish have a good shot at a W against OU:
One way to look at whether or not the Irish have a chance, is to look at
the statistics in regards to how well a team does against what their
opponents normally give up. It is kind of useless to look at a team's
total scoring average without factoring in the strength of opposing
defenses. In ND's case, we have faced three top 10 defenses (by yards)
so far this year (BYU, Michigan, MSU, with Stanford joining them as a
top 24 scoring defense nationally).
I think to be accurate you can't just look at points plus or minus
versus points given up. My preference is to use percentages to capture
the impact of those strengths.
Under that basis, ND's defense is holding opposing offenses to 33.5%
of their normal total points scored. So if you instead look at it that
way, Oklahoma should score about 15 points against the Irish. The other way to
analyze it is based on the strength of OU offense. How many more
yards does OU offense score than what opposing defenses give up. Now
if you look at OU offense in comparison to the strength of the
opposing defense, you see that Oklahoma tends to score about 72% more
points than the opposing defense normally gives up. So based on that
metric, they should score about 16 points against ND, based on 9.43
points per game that the Irish currently give up. Interestingly, these numbers
are actually quite close to each other.
Now using the same formulas in regards to ND, Oklahoma is holding
opponents to 50% of their scoring average, which would mean they should
hold ND to 13 points. However, on average ND's offense scores 24% more
points than opposing teams on average normally give up. So based on
that fact, ND should score around 19 points in this game offensively.
So if you average the two approaches out for each, you would get
15+16=31 divided by 2 or Oklahoma scoring 15.5 points. While ND is
13+19=32 divided by 2 or ND scoring 16 points.
I think that both observations are somewhat backed up by empirical
data. My opinion is that points mean a lot less against the really bad
teams in college football, while against the really good ones,
especially defenses, you get less variance and chances to "run up the
score". It also means less points and turnovers from getting outmatched
teams out of their comfort zones and playing one dimensional football
to try and catch up. Hence it makes sense to weed out the weak sisters
from reviewing opponents. In this veign, we should be able to look at
just the very top defenses each team has played to get a feeling how
each team will do against a good all around team, top 15 defenses in
total defense or top 15 in scoring defense (Oklahoma hasn't faced a
defense that is in the top 20 of total defense, but KSU is #23 and #14
in scoring defense, so we will have to make due).
We have three examples of ND playing top 10 total defenses, and ND
scored 20, 13, and 17 points against those defenses that give up 15, 16,
and 14 points per game. Just based on those top defenses, we scored
+5, -3, and +3 points in those games (or roughly +3). Oklahoma is
giving up 15.33 points per game, so it is reasonable to think we will
score 18 points against Oklahoma's defense. In contrast, Oklahoma has
played one top 15 scoring defense in KSU (they are actually ranked 23rd
in total defense for yards given up). They scored 19 points against
KSU, or +3. Granted, the sample size is really small, but that would
seem to indicate that they would only score about 12.5 points against
us. Note, in this case I am using actual points because the difference
in the base is so small, that percentages don't make sense.
Statistically from a number of different approaches, it seems to be
almost a dead draw/coin flip with ND having an ever so slight numerical
advantage.
In other words, ND has a much better chance than the national
prognosticators are giving them, or even Vegas at this point.
ND plus 10.5 or more if the line goes up!!!!
GL
thats funny. As a ND fan, I cringe having to hear Holt's homerism 
Hot Forum Topics
