DBB's Statistical Breakdown
You've seen my breakdowns of the games...
Here is the breakdown of this "controversy"
Let's look at how well each team did vs. their opponents' average output both offensively and defensively. Then we'll use those numbers to see which team might be expected to win a head-head contest (does that even matter to the Pac 10?). Finally, I'll add my take on the numbers and what they mean.
Recall that
Differential YPG/PPG is a measure of how many more yards/points you produced than what your opponent gave up on average against their other opponents (their average with the game against you factored out).
OFFENSE
_________________Differential YPG_______Differential PPG
OHIO STATE__________19.03________________7.57_____
Notre Dame__________108.64_______________14.72_____
Oregon_______________62.64_______________11.54_____
OK, it looks like ND has the best offense of the group; with OSU respectable but still pulling up the rear. No surprise there, but it doesn't tell the whole story.
DEFENSE
_________________Differential YPG_______Differential PPG
OHIO STATE_________171.86_______________18.06_____
Notre Dame___________25.25________________6.40_____
Oregon_______________70.58________________8.21_____
Um... WOW. That's not close. OSU has a much better defense than either team. Again, not really a surprise.
MATCHUPS
OSU vs ND
Matching avg offense vs differential defense suggests:
OSU Wins 26-20 (379 yards to 300 yards)
Matching avg defense vs differential offense suggests:
OSU Wins 30.47 to 29.54 (So close I have to show 4 digits) (400 yards to 384 yards)
OSU vs Oregon
Matching avg offense vs differential defense suggests:
OSU Wins 24-18 (334 yards to 277 yards)
Matching avg defense vs differential offense suggests:
OSU Wins 31-26 (376 yards to 338 yards)
Analysis
So Ohio State is the better team when you do the RAW analysis. But there is a hole in this type of analysis that I didn't really get into during the Big 10 season. During league games, this "hole" isn't very big; but when comparing teams from different conferences (or no conference) it can be huge.
The "hole" I speak of (something only Mililani pointed out) is that this analysis compares you to everybody else on your opponents' schedules. But what if your opponents played teams that built strong numbers by playing weak schedules? Then when they play you, it is easy to hold their numbers way down. In other words, what if your opponents' opponents sucked?
So to balance this analysis, you have to compare
not your schedule but your
Opponents' Schedules to the other team's Opponents' Schedules. Again, in league play, this isn't usually a big deal. But when comparing OSU to ND and Oregon, it's a very big deal.
I decided to look at opponents' schedule in two ways.
- Raw winning %
- Sagarin Strength of Schedule Ranking
I include the Sagarin numbers because I believe strongly that a weak conference can artificially inflate SOS numbers, both for you and your opponents. I include the Raw numbers because the Oregon and ND fans won't like the Sagarin ratings, but the Raw numbers STILL show Buckeye dominance.
Opponents Schedule Strength
_________________Raw % avg rank________Sagarin avg rank
OHIO STATE___________31.5___________________15.5_____
Notre Dame____________39.2___________________32.5_____
Oregon________________45.9___________________46.2_____
Let's TRY to wrap our minds around how stark a contrast this is.
- Differential Numbers compare you to the rest of the teams on your opponents' schedules.
- OSU is therefore compared against a MUCH better set of teams via this analysis than are ND and Oregon.
- OSU came out on top ANYWAY.
- To Repeat: Compared Against a Better Standard - Look Better Anyway
So if it matters which team is better; the clear FIRST choice for the FIESTA BOWL is Ohio State. Who they choose with their second choice is their business.
They're going to get their ass kicked anyway.