well done...i still think we win ND28 - OSU24
Nick | Email | Homepage | 12.26.05 - 11:52 am | #

Time to start tracking prediction accuracy. USAToday (http://www.usatoday.com/sports/college/columnist/ weekendforecast/2005-12-19-column_x.htm) is predicting a 24-21 OSU victory.

If (*if*) we can stop Troy Smith's running, I think we can do it. Many of you know football better than me, but our outside linebackers and their ability to make tackles could be the key to the game.
Dad the Domer | Email | Homepage | 12.26.05 - 3:15 pm | #

Jay, you completely forget the W-factor which is good for at least 14 points. Notre Dame 38, Ohio State 29.

It's not going to be as close as people think and Notre Dame will finally shut the media up for a little while.
Irish '81 | Email | Homepage | 12.26.05 - 7:12 pm | #

This one comes down to Special Teams and turnovers. ND wins the turnover battle and special teams may be a push.

Tressel is good at bowl games and tOSU are veterans at the Fiesta bowl, and Charlie Weis has never taken a team to a college bowl game. Advantage Tressel.

In game adjustments, advantage Weis.

That's 2 for ND and 1 for tOSU. Hence, ND wins by 3-7 points.

ie Somewhere between, ND 27-tOSU 24 and ND 31-tOSU 24.

Go ND.
atepesm | Email | Homepage | 12.26.05 - 10:00 pm | #

Weis may not have taken a team to a Bowl game, but he does have 4 Super Bowl rings. ND in a blowout, give him a month to prepare, it shouldn't even be close.
Matt | Email | Homepage | 12.26.05 - 11:52 pm | #

Don't be naive to think that this game will simply be a blowout. Nobody has "blown out" OSU this year. Tressel is a good coach with a lot of experience. They have a lot riding on this bowl as well. They can be the team that will make the media say "Same old ND, can't come through in the end". I'm sure they'd like to be the team to make the media question ND again. I still think ND wins, but both teams have a lot to play for in this game. It won't be a blowout.
ND101 | Email | Homepage | 12.27.05 - 9:52 am | #

ND wins 4-2 on a last minute safety.
Yago | Email | Homepage | 12.27.05 - 9:53 am | #

Prediction: Due to tOSU's fantastic defense, Charlie's boys are held to a mere 750 total yards, and only 58 points.
Brad | Email | Homepage | 12.27.05 - 2:38 pm | #

Interesting work; here's an observation. OSU has three "outlier" wins of 31, 38, and 41 points (vs. IN, IL, and NW, resp.). Those disparate scores explain most, if not all, of the 4+ point difference you have derived here. Moreover, those three teams were 6-19 in conference and had only one win was vs. a "good" opponent (Rutgers in IL's 1st game). The game vs. The Irish may be closer than you predict.
Brian Murphy | Email | Homepage | 12.27.05 - 2:40 pm | #

Very good analysis. One thing it doesn't adjust for is the difference in schedule. ND and OSU have two common opponents in Meatchicken and Meatchicken State. However, this leaves a significant lack of samples. I did the analysis anyway.

MSU vs. OSU D diff - -9.82
MSU vs. OSU O diff - +6.27

UM vs. OSU D diff - -7.82
UM vs. OSU O diff - +5.73

MSU vs. ND D diff - +4.18
MSU vs. ND O diff - +9.72

UM vs. ND D diff - -18.82
UM vs. ND O diff - -2.27

OSU D avg - -8.82
ND D avg - -7.32

OSU O avg - +6
ND O avg - +3.5

OSU PPG offense for MSU, UM = 30
ND PPG offense for MSU, UM = 27.5

OSU PPG defense for MSU, UM = 22.5
ND PPG defense for MSU, UM = 24

OSU o w/ ND d = 22.68
ND o w/ OSU d = 18.68

OSU d w/ ND o = 26
ND d w/ OSU o = 30

OSU Score – 26 (26.34)
ND Score – 22 (22.34)

Still the same outcome, but predicting a closer game. The pros of this analysis is that the data is against common opponents, thus not influenced by cupcakes like IN, IL, Navy, and 'Cuse. However, it means that all the data comes from four games, most of them with weird goings on; Henne's suckage in the UM-ND game, the halftime meltdown at OSU-MSU, the pass-wacky MSU-ND game. The only mostly-normal game in that set is UM-OSU. So the numbers are suspect, but in a different way.

The teams appear to be fairly equal. The game is going to be decided by unpredictables: playmakers, which both teams have in abundance (Brady/Walker/Shark/Zibby and Smith/Ginn/Holmes/Hawk), turnovers, last second heroics, and coaching. I have to think that ND's got the advantage in those catagories from what football I've watched this year. It's going to be close. Hopefully the luck of the Irish is present.

Biased Prediction: 27-24 Irish: ND comes from behind on one of those patented 2-Minute drills observed at USC and Stanford.
tedwick | Email | Homepage | 12.27.05 - 6:02 pm | #

Does anyone know how the oddsmakers really come up with the lines for the games? The analysis done by Jay may come close the line determined by the brethren in Vegas, but is this a coincidence? I had delusions that the odds were somehow correlated with Sagarin's Pure Points Predictor, but now that I look at some of the odds on the games, I would have to say I was incorrect. The difference in the Pure Points Predictor for tOSU and ND is 4.73, which is pretty close to the 4.5 current line. What makes me think I am really wrong on this is that Miami is giving 7 to LSU (per Vegas), yet the difference in their Pure Points Predictor is only 2.5.
snark | Email | Homepage | 12.27.05 - 6:07 pm | #

Does anyone know how the oddsmakers really come up with the lines for the games?
PantsB | Email | Homepage | 12.27.05 - 6:43 pm | #

Nice analysis. It almost made sense.

BTW, is there something to be said for the role of ND/Weis as "destroyer of seasons?" Think about it - teams like Pitt, Purdue and Michigan all collapsed against expectations after being dissected by the Irish. Weis is pointing out fundamental flaws in his opponents that others are others are capitalizing upon - and hurting those opponents' strength of schedule, points per game, points allowed, etc.

Even USC - after getting the scare of their life, they get a huge scare from Fresno. Tell me with a straight face that the film study of how Weis approached USC didn't affect the Fresno planning. (And tell me that Mack Brown hasn't memorized that game by now.) I submit that ND/Weis taking USC to the limit with superior scheme and inferior talent made USC eminently beatable. Perhaps that's why USC is mailing DVDs of the ND-USC game to their season ticket holders...it's the last hurrah.

I know it's blind loyalty, but I really think that Weis and his staff are that much smarter than the rest of the college football world.

(I also think that Tressel is not too smart but able to play predictable football with excellent talent. Run the ball, control the ball, stifle on defense. Tressel is not smarter than Weis. ND should win.)
IrishAlum | Email | Homepage | 12.27.05 - 10:57 pm | #

HOW the oddsmakers really come up with the line for the game? I have no idea. WHY they come up with the line that they do? Simply to balance the wagering. That's it. Their only objective is get equal money bet on both sides of the game. That way the "house" wins, no matter the outcome of the game. As game time approcahes, the line "moves" simply to induce additional wagering one way or the other. In a bookie's perfect world, there would be equal money bet on both sides of the game. Go Irish!
Dozier | Email | Homepage | 12.27.05 - 11:24 pm | #

The analysis is interesting. the one thing I did notice is that we gave up a lot of garbage time points, final touchdown to Syracuse, points to NAvy and Pitt, etc. that seem to influence the outcome a little. I am sure that tOSU did the same so perhaps it evens out. I think the score predictor is about right, but it comes out reversed. The one intangible you ignore is this whole Hawk-Quinn relationship factor. I am sure BQ would like nothing more than to embarrass big sis' boyfriend on national tv. The family dynamic is important. ;-).
Andy | Email | Homepage | 12.28.05 - 9:20 am | #

That is surpisingly close to my prediction of a 48-3 Notre Dame victory.
Ken | Email | Homepage | 12.28.05 - 11:52 am | #

One quibble- when doing an opponent's average points for/against, you really should take out the game against ND and tOSU (e.g., Pitt, other than ND, was 5-5 with an average score of 24.6-20.1). This is especially true for teams like Tennessee, who gave up one-fifth of their season's points to ND.
gozer the traveller | Email | Homepage | 12.28.05 - 4:36 pm | #

good call gozer
tedwick | Email | Homepage | 12.28.05 - 6:13 pm | #

Your numbers are all fine and good, but they don't take into account the offenses that each defense faced. OSU faced 7 teams ranked in the top 40 in total offense. In other words...

-OSU's opponents averaged 31.59 ppg on the season.
-OSU's defense gave up 14.82 ppg.
-OSU held it's opponents to 46.19% of their norm.
-That translates into a 17.78 score for ND based upon 46.19% of your 37.91 ppg.

-ND's opponents averaged 28.31 ppg on the season.
-ND's defense gave up 23.09 ppg.
-ND held it's opponents to 81.56% of their norm.
-That translates into a 26.55 score for OSU based upon 81.56% of our 32.55 ppg.


OSU 26.55
ND 17.78
Mike | Email | Homepage | 12.30.05 - 8:23 pm | #