October 4th, 2015 at 10:24 AM ^
we are look pretty great on defense
October 4th, 2015 at 10:51 AM ^
yes defense have it we
October 4th, 2015 at 10:53 AM ^
October 4th, 2015 at 11:15 AM ^
Sent from MGoBlog HD for iPhone & iPad
October 4th, 2015 at 11:24 AM ^
Going on what is?
October 4th, 2015 at 12:34 PM ^
Sent from MGoBlog HD for iPhone & iPad
October 4th, 2015 at 1:41 PM ^
October 4th, 2015 at 11:33 AM ^
look as good as me you will not.
October 4th, 2015 at 11:45 AM ^
October 4th, 2015 at 11:41 AM ^
This is all well and good but the shutout of Maryland makes me feel as good as if I pummelled the village idiot in the town square. That offense is absolutely putrid.
October 4th, 2015 at 1:22 PM ^
October 4th, 2015 at 5:27 PM ^
A lot of sparty fans on here.
October 4th, 2015 at 12:08 PM ^
would start for OSU
s/
October 4th, 2015 at 12:54 PM ^
Sent from MGoBlog HD for iPhone & iPad
October 4th, 2015 at 1:02 PM ^
Sent from MGoBlog HD for iPhone & iPad
October 4th, 2015 at 6:42 PM ^
Thank you for posting this.
And I hope you repost this every time we win convicingly. Especially the final regular season game.
Spielman sucks. /not s
October 4th, 2015 at 6:40 PM ^
Hopefully the loss of Ojemudia won't hurt us too much.
October 4th, 2015 at 10:24 AM ^
October 4th, 2015 at 10:25 AM ^
October 4th, 2015 at 10:26 AM ^
Yep. We always hate the coaches poll, because it's a pointless, biased poll that a lot of coaches don't even take seriously.
October 4th, 2015 at 10:30 AM ^
October 4th, 2015 at 11:15 AM ^
Sent from MGoBlog HD for iPhone & iPad
October 4th, 2015 at 12:18 PM ^
October 4th, 2015 at 11:00 AM ^
Unless UM was the top rated Big 10 team in the coaches poll, at which point it would be the best poll ever.
October 4th, 2015 at 12:36 PM ^
bias-free speaks so well for the integrity of the game!
October 4th, 2015 at 10:38 AM ^
Considering Playoff team selection is human-based and susceptible to influences other than numbers, I think the coaches poll still matters.
October 4th, 2015 at 10:50 AM ^
October 4th, 2015 at 5:16 PM ^
Or that ranked us behind Nebraska in 1997.
October 4th, 2015 at 10:27 AM ^
October 4th, 2015 at 11:05 AM ^
Because a whole generation of people have grown up using this one weird trick....
October 4th, 2015 at 11:33 AM ^
You won't believe what happened next.
October 4th, 2015 at 11:51 AM ^
Sent from MGoBlog HD for iPhone & iPad
October 4th, 2015 at 12:42 PM ^
More at 11
October 4th, 2015 at 10:51 AM ^
October 4th, 2015 at 11:25 AM ^
Sent from MGoBlog HD for iPhone & iPad
October 4th, 2015 at 11:26 AM ^
I like those things, too.
Anyway, our blowouts aren't like Baylor's or anything and we haven't truly run up the score on anyone.
October 4th, 2015 at 11:32 AM ^
Is that what the data says though? Great teams blow bad teams up because they have no other option - play your starters and they're going to kill the other guys, get up big and cruise. I realize M struggled on O in the first half but that game wasn't competitive at any point - Maryland would never score on that defense in 100 tries. They completely shut down what had been a decent rushing game from Maryland. Utter domination. Is that what highly ranked teams have historically done? If yes, it has value in a statistical model, and I assume he has data to support how much he weights it. Same thing with close losses vs. good teams.
The only thing I would question is the Michigan result: were the blowouts flukes, and was the close game not really as close as the score made it appear? I'd say no to both: BYU and Maryland were at least as bad as the score indicated, and if a couple of those deep balls had connected against Utah the result could have been completely different. M didn't deserve to win, but it was a closely contested game.
October 4th, 2015 at 12:52 PM ^
October 4th, 2015 at 3:49 PM ^
October 4th, 2015 at 1:06 PM ^
It is unclear from your post how you formed that opinion. By contrast, Sagarin's model is based on many years of data. It is unbiased and actually predicts the outcomes of future games quite well---much better than the rankings based entirely on opinion and subject to other biases (eg coaches' poll).
Sagarin's ELO-CHESS rankings ignore lopsided scores, a process yoiu seem to favor. Although they are more "politically correct,", they are not nearly as good at predicting future games as are the ratings in the OP's link (eg PREDICTOR rankings).
October 4th, 2015 at 2:44 PM ^
I have been studying Sagarin's ratings for 30 years now. His ratings have ALWAYS been overly biased towards teams that put up lopsided scores over mediocre opponents. This isn't just a football problem--it happens in basketball too.
His ratings have always been biased towards teams that win with offense over teams that win with defense. Now, we happen to be putting up lopsided scores recently while winning with defense. But his model never really values defensive teams. And that is why his ratings always said the Bad-Boy Pistons teams weren't all that good. They never won with the offensive style-points his system requires. In contrast, he loved the '89 Michigan basketball team. They scored points in buckets.
My biggest beef with his model is that for a team with a lockdown defense, I really don't see much difference between winning by 28 and winning by, say, 14, yet he seems to think there is a large differece in these results . For a great defensive team, a 14-point win is basically a blowout. But his model doesn't really account for that, probably because if you've got two great offensive teams, a 14-point win is still a close game.
No I haven't published some sort of paper on this, so yeah, it is all just IMO. But to say that the Sagarin ratings are "unbiased' is a load of horsecrap. They are completely biased--towards whatever Jeff Sagarin believes is the proper way to evaluate the strength of teams. They are singularly biased towards HIS formula. They may not be emotionally biased, but that doesn't mean they aren't biased.
October 4th, 2015 at 7:43 PM ^
His ratings are unbiased in the sense that they are derived from actual data and do not under or overestimate the actual margins of victory in future games, I believe.
Granted, you could start a priori with a different type of model. Intuitively, I agree with you that it seems logical to consider not merely the difference in scores but also their sum (the total amount of pints scored). The question, however is not whether the model seems logical but whether it actually performs better than Sagarin's.
One simple-minded way to construct such a model ould be to divide the difference in the two teams' scores by the sum. So a 28-21 win would get 7/49=1/7 of a point, while a14-7 win would get 7/21=1/3 of a point---moie than twice as much. That model would not however distinguish a 14-0 win from a 28-0 win, as both would get a full point of credit. So, you would need to fiddle around with the formula to get more reasonable results across a wider range of examples. Also, you would need to work out some fancier statistics to test the signifance of your results, since your statistic would not be normally distributed.,
If you have studied this problem for 30 years, and you are confident that Sagarin's model is worng, you certainly must have thought about how to replace it with a better one.
October 4th, 2015 at 11:54 AM ^
We're about as good as anyone in the B1G over 21 positions. But Rudock and that goofy horizontal/short passing game he forces us to use is going to get pick sixed and stymie our offense against elite teams. We can't afford missing wide open receivers for six in those games.
October 4th, 2015 at 5:22 PM ^
This. Sooner or later the offense is going to have to actually win a game for us. We can't expect the D to always keep us in a game by holding the opponent scoreless while we fiddle-fuck around on offense for 3 quarters. It's a risky way to live.
October 4th, 2015 at 1:29 PM ^
We did not have a 1000 yd rusher or even a receiver with 40 catches.
They're are lots of teams out their with bright shiny loud offenses. I thought yesterday's smothering was a thing of beauty.
Give me these ugly wins all year long.
October 4th, 2015 at 5:21 PM ^
October 4th, 2015 at 5:28 PM ^
Our '97 offense was not dynamic. But it was functional and sound in all phases of the game.
They could run the ball, they could throw the short stuff to the TEs, and they could hit the long pass when the other stuff lulled you to sleep (all 3 of our Rose Bowl TDs were off long passes . . . because the run and short passes were working reasonably well).
I'd absolutely settle with pairing '97s offense with this year's defense.
But this year's offense is not '97s offense. They have a long way to go to get there.
October 4th, 2015 at 1:59 PM ^
Although we all realize there is much fine tuning to finish, we are rapidly turning into the MACHINE, I thought we might when JFNH was hired. No question we had the talent. No question said talent had to mature into MEN. That said a catylst was needed.JFNH, filled that need. After the slow start yesterday, followed by the second half thrashing; there is zero doubt in my mind this TEAM can and will make a strong run for the B1G Title. I've been watching our men since '68. We are back. Now. Haven't seen this for yrs. Now? Sit back and enjoy.
MUCH WORK TO BE DONE. BUT WE ARE BACK. GO BLUE!
