BigTen Strength of Schedule
Looking through Sagarin's rankings to get any insights into the coming match-ups, I noticed something interesting about SoS of BigTen teams. Several BigTen teams have really abysmal SoS and that may be making them look better than they are. Cases in point (figures in parantheses are the current SoS ranking followed by current national rank) :
- OSU (118, 6): Easily the worst SoS in top 20. It has been easy for them to win against easy competition - will they be able to elevate their game agaist sterner tests?
- Iowa (128, 27): Second worst SoS in top 30 after Nebraska. I am picking Iowa for this year's second most disappointing team compared to pre-season expectations. The first choice is of course Penn State.
- Wisconsin (133, 31): Second worst SoS for FBS teams in top 40 after Nebraska. Although has had trouble putting away teams (except the last one) convincingly - it is also their style. The low variance run oriented approach fives you Ws without too much drama. So I expect them to rise up to stronger schedule coming up.
- MSU (157, 41): Worst SoS among all top 50 teams. I doubt they will hold up to better scrutiny. I bet they will fold against Wisconsin this Sat because like last year, they will be totally working for next week's Michigan game. Sparty does not mind a loss if it gives them the Michigan win.
- Indiana (200, 85): Worst SoS among all top 100 teams. I think their 3-0 is a product of the ridiculous schedule. Their defense will soon get exposed and their high variance pass oriented offense will stall enough for not being able to make up for the defense's shortcomings.
The four teams above are notable for having the worst SoS in the country for their national rank. I had some idea that Big Ten's SoS is weak but I had not thought it to be this bad.
Michigan's (77, 26) has had a decent scheule for its rank. Penn State has a better SoS at 66 because of the Alabama game but since they lost that game convincingly, I don't think the current rank of 22 is justified for them. Illinois (113, 61) seems to have an okay SoS for its rank.
Well - what I am concluding from the above is that we have a decent shot against even apparently well ranked teams like OSU, Iowa and Wisconsin - and surely a very very good shot against teams that have hardly played anybody like Indiana and MSU. I am a little more calm in my head for next two weeks after looking through the data. Of course, the next two weeks of Big Ten matchups will give us much more information to have a better idea about Big Ten teams.
Sagarin's rankings are available here: http://www.usatoday.com/sports/sagarin/fbt10.htm?loc=interstitialskip
PS> My very first diary. Yay!
September 28th, 2010 at 11:17 AM ^
These teams have crappy out of conference schedules ... what does this have to do with our chances to win a game?
September 28th, 2010 at 11:24 AM ^
Just that it may mean that they are looking better than they actually are. So we may have a better chance of winning than what their current national rank may imply.
September 28th, 2010 at 12:27 PM ^
Except that only a gigantic fucktard forms an opinion about the likelihood of winning a game based on the rankings, which are entirely subjective.
September 28th, 2010 at 12:44 PM ^
Last I checked sagarin was not subjective but I know how terrible buckeye fans reading comprehension is...
<br>
<br>In sum you're an idiot. While I don't agree with the OP that osu is over-rated I think that PSU is going to be god awful.
September 28th, 2010 at 4:35 PM ^
Well technically, the rankings are subjective, because they reflect however he chose to weight the categories in his formula. But they're less subjective than say, the AP and Coaches' polls.
September 28th, 2010 at 10:51 PM ^
While the formula is subjective. I believe the actual rankings are objective because the formulas that give the rankings were created without a bias for any particular teams. A subjective and formula based ranking would be something like, "I like Michigan so I'm going to wait the total yards rushing by the quarterback as the largest weight for ranking a team." An objective and formula based ranking would be something like, "Winning is a decent metric to determine how good a team is so I am going to base my formula on win-loss record."
September 28th, 2010 at 2:29 PM ^
Yes, and strength of schedule is pretty much meaningless at this point, as most top teams (teams we know are pretty good regardless) have played a lot of overmatched opponents, so neither ranking nor sos is useful for predicting outcomes - why bother to use one to debunk the other?
September 28th, 2010 at 12:29 PM ^
He is saying PSU will turn out to be the most overrated team from the Big Ten in respect to pre-season rankings. --Think Michigan like every other year under Carr or Mo.
EDIT - Response was to Braylon 5... not sure how his/her post moved...
September 28th, 2010 at 11:19 AM ^
Although the Sagarin ratings are still biased and Bayesian. One thing is clear though - the Big Ten has about the worst SoS of any major conference so far on average, even assuming that the ratings aren't super informed without more data (and removal of bias). Without doing the math, the Big Ten looks pretty far behind the Pac-10 and SEC.
Having said that, the numbers get in the way a bit. If you are a top 10 team, it doesn't much matter whether you play the 50th best team or the 100th best team, you are very likely to win. A tough schedule to me involves games you might lose. By that measure OSU's schedule is tougher than the numbers say because they played a top 20 team that (on paper) should have enough talent to compete or beat them. Take Iowa - they played a top 20 team on the road and lost. Many top 20 teams can't make a claim that they played a team that could beat them, despite what the math says.
September 28th, 2010 at 4:01 PM ^
You say "Bayesian" like that's a BAD thing....
September 28th, 2010 at 11:33 AM ^
is to take a look at the individual rankings of each of the teams Michigan and our conference peers have played.
OSU: #11 Miami, #97 Marshall, #107 Ohio, #160 EMU
Iowa: #12 Arizona, #65 Iowa State, #152 Ball State, #204 Eastern Illinois
Wisco: #45 ASU, #101 UNLV, #106 SJSU, #187 Austin Peay
Michigan: #64 Notre Dame, #69 UConn, #84 UMass, #94 BGSU
MSU: #64 Notre Dame, #125 Florida Atlantic, #136 WMU, #171 Northern Colorado
PSU: #1 Alabama, #50 Temple, #119 YSU, #123 Kent State
Need to get back to work. Basically, it looks like OSU, PSU, and Iowa have played the single toughest OOC games, but the rest of their schedules are mediocre, at best.
To that end, Michigan's schedule doesn't have any truly horrid teams on it, but is pretty meh overall due to the underachieving ND and UConn squads.
September 28th, 2010 at 1:20 PM ^
Notre Dame will improve over the next few weeks. Their schedule was really front loaded and they lost heartbreakers to both Michigan schools. They'll be a fringe top 25 team by the end of the year.
September 28th, 2010 at 3:22 PM ^
If ND is a fringe Top 25 team at year's end, it would only prove that the polls are totally biased in favor of "name" teams. ND will likely finish no better than 7-5 at this point (I think 6-6 is the most likely result) as they still have the following games:
@BC - Yes, BC isn't that good, but ND has had a lot of trouble beating BC lately, particularly in Chestnut Hill.
Pitt - I think ND will win this game as Pitt looks to be seriously struggling this season, but ND doesn't look all that great right now either. Toss-up game with slight edge to ND as it is in South Bend.
Navy (in NJ) - Navy doesn't look quite as good this year as they have in the last 6-7 years, but this still looks like a toss-up game to me.
Utah - Still one of the best non-AQ teams out there. I expect them to give ND all they can handle.
@USC - Armageddon for ND.
That is five games ranging from moderately tough to nigh unwinnable for ND on the remaining schedule. I would be very surprised to see them win more than 3 of those games which would get them to 7-5 at best assuming wins against the weaker teams on their schedule.
September 28th, 2010 at 4:06 PM ^
I look at the first three as almost certain wins and I think USC will get into trouble as the season progresses. Their marquee win is a 17-14 win over Virginia. If ND beats either USC or Utah, they're at 8-4 winning seven of their last eight games. That's a fringe top-25 team for most pollsters, provided it's a BCS school.
We may just have to agree to disagree here.
September 28th, 2010 at 11:55 AM ^
How is Penn State the biggest disappointment in the conference compared to expectations? They are 3-1 with 3 basically meaningless wins, and 1 road loss to the #1 team in the country while starting a true freshman quarterback. I don't think they are particularly good, but I am not sure how they are disappointing anyone so far.
September 28th, 2010 at 12:44 PM ^
I expect them to implode going forward. Fr QB, weak O-line and recent injuries all combine to make them weaker and at least put our game with them as a toss-up - even though it is a night game and a possible white-out.
September 28th, 2010 at 12:57 PM ^
I don't think they'll implode, but they will struggle. Expect a lot of 20-17 type games. I wish Michigan had Penn State this week, because the Bolden should improve faster than any QB in the Big 10, and Penn State's offense will improve correspondingly. That said, Michigan's defense is also going to improve more rapidly than any other defense in the conference due to its relative youth in the two-deep.
All that said, Michigan v Penn State is better than a tossup on a neutral field, and is probably a tossup in Happy Valley.
September 28th, 2010 at 1:12 PM ^
Well I think the key will be the Penn State offensive line play. If we can manage to control Royster, and the freshman QB doesn't play a flawless game, that should be more than enough to keep our offense with a fighting chance to be in there. That's why I love the Notre Dame win so much, it was a struggle for portions of that game but we still had a fighter's chance offensively at the end and we got the win. No matter how terrified I am of our defense, it's going to take a really strong performance by our opponents to blow us out this year.
September 28th, 2010 at 2:00 PM ^
Right now Indiana is an unknown. They have only played 3 games (already had their bye) and those teams have been horrible. If you just look at the numbers Indiana appears to be relatively even with M. But, considering their SoS, they don't look close to M.
We'll know more ina few days.
September 28th, 2010 at 2:36 PM ^
Both suck. IMO this comes down to turnovers and FGs (who has the least of each).
September 28th, 2010 at 4:41 PM ^
The gap in talent/ability between our offense and their defense is much greater than the gap between their offense and our defense. As long as our center doesn't snap the ball over Denard's head five times, we'll win comfortably.
September 28th, 2010 at 2:29 PM ^
I stayed up (I'm deployed in Afghanistan) and watched the Alabama at Arkansas game. Alabama shouldn't have won that game. To me that suggests that Alabama shouldn't be the #1 team in the country, and that Penn State should have done better if they were ranked correctly. Penn State is the biggest disappointment of the teams in the Big Ten that were supposed to be good this year.
September 28th, 2010 at 3:59 PM ^
and +10,000....to you for serving, sir.
Hope you can stay up and watch Denard run clear through the state of Indiana without stopping.
September 29th, 2010 at 8:21 AM ^
And thanks for all you do for all of us back home so we can worry about Michigan Football and not our way of life.
September 28th, 2010 at 4:50 PM ^
I didn't realize MSU's schedule was this horrible. And Wisconsin, Iowa and OSU - wow. It shows that we really can't draw many conclusions from most of the games that have taken place thus far.
September 28th, 2010 at 7:05 PM ^
Did anyone expect them to be 4-0 at this point? I know I figured they'd lose to Alabama. They might not have been that impressive in their 3 wins but 3-1 is what I guessed they would be heading into big ten play.
It seems like your penalizing Iowa for losing on the road to a pretty good Arizona team. They've won their other 3 games by a combined score of 117-14. For all the grief they took on this board over the summer for barely beating the non-conference cupcakes on their schedule last year, they've done a 180 this year and taken care of business against those teams.
September 29th, 2010 at 1:18 AM ^
SOS is completely subjective. college football rankings don't account for how in the last 10 years, the difference in the level of talent between traditional powerhouses and no-name teams has become far narrower than it has ever been. nobody doubts osu is a top 10 team.
The actual rankings as i look at them are more like groupings, like a grading curve. What this means is that anyone within a class has the ability to beat another member of its class at least half of the times they would hypothetically play each other, or the scores would be close every time they played. a hypothetical 10 games.
Group A (teams in the top 5) These teams beat B and C teams 9/10 times where the wins are decisive but they split games amongst each other 5/5)
Group B(teams below the top 5 but in the top 40) These teams beat A teams 1/5 or m times and lose no less than 4 out of 10 games to any other team in group B)
Group C(Teams starting at the 15th actual best team all the way down to the 100th actual best team) these teams lose 10/10 games against top 5 teams and lose more than 5 to Group B teams, while splitting amongst each other 3 out of 10 to 7 out of 10)
So basically it goes like this. if your team is in group c you are among the third echelon of teams but you might be ranked in the polls as high as 15 or you could be temple or toledo or umass. you are in group c becasue hypothetically if you play a group B team 10 times, you lose more than 5 times, or on average, lose by more than a touchdown. example: FSU 2009 loses more than 5/10 times to cincinatti 2009, iowa 2009 beats slippery rock more than 5/10 times.
IF your team is in group B you may be ranked very high or you might not be in the top 25 at all. Group A teams beat you 4/5 games and consistently win by more than one score. I. Either way, Group A teams beat B teams like you 9/10 times whereas you beat those B teams only 5/10 times. example: Florida 2010 beats Cincinatti. 9/10 times | Florida's single 2009 loss was roughly 1 in 2 odds to alabama , another group A team from 2009) Cincinnatti was 12-0 prior to the loss on wins over group B and group C teams (very close wins over Pitt, Uconn)
So the basic premise behind this system is that there are no such thing as rankings as the media portrays them. ohio state is probably a top five team but the fact they beat miami (who is either a B but could be a C team) means no more than if they would have beaten Notre Dame (who could be a B team or a C team)... However, if OSU goes undefeated in the big ten, that would be huge for them only if one of the teams they beat only lost to osu (which would be evidence that that team was in fact also a group A team)
September 29th, 2010 at 1:40 AM ^
I think we are a top 5 team even with our defense. sorry for the long post below!
Comments