Interesting findings. I think this basically shows that the consistently good teams in the Big 10 - UM, OSU, PSU, Wiscy - tend to take care of business against the lesser teams, while the rest of the league kind of fluctuates. I would be interested to see how the big 3-4 in the conference stack up against each other - my guess is that most of the "upsets" for them come against each other, where one is favored by a couple of points but are generally equal from a talent perspective.
The Over and Underrated
Given all the talk before the Michigan – Michigan State game about “respect” as well as under and overrated teams, I found myself wondering which team in the Big Ten is typically the most overrated. I feel like MSU never meets their preseason expectations, but the same can usually be said about Michigan as well. Ohio State has been very solid within the conference, but obviously slips up once it goes up against the elite outside the Big 10. But which team is the most consistently overrated during the entire season?
I analyzed the point spreads for all Big 10 games for the past 10 years, 1999-2008. Team performance against the spread is well documented, but we don’t really care about ATS; hanging 50 points on Wofford when the spread is 40 does not an underrated team make. We want to break the lines down to victory or defeat, and see how the team performs in comparison.
W # of upsets against opposing team
L # of times being upset
TOT Total deviation from expected over total games
Stdev Standard deviation
PW Predicted win percentage
ATS Performance against the spread
46% Penn State
51% Ohio State
53% Michigan State
Move over Mark Dantonio, there is a new Rodney Dangerfield in town. Yes, humble Northwestern earns the title of most misunderestimated in the Big 10. Meanwhile, Purdue is the most overrated team, although they have several teams not far behind, notably Michigan. I expected the Big 10 bottom-feeders to be the most underrated; when everyone expects you to lose every game, there is nowhere to go but up. Likewise, the big boys would be near the top. But Purdue has no excuse ... they have been given the modest task of winning 2/3 of their games, and they consistently blow it.
- Ohio State has the highest predicted win percentage @ 83%, as well as the lowest standard deviation. People expect them to win, and they oblige.
- Michigan State appears to perform as-expected @ -2.5%, but they have the highest standard deviation. Sparty wins a lot of games they have no right to win, and loses a lot of games for no reason, and basically acts very Sparty-like
- The third most underrated teams is Iowa. Given Michigan's overrated-ness, this does not bode well for this Saturday. Or it has no relevance, since Iowa is already favored ... I haven't decided.
- Purdue has only upset 6 teams in the past decade. Northwestern performed the same feat between 2005 and 2006.
- I expected the ratio of underrated teams to overrated teams to be closer. The Big Ten, not surprisingly, is not performing.
Thoughts? Is this a useful analysis of overrated-ness? Should this be expanded to additional seasons and teams? Spoiler-alert: I have already looked at Notre Dame, and they are not, repeat NOT the most overrated team in the universe.
all that talent and couldn't do much with it since his BCS bid. I do like the analysis, thanks. Perhaps doing it over a coach's tenure to see when they started falling of the deep end or when they began turning a team around? You can also test some hypotheses regarding recruitment drop off/pick up and overrated-ness, etc. Lots of potential i think.
Yeah it's very well done, great job. I'd add a graph or something to track the ups and downs of it per year (or per team weekly for the whole span) but that'd require a lot of work on your part, and I'm just getting greedy.
Does anyone else feel spoiled by the quality of this e-blog, the blog itself is tremendously written and the user generated content is just as good.
Spoiler-alert: I have already looked at Notre Dame, and they are not, repeat NOT the most overrated team in the universe.
I could have told you that :)
What about to Lou Holtzth?
Notre Dame is always underrated. In his universe, they are always No. 1 and nobody can beat them.
...this is interesting. I wonder if a useful alternative to the "TOT" stat might be something like [W/(expected losses)]-[L/(expected wins)] rather than (W-L)/total games. That way teams with far more opportunities for upset wins than upset losses would not be "rewarded" quite as much; you would just be evaluating what percentage of of a team's expected wins they win versus what percentage of their expected losses they lose. Its just a thought, but that may correspond better to the conventional senses of "underrated" and "overrated." Under my suggestion, the numbers might skew a little (or a lot) for teams with very few expected wins or expected losses (imagine a team with only 1 expected loss over 10 years--a game in which they were 3 point underdogs or something--who pulls that game out; they look like a hugely underrated team, and that doesn't seem right either). Again, just a thought.
However, in the last scenario, their TOT score would be 1/120 (or however many games they play), so they will not come out as hugely underrated. I will see how your suggestion pans out.
I still can't get over how highly Purdue is thought of. They were predicted to win almost as many games as Wisconsin and Penn State! I know I personally do not group Purdue with these other teams ... this sort of explains why.
...and in fact, I expect my suggestion would be pretty favorable to teams like Michigan and OSU (with relatively few games as underdogs, and I am guessing very few in which they were more than a few-point underdogs). But, at the same time, it strikes me as off (or at least out of step with *my* conception of "underrated") that Indiana might be the 2nd most "underrated" team in the Big 10 over the last 10 years. That they manage to pull out some 4 and 5 win seasons when expected only to win 3, doesn't really make them underappreciated, IMO (they've only played in 1 bowl game over that time period).
I think the reason Northwestern is so far ahead is that no matter how many times they "upset" teams that are favored to win, people fail to give them the requisite respect, falling back on the "oh, it's just silly little Northwestern" mantra. I see Northwestern constantly singled out on message boards as a team that "doesn't belong" and "can't cut it" in the Big Ten. The reality is that, over the past 15 years or so, the Northwestern team people continue to expect to be awful has instead been average, with a few above average teams mixed in, including 3 Big Ten title teams.
"overrated" is not a matter of statistics & variances, rather perceptions by the public as they are informed/reported to by the sports media. The media tells you this team/coach/player is great, and if enough people buy it...it sticks. Then sterotypes stick as a result of that. That all feeds into teams, in this case, being "precieved" as overrated/underrated. And generally "overrated" is a function of the teams said perception by the public (who gets reported to by the media), and the media themselves, which orchestrates the prime measure of "rated" ie. The POLLS!
Excellent analysis. I'd be interested to see this expanded to analyze the performance of the so-called "underrated" conferences - most notably the WAC and the Big East.
You've confirmed what many people have been saying about Michigan (anecdotally) for years.
We fans are frustrated that the Wolverines too often lose games they're supposed to win, or win by nervously narrow margins. Non-fans sometimes take this as evidence that Michigan is a soft team, prone to choke away their advantage. You've shown that there is some truth to this. We aren't just imagining it.
I realize that 10 is a nice round number, but you are only looking at the Carr era, and cutting off the best performing season of that era.
But overall I agree with EvilEmpire that Michigan Fans have traditionally (yes this is one of our traditions) expected Michigan to win the National Championship every year.
Growing up watching the Bo era, many years seemed to be "the one" only to find out that Bo didn't have the National Championship as his goal, feeling that a popularity contest was something out of his control. His stated goal was to win the Big Ten Championship and GO to the Rose Bowl.
Note, my memory is that he always stated it as "Go to the Rose Bowl", since he viewed the trip as a vacation reward for winning the Big Ten Championship, and actually winning the Rose Bowl was not as important (I'm not saying any coach likes losing, they do NOT, but they also know you can't win them all, and place their importance differently on different games. For Bo the Non Conference schedule was a secondary priority to the Conference and the Championship.
Now my very flawed memory has statistically created the belief that in the Bo Era, whenever Michigan was rated highly, they would lose to Notre Dame, and usually another non-conference game (remember in the Bo era Notre Dame was the first game, and he usually had another top level team to play, like Miami or Florida State right before the Big Ten conference games) and drop in the polls, although never out of the polls, and pretty consistently Michigan would be the highest rated 2 loss team on the list. Then the buildup to Ohio State would begin.
We are clearly in a new era of expectations. Not that Michigan's have changed, but everyone else's in the Big Ten is changing.
Soon, there will be enough parity to warrant adding Notre Dame to the conference and establishing a Big Ten(12) championship game.
A big step forward to this eventuality is the press helping Dantonio establish the UM-MSU game as a "rivalry" instead of just a grudge match for in state bragging rights.
Which incidentally this year belong to Central Michigan.
Thanks Spartans, no how about winning some other Big Ten games this year so we have a shot at the Champtionship.
I'd respect that.