Last offseason, I examined the various coaches of the Big Ten and tried to determine how each coach had performed relative to what a fan could reasonably expect (tradifference, the difference between tradition and reality). Simply put, I assumed that fans would be happy if their current coach won more often then an average of their school's past performance, weighted toward more recent seasons. However, I forgot an important fact: college football fans are generally unreasonable. For every reasonable fan, there's at least one of a more optimistic variety. This type of fan remembers their school at its best and sees no reason why they can't win like that all the time.
This time, I based the expectations for a coach on two values: the weighted average of past performance ("realistic expectations") and the record of the school in its best decade in the forty years preceding the current coach ("best ten"). I also only considered conference records to adjust for variance in scheduling in the non-conference, as some schools that formerly played muffins have drifted towards the cupcake range. (If anyone has a better idea for "the opposite of a cupcake", I'd love to hear it.)
As a result, the average expected winning percentage for the current coaches is .609, while the average actual winning percentage in a Big Ten game is remarkably close to .500. Every school cannot have a successful coach at the same time. Some school's fans are always unhappy about their coach.
I've also disregarded Penn State because Joe Paterno has been the coach there for its entire existence in a conference as well as almost 30 years before that.
(I actually had this written before BamaHawkeye's article but hadn't posted it due to the recent hubbub.)
First, the schools changing coaches in order of how well the new guy is expected to perform:
Jim Tressel (2001-2010)
Winning Percentage: .825 (6.6-1.4 average conference record)
Realistic Expectations: .730 (5.8-2.2)
Best 10: .892, 1968-1977 (7.1-0.9)
Combined Expectations: .811 (6.5-1.5)
Tradifference: +.095 vs realistic/ +.014 vs combined
Of the 107 coaches hired in the Big Ten + Nebraska since WWII, only 18 (listed at the bottom) stayed at least 5 years and finished better than the combined expectations. Jim Tressel was one of them. Of those 18, none of them came in with a higher level of anticipated success. Whatever his other failings, his teams won at a historical level even by OSU's standards.
Of course, only a select few of these coaches have been fired or forced to resign: Tressel, Woody Hayes (punched a kid), Mike White (recruiting violations at Ilinois), Bill Mallory (winningest coach in Indiana history fired after two bad years) and Bob Blackman, who holds the dubious distinction of the only coach in the period to be fired with a winning conference record at a school other than OSU and Michigan with no off-the-field problems (that I know about).
Realistic Expectations: .784 (6.3-1.7)
Best 10: .881, 1972-1981 (7.1-0.9)
Combined Expectations: .833 (6.7-1.3)
After a coach like Tressel, the expectations go up even higher. Whoever takes this job faces the highest "realistic" expectations of any coach not succeeding Woody Hayes or Tom Osborne, and knowing OSU fans they will not care about any sanctions. They might realize on some abstract level that they shouldn't demand as much, but if a 4-4 season comes around with losses to Michigan, Wisconsin, Illinois, and Penn State, they will not understand how OSU could possibly lose so many games to inferior programs. With the scandal, I can safely say that the eventual occupant has little chance of surviving.
Winning Percentage: .250 (2-6)
Realistic Expectations: .781 (6.3-1.7)
Best 10: .894, 1971-1980 (7.2-0.8)
Combined Expectations: .838 (6.7-1.3)
Tradifference: -.531/ -.588
I don't want to overstate this point, so I'll going to word this very carefully: Rich Rodriguez was the most disappointing coach in the Big Ten since at least WWII. His -.531 against realistic expectations surpasses every other coach by far. Northwestern's Rick Venturi got a -.339, whom most fans have not heard about but who holds a special place in Wildcat lore. His conference record was 0-26-1 with the one tie occurring in his first game when the Wildcats fought Illinois to a 0-0 draw. Still, he was less of a disappointment than Rich Rodriguez. Cornhuskers consider "Bill Callahan" a swear word in that state after he had a -.347. However, he was again substantially better than Rodriguez.
Inspired by HilaryLee's article, I also checked the coaches of the "Kings" (that list of 10-15 top programs that everyone knows) and actually found a more egregious destroyer of expectations. Jennings Bryan "Ears" Whitworth posted a 3-18-1 SEC record for Alabama 1955-1957, scoring a -.604 against the combined expectations. His Wikipedia article merely says he was fired, presumably because whatever 1950s Alabamans did to him was too gruesome to describe. (The Alabama administration then decided that if nothing else, their next coach would have a more fearsome nickname than "Ears".)
Realistic Expectations: .683 (5.5-2.5)
Best 10: .894, 1971-1980 (7.2-0.8)
Combined Expectations: .788 (6.3-1.7)
RichRod's performance has tempered the expectations for Hoke a bit, but not a lot. They still exceed those of any other current coach except Bo Pelini and Luke Ficknell.
Winning Percentage: .222 (1.8-6.1)
Realistic Expectations: .380 (3-5)
Best 10: .525, 1967-1976 (4.2-3.8)
Combined Expectations: .453 (3.6-4.4)
While not at the unprecedented level of Rich Rodriguez, Brewster still had to go. Of the 34 coaches with a worse tradifference, only two (the predecessors of Joe Tiller and Bo Schembechler) stayed more than 5 years.
Realistic Expectations: .350 (2.8-5.2)
Best 10: .465, 1972-1981 (3.7-4.3)
Combined Expectations: .408 (3.3-4.7)
Coach Kill (isn't that fun to say?) benefits from Brewster's ineptness. His expectations of .408 are the lowest of any coach ever*.
(*Offer not valid at Indiana or Northwestern)
Winning Percentage: .188 (1.5-6.5)
Realistic Expectations: .291 (2.3-5.7)
Best 10: .469, 1985-1994 (3.8-4.2)
Combined Expectations: .380 (3-5)
Indiana is an extremely tough place to win, but 6-26 doesn't suffice anywhere. He also took a big hit by going strictly by conference records instead of overall. In terms of coaches at other schools, he was just a tad better than John L Smith and Ron Zook.
Realistic Expectations: .265 (2.1-5.9)
Best 10: .469, 1985-1994 (3.8-4.2)
Combined Expectations: .366 (2.9-5.1)
Coach Wilson has some good news and some bad news. The good news is that he arrives with the 4th lowest expectations of any coach, after Gary Barnett, the guy before Gary Barnett, and Bill Mallory (Indiana 1984-1996). Two of those three had successful runs.
The bad news is that he's the coach at Indiana.
And now for the few remaining schools that haven't changed coaches recently:
Winning Percentage: .675 (5.4-2.6)
Realistic Expectations: .482 (3.9-4.1)
Best 10: .581, 1992-2001
Combined Expectations: .531 (4.3-3.7)
Any coach that wins at a higher rate than the best decade in their school's recent memory should get a statue built of them. In terms of overall tradifference, he's currently sandwiched between Hayden Fry and Bo Schembechler. Now he just needs to keep it up for the next 15 years.
Winning Percentage: .625 (5-3)
Realistic Expectations: .484 (3.9-4.1)
Best 10: .644, 1987-1996
Combined Expectations: .564 (4.5-3.5)
While not quite at Bielema's pace, Dantonio still has substantially exceeded the bar set by the inestimable John L. Smith and Bobby Williams. If he keeps winning at this rate, he will be the first MSU coach since at least WWII with a positive tradifference. Every other school except Minnesota has had at least one such coach. (I didn't include Duffy Daugherty 1954-1972 because I didn't have enough conference records before him. George Perles was -.002 even with the vacated games included.)
Winning Percentage: .450 (3.6-4.4)
Realistic Expectations: .252 (2.8-5.2)
Best 10: .463, 1995-2004 (3.7-4.3)
Combined Expectations: .408 (3.3-4.7)
Fitz falls back from his top position last year, partially due to the exclusion of nonconference games but mostly due to Northwestern's 3-5 record in 2010 compared to Wisconsin and MSU going 7-1.
Winning Percentage: .552 (4.4-3.6)
Realistic Expectations: .519 (4.2-3.8)
Best 10: .708, 1982-1991 (5.7-2.3)
Combined Expectations: .613 (4.9-3.1)
I think Iowa fans have accepted that Ferentz isn't evil enough to achieve quite the same high point of Hayden Fry. Few coaches are.
Winning Percentage: .708 (5.7-2.3)
Realistic Expectations: .713 (5.7-2.3)
Best 10: .925, 1988-1997 (7.4-0.6)
Combined Expectations: .819 (6.6-1.4)
Nebraska is an interesting case. According to tradifference, Cornhuskers expect more out of Pelini than any other current coach and more than all but a small handful of Nebraska, Michigan and Ohio State coaches historically. However, I don't know how much Nebraska fans' expectations have adjusted based on conference changes and the fact that Tom Osborne's run was simply absurd. Over the past 50 years, Nebraska has had two spectacular coaches (Bob Devaney, Tom Osborne), one ok-but-not-great coach (Frank Solich), and one other coach (Bill Callahan). So far, Pelini has been more Solich than Osborne, though his teams have improved each year.
Winning Percentage: .375 (3-5)
Realistic Expectations: .472 (3.8-4.2)
Best 10: .600, 1997-2006 (4.8-3.2)
Combined Expectations: .536 (4.3-3.7)
Obviously he still has some time to improve, but I don't think Hope's results so far will keep him employed. It's hard to say whether Hope is the problem or if Purdue is simply regressing to pre-Tiller levels, but Purdue fans have come to expect better than 3-5.
Winning Percentage: .333 (2.7-5.3)
Realistic Expectations: .421 (3.4-4.7)
Best 10: .669, 1981-1990 (5.4-2.6)
Combined Expectations: .545 (4.4-3.6)
The only coach to survive longer than Ron Zook with a lower level of tradifference was Ron Turner, his predecessor at Illinois.
Potential future improvements:
1. Add a rivals component (the "John Cooper Memorial Ranking")
Rivalry games count for more than regular games, and fans judge coaches disproportionately by those games. The difficulties here are in deciding what rivalries to include and how to weight them. For example, the Big Ten has one-way rivalries (Illinois-OSU), rivalries that have different levels of emphasis for a school (Michigan-MSU vs Michigan-OSU), defunct but still played rivalries (Minnesota-Michigan), defunct rivalries that may never be played again (Nebraska-Oklahoma, Iowa-Illinois), non-conference rivalries (Purdue-Notre Dame), rivalries against schools that no longer even play D1 football (Northwestern-Chicago, Iowa-Iowa State), budding rivalries (Nebraska-Iowa), ballot-box rivalries (Nebraska-Penn State), rivalries that a lot of fans thought were important but apparently weren't important enough to keep around (Iowa-Wisconsin) rivalries that were once pretty big but then weren't played for awhile but then started up again but might not be a rivalry anymore but could be who knows (Nebraska-Minnesota), the LoL rivalry (Northwestern-Illinois), faux rivalries (Iowa-Purdue) and fake rivalries (MSU-PSU). It's a lot to sort out.
Relative record to a rival (especially one thought to have similar resources) would be another factor to include. Purdue losing to Indiana is one thing, Purdue losing to Indiana and then watching the Hoosiers in a bowl game while the Boilermakers freeze in West Lafayette is even worse.
(Just a quick comment on Cooper: I know he's the poster boy for fired because of performance against a rival, but the guy coached for 13 years at OSU. The only Big Ten coaches since WWII with longer tenure were Woody, Bo, Hayden, Murray (Warmath of Minnesota), Barry, and Jack (Mollenkopf of Purdue) (and Joe). If constantly losing the biggest rivalry game can get a coach fired, that definitely didn't happen quickly in Cooper's case.)
2. Come up with a better way to judge the current situation of long term coaches (the "Glen Mason Memorial Ranking")
When a coach has been at a school for a decade, he's mostly judged against his own previous success and whether or not he continues to improve upon it. Several coaches had been a success early on, but later on they had a couple bad years in a row and were let go (Bill Mallory, Glen Mason). The current system doesn't really capture that problem.
3. Adjust the performance grade for progress (the "Frank Solich Memorial Ranking")
Suppose a coach comes into a school where .500 would be a reasonably good performance but not unprecedented (e.g. Purdue). If that coach has a 12-12 record after his first 3 years, his job security is much better if he went 2-6, 4-4, 6-2 rather than 6-2, 4-4, 2-6. Progress can count more than overall performance.
4. Include the great year halo effect (the "Ron Zook Should-Be-Memorial Ranking")
One great year can cover up a few mediocre ones. A 6-2 season can keep a coach around for longer than his overall winning percentage might suggest, even if he is 10-30 in the rest of his time.
For reference, here are the 18 coaches in the Big Ten + Nebraska hired since WWII who finished their career with a positive tradifference:
1. Ivy Williamson, +.184 over combined expectations
2. Forest Evashevski, +.161
3. Bo Schembechler, +.144
4. Hayden Fry, +.133
5. Bob Devaney, +.118
6. Mike White, +.118
7. Tom Osborne, +.115
8. Woody Hayes, +.093
10. Gary Barnett, +.067
11. Randy Walker, +.042
12. Jim Young, +.040
13. Jack Mollenkopf, +.028
14. Bob Blackman, +.028
15. Bill Mallory, +.018
16. Jim Tressel, +.014
17. Joe Tiller, +.013
18. Barry Alvarez, +.004
There's also the group of coaches that exceeded the realistic expectations of their school, but weren't quite up to the ghosts of great teams past. I call these the Lloyd Carrs.
John Pont at Indiana, +.120 over realistic expectations
George Perles, +.066
David McClain, +.062
Lee Corso, +.050
Glen Mason, +.045
Ara Parseghian, +.029
Nick Saban, +.014
Stu Holcomb, +.006
Lloyd Carr, +.000
Inspired by BamaHawkeye's article, here are the 20 most disappointing coaches hired in the last 50 years ("The Rich Rodriguezes"):
1. Rich Rodriguez, Michigan, -0.588
2. Don Morton, Wisconsin, -0.411
3. Rick Venturi, Northwestern, -0.407
4t. Bill Callahan, Nebraska, -0.400
4t. John Coatta, Wisconsin, -0.400
6. Muddy Waters, Michigan State, -0.390
7. Frank Lauterbur, Iowa, -0.368
8. Jim Wacker, Minnesota, -0.343
9. Jim Valek, Illinois, -0.342
10. Bobby Williams, Michigan State, -0.327
11. Fred Akers, Purdue, -0.322
12. Joe Salem, Minnesota, -0.314
13. Gary Moeller, Illinois, -0.312
14. Jim Colletto, Purdue, -0.290
15. Gerry DiNardo, Indiana, -0.276
16. Ron Turner, Illinois, -0.260
17. John Pont, Northwestern, -0.244
18. Jerry Burns, Iowa, -0.242
19. Tim Brewster, Minnesota, -0.230
20. Leon Burtnett, Purdue, -0.222