|
Contributed by Peter Tiernan, ESPN.com Bracketologist
|
Struggles of top teams open door for upstart schools
It's true that the modern era's top 10 schools in terms of tournament winning percentage haven't changed since last year, save for a little shuffling at the bottom (see below). And they've combined to win 405 games -- an impressive 29 percent of the 1,386 games played since 1985, including 15 of the 22 championships. But beneath the surface, the laws of tourney supremacy are beginning to crumble. You can see signs of a new order in the fact that not a single top 10 team improved its winning percentage from the 2006 NCAA Tournament. Maryland climbed three positions on the list -- simply by not playing in the tourney.
To get a better perspective of how big a scrum the tourney is becoming, look no further than the results of the last decade. Since 1997, the NCAA has crowned nine different champions -- with UConn being the only two-time winner. Perhaps even more telling, 26 different schools have filled the 40 Final Four slots over the last decade. Only nine teams have gotten to the semifinals more than once since 1997, led by Michigan State and North Carolina with four. As a point of comparison, only 18 different teams have reached the NBA semifinals since 1997 and there have been just five unique champions.
After the 2007 tourney, the effects of college hoops parity should become even more evident. The bottom four teams in the list of winningest schools -- Maryland, Michigan State, Syracuse and Arkansas -- are in serious danger of falling out of the top 10. Which programs are vying to take their place? The most likely candidates are Florida (currently 23-12), UCLA (30-16), Georgetown (26-14), Louisville (26-14), Arizona (39-21) and UNLV (21-8).
All the Gators have to do is reach the Elite Eight, and they'll jump to seventh place on the list. With a repeat championship and another early Kansas exit, Florida could jump as high as fifth on the modern era winning percentage list. UCLA, Georgetown and Louisville need to reach at least the Elite Eight to crack the top 10 ?and can go as high as seventh with championship runs. Arizona has the toughest task in cracking the top 10. If the Wildcats get to the Final Four, they could sneak into ninth place.
The easiest road to the top 10? That belongs to UNLV. All the Runnin' Rebels have to do is make the tourney. They already have a good enough record and can post no worse than a .700 winning percentage with a first-round exit. They just need one more bid to meet the "10 appearances" requirement. Other teams with an outside shot of sneaking into the top 10 winning percentage list include Villanova, Seton Hall, Ohio State, Georgia Tech, Oklahoma State and Indiana.
Before we get too carried away tearing down the top 10 tourney winners, let's give credit where credit is due -- at the very top of the list. Based on records alone, Duke is clearly the best school in the 64-team era. It has the best record by a solid margin, has reached more Final Fours (10) and won more championships (three) than any other program. For Duke to lose its stranglehold on the top position this year, it would have to bow out in Rounds 1 or 2, while UConn won the tournament. To put Duke's dominance in perspective, the Blue Devils could suffer 17 straight first-round exits -- and still remain in the top 10 list of tourney winners.
The best performers against seed expectations
It's fair to say that Duke's average seeding of 2.14 has preordained a reasonably strong tournament record. After all, the typical No. 1 seed has a .793 tourney winning percentage. How can you tell that Duke's tourney performance, when adjusted for seed bias, isn't actually worse than some schools with solid records built from lesser seed positions?
We created the PASE statistic to answer such a question. "Performance Against Seed Expectations," or PASE, compares a team's actual number of wins to the amount it should've won based on its seed position. Over the last 22 years, each seed has recorded an average number of wins per tournament. Top seeds win 3.32 games each tourney, two seeds win 2.41, three seeds win 1.78 and so on. When a top-seeded team wins only two games, like Duke did last year, it underperforms by 1.32 games -- despite a solid 2-1 record. If you add up a team's year-by-year performance, you can come up with the average games per tourney that it deviated from the expected victory total. Duke should've won 54.2 games based on seeding. The Blue Devils actually won 68 -- about 14 games better than expectations, or two-thirds of a game more per tourney.
How does Duke's PASE value stack up against other schools? Well, if you're looking for a chink in the Blue Devils' armor, you won't find it here. These are the top 10 schools with at least 10 tourney appearances that perform better than their seeding would dictate:
Adjusting tourney performance for seeding doesn't change which school comes out on top. Duke's +.657 PASE can be looked at in one of two ways: either every Blue Devils squad wins two-thirds of a game more than seeding says it should, or two out of three Duke teams win one more game than they should. However you slice it, the Blue Devils have had the greatest expectations to advance the farthest in the tourney ?and have exceeded those expectations the most. UConn's record is also impressive, coming second both in terms of overall record and surpassing expectations. But the Huskies are clearly a notch below Duke -- at least over the full 22 years of the 64-team era. As we'll soon see, the last decade is a different story.
The real story of the 2006 PASE results, however, isn't about how dominant the leading teams are, but how much they've lost off their lead. Only two of the modern era's top 10 PASE performers actually overachieved last year. Florida was the only substantial gainer, adding .326 to a PASE that was practically flat before 2006. The Gators' championship run vaulted them from 24th place on the modern-era PASE list all the way up to seventh.
Kentucky also added to its PASE value with a first-round win as a No. 8 seed, increasing a modest .003 -- yet leaping from seventh to fourth. (Maryland's slight increase was the result of small differences in the "wins by seed" calculations.)
No offense to the Wildcats, but it didn't take much to climb the top 10 list after last year's tourney. Michigan made a similar three-position jump, from eighth to fifth ?simply by not making the tourney. (I'm sure this comes as small consolation to Tommy Amaker.) Half of the top 10 suffered significant PASE declines in 2006.
At -.091, Duke's was the biggest drop ?but the least painful. That distinction belongs to Syracuse. The Orange's first-round failure as a No. 5 seed dropped Syracuse from a +.343 PASE to +.262 -- and out of the top 10 to 12th place.
Arkansas and Villanova also lost ground, each falling three slots with narrow underachievements at its seed position. The Razorbacks lost their opening-round game as a No. 8 seed, and the Wildcats only made the Elite Eight, ?their three wins just a hair shy of the average 3.32 wins for top seeds. UConn and North Carolina also underachieved last year, but the Tar Heels' second-round exit as a No. 3 seed proved more costly, dropping them from fourth to sixth on the modern era PASE list.
One more team merits mention for its PASE decline?. If you're a Kansas fan, you may want to move on to the next section. Two years ago, the Jayhawks were the fifth-best PASE performers of the 64-team era. Then they lost to Bucknell as a No. 3 seed in 2005 and fell to 12th place. They followed that with an upset to Bradley last year -- and plummeted all the way to 21st place with a PASE of +.217 -- still in the black, but, as I'm sure many Jayhawks fans will agree, in serious need of some tourney overachievement.
The worst performers against seed expectations
Of the 46 schools with at least 10 tourney trips in the modern era, 20 have fallen short of the number of wins their seed position would dictate -- and the worst of that list is Purdue. Seven disappointing tourneys from the top three seed positions have doomed the Boilermakers to "half-game-per-tourney" underachievement. Purdue fans can take heart, though, that their team has reversed its fortunes over the last decade. As we'll see shortly, the Boilermakers are actually one of the top overachievers of the last decade.
The same can't be said of their Big Ten counterpart, Ohio State. The Buckeyes were the only new entry in the bottom 10 of modern era PASE underachievers. Their second-round loss as a No. 2 seed to Georgetown knocked .121 off their PASE, dropping them to -.285. At least they did a favor for another Ohio school. Cincinnati escaped the bottom 10 without having to play a game. That's one good thing that came from the Bearcats not making last year's tourney -- though Cincy fans no doubt would've rather been dancing.
The fates of two other tourney underperformers are noteworthy. Oklahoma and Pittsburgh swapped positions in the bottom 10. Oklahoma fell from fourth to second worst on the PASE list by virtue of a first-round loss as a No. 6 seed. Meanwhile, Pittsburgh jumped from second to fourth worst with its one-win performance as a No. 5 seed.
PASE powers change in last decade
Six of the top PASE performing teams of the modern era do not appear on the top 10 list for the last decade. The biggest omission is Duke. Since 1997, the Blue Devils are actually underachievers, winning 29 games when they should've won over 30. (Of course, they were saddled with the expectations of eight top seeds over that time period.) The five other top PASE teams not to make last decade's list are Louisville, Michigan, North Carolina, Arkansas and Villanova.
The top PASE performer since 1997 has been Connecticut. The Huskies have won nearly one game more than they should have per tourney. Michigan State comes in a distant second, having fallen from the top position last year with its first-round exit. At least MSU stayed in the top 10. Four teams dropped out of last year's "decade's best" rankings: Syracuse, North Carolina, Arizona and Wisconsin.
To be fair, a rolling "10 year's best" list is bound to be fluid because while one year is gained, another is lost -- and those two years can work more in favor of some teams than others. Syracuse is a perfect case in point. In 1996, the Orange went to the finals as a No. 4 seed before succumbing to Kentucky. Last year, 1996 counted in the decade's-best PASE list. This year, that overachieving tourney performance doesn't count and is replaced by Syracuse's first-round flameout as a No. 5 seed. The result: a .562 PASE decrease -- and a precipitous fall from second place all the way to 14th place on the 10-year PASE list. The same problem hounded North Carolina (from seventh last year to 12th), Arizona (ninth to 13th) and Wisconsin (10th to 18th).
The four vacancies in the decade PASE list were filled by Purdue, UCLA, Florida and Maryland. With a +.670 PASE, Purdue jumped from 15th to fourth place, not because it accomplished anything last year, but because its second-round exit as a top seed in 1996 no longer counts against the Boilermakers. Plus, Purdue has quietly strung together three straight overachieving tourneys, highlighted by an Elite Eight run in 2000 as a No. 6 seed. Not bad for a program that has the lowest overall PASE in the modern era.
It doesn't take a bracket scientist to figure out why UCLA and Florida climbed into the 10-year PASE list from 26th and 27th place, respectively. Last year's finalists, the Bruins and Gators each added a huge overachieving season to their PASE values -- and UCLA shed a bad year in 1996 as a winless No. 4 seed.
The weakest PASE performers of the last decade
Among the seven newcomers to the 10-year list, Arkansas stands out. The Razorbacks are effectively the flip side of the Boilermakers: They rank as the eighth-best overachiever of the 64-team era, yet appear as the third-worst underachiever of the last 10 years.
Arkansas is one of only two teams that weren't on last year's decade-long PASE underachiever list. The other is Tennessee -- which had the misfortune of meeting the "five appearances" requirement and thereby found itself as the fourth-worst underperformer. The addition of Tennessee and Arkansas helped Charlotte and Murray State climb out of the PASE basement.
What to look for in the 2007 tournament
Freelance writer Pete Tiernan has been studying the NCAA Tournament for 17 years. E-mail him here. Or go to www.bracketscience.com and check out the BracketMaster, which lets you do your own bracket analysis. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||