NCAA Tourney Conference Overachievers and Underachievers (1985-2007)

Posted by rtmsf on July 11th, 2007

Today we’re ready to unleash the last installment of our analysis of NCAA Tournament stats of the 65 (64) team era… that is, unless we decide to analyze the coaches too… well, it is over three months until Midnight Madness, so ok, hold that thought.  Anyway, as you hopefully recall, during the weekend we took a look at the raw numbers of the era by conference, and essentially concluded that the ACC has been the most successful conference of the last 23 years, the Pac-10 SWAC/NEC the worst, and that the mid-major conferences may not have been as consistently good as we had hoped over the years.

Now let’s take a look at the conferences who have overachieved and underachieved over the 65 (64) team era. In our analysis of this measure by school, you may remember that we looked at two different models – a Standard Model of expected wins by seed (e.g., a #1 seed should win 4 games per appearance), and a Historical Model of expected wins by seed (e.g., a #1 seed has actually won 3.36 games per appearance from 1985-2007). We concluded then that the true value lies in considering the Historical Model foremost because the Standard Model places too unrealistic of an expectation on high seeds and not high enough of one on low seeds, which ultimately skews its results in favor of lower-seeded schools and conferences. Given that condition, we now show the Overachiever and Underachiever conferences of the 65 (64) Team Era using the Historical Model. See Table A below.

Table A. Historical Model applied to 65 (64) Team Era

Notes: the table is sorted by “+/- per App,” which represents the number of games won above or below the expected number of wins for that seed per NCAA appearance (1985-2007). The conferences whose names are in bold are BCS conferences. The conferences whose names are in red are conferences that no longer exist.

NCAA Tourney by Conf v.3

Not Just George Mason. The first thought everyone will have (because we had it too) is that George Mason‘s miraculous run in 2006 accounts for the Colonial Conference’s rather aristocratic pedigree at the top of our list. But looking a little further inside the numbers somewhat mitigates this idea. Sure, the Masonites (as a #12 seed) won 3.52 games beyond its expected value of 0.48 wins per appearance in 2006, but that only accounts for half of the Colonial’s wins beyond expectation during this era. So where are the rest of the wins coming from? Thank David Robinson’s Navy squads of the mid-80s and Dick Tarrant’s Richmond Spiders in the immediate aftermath for the CAA’s perch as the biggest overachiever on our lofty list.

David Robinson

George Mason isn’t the only CAA School to Overachieve

BCStriation. Unlike our previous posting that used standard objective measures (wins, F4s, titles, etc.) to show that the six BCS conferences were without question the top six leagues of the era, today’s posting paints a substantially different picture. A league can be very successful objectively and still considerably underachieve, as in the strange case of the Pac-10 (and to a much lesser extent, the Big 10). Although the Pac-10 was clearly the weakest of the BCS conferences by the raw numbers, we certainly didn’t expect that it would be the second-worst underachiever of the 65 (64) team era – but it unquestionably is. The Pac-10 has won sixteen fewer games than it should have during this period, which dwarfs the negative output of any other conference – next in line for public shaming are Conference USA (7.6 wins fewer) and the Big 10 (6.5). Looking back at our list of chronic underachieving schools, we note that Stanford, Arizona and Cal all fall into the frequent NCAA underachievers list, which should have tipped us off that this was coming.

High Achievers. On the other side of things, the ACC and the Big East fall in line behind the CAA as the biggest overachievers of the era, which proves that you can get great seeds, have tremendous objective success in terms of wins and titles, and still overachieve as a conference. The ACC has won a whopping 22 games and the Big East 18 games beyond expectation; and the SEC isn’t far behind with 13. We also want to nod a tip of the hat to the Mid-Continent (+4.5 wins), MAC (+4 wins) and Horizon (+4 wins) conferences, each of which shows that leagues with consistently low seeds can do some damage on a regular basis in the NCAA Tournament.

Bradley

Missouri Valley Teams Need to Do More of This

What About…? If anything, these last two posts have opened our eyes to just how traditionally overrated the Missouri Valley Conference has been. For a so-called mid-major who gets multiple teams invited every year, its performance leaves a lot to be desired (4 wins below expectations). We realize that things change – conferences get better and worse over the years – but the MVC is going to have to really start producing in the next 5-10 years to lose our proffered overrated tag. As a comparison, the Horizon and West Coast conferences have performed nearly as well (19 wins each) as the MVC (22 wins) despite earning far fewer NCAA bids and having a slightly worse average seed.

Ivy League Paradox. We suppose that if you asked a hundred college basketball fans whether they believed the Ivy League traditionally overachieves in the NCAA Tournament, 99 of them would likely agree. This is probably due to a memorable upset or two over the years in addition to a common perception that the Ivies are a “tough out” every year. But looking above, we see quite starkly that the Ivy League has been one of the biggest underachievers of this era, earning only three wins versus an expected total of seven. This is largely because the Ivy champs (usually Penn or Princeton) have consistently earned seeds ranging from #11-#13 over the last decade, but haven’t been able to earn a single win during that period. The lesson here, we suppose, is to never take an Ivy team in your brackets (we’ve heard that taking an Ivy team against the spread in the first round is a good bet, however).

rtmsf (3726 Posts)


Share this story

Leave a Reply