Over the past week or so we have heard much from two outspoken critics of the college basketball mid-majors--namely Billy Packer and Gary Williams.

A few years ago Packer famously blew a gasket a few years ago when St. Joseph's was named a 1 seed, only to see them fall one buzzer-beater shy of advancing to the Final Four (and substantially further than two other number 1 seeds that year which were knocked out in the first weekend). This year he complained about the inclusion of "too many" mid-major at-large teams instead of more teams from power conferences, noting the general success of the power conferences in recent years. The relevant comparison, it would seem, would be team to team--e.g., the 5th, 6th, or 7th place teams in the power conferences versus the 2d in the CAA or 4th in the MVC. Duke's success seems largely irrelevant as to whether Bradley or Maryland deserves an at-large berth.

As John Feinstein notes in his excellent column in today's Washington Post, the mid-majors have more than held their own through the first weekend of the tournament. In fact, he observes that the mid-major at-large teams were 4-4 in the first round, notwithstanding the fact that only two of them were higher seeded than their opponents.

As for Gary Williams, thet other outspoken critic of the inclusion of mid-majors in the tournament, Maryland went down last night in the NIT to--wait for it--mid-major Manhattan. At home. Moreover, Old Dominion and Hofstra, two CAA teams that didn't make the NCAA, both knocked out Big 12 teams in the opening rounds of the NIT as well.

As a George Mason fan, I obviously have my personal bias in favor of the mid-majors here. But it seems to me that there may be some lessons here for critics of the selection committee and the continued rise of the mid-majors in college hoops.


I missed this excellent post which looks at some of the numbers on the relevant comparisons between the mid-majors and power conferences.

Dutch (mail):
Not to get too political about this, but it's strikes me as a "group identity" kind of thing. I think it's difficult for some folks to separate Maryland, for example, from "ACC basketball" which lets them draw on Duke's strength, and the strength of past Maryland teams. On selection Sunday, when I heard the pundits (like Seth Davis) talk about "the history of the program" as a reason to let a team into the Dance, I think, what are you smoking? Because the 2002 Maryland team was good, that means what exactly about this year's team?

I understand that many people use conference as a quick heuristic to determine how good someone is, but isn't that what the RPI was designed for?
3.19.2006 12:17pm
DNL (mail):
In no universe is Manhattan a "mid-major." They're from a one-bid conference. Same with George Mason. Mid-majors are conferences that often get three or so bids, and have gotten a number of high bids.

That makes it more embarrassing for UMD, but that's neither here nor there.
3.19.2006 12:35pm
Steven Vickers:
George Mason is from the CAA, which had two teams, GMU and UNC-Wilmington, make the tournament, and a third, Hofstra, who just missed.
3.19.2006 12:45pm
DNL (mail):
Oh, yeah, that's right. I forgot UNC-W is there. My bad.
3.19.2006 12:49pm
John M (mail):
I am a graduate of a school with a major program (Indiana), but I enjoy the drama that the mid-majors add to the tournament. It's more fun to see a mid-major with a great record play one of the elite teams that to watch an underachieiving Michigan or Maryland team do the same. On the other hand, the "last in" major conference teams have acquitted themselves pretty well this year. By my count, there were five major conference at-large teams playing higher seeded teams in the first round, and three (Texas A&M, NC State, and Alabama) won, while two lost (Wisconsin and Seton Hall, which lost to mid-major Wichita State). The lower seeded mid-major at large teams went 3-3, so it's basically a wash.

The point where I disagree with Todd's entry is with the notion that tournament performance necessarily validates or invalidates a particular decision by the committee. The NCAA tournament is such a great event partly because the single-elimination format makes it so unpredictable. If the regular season performance of the at-large teams from the MVC and CAA warranted at large bids for those teams, they could have gone 1-7 and still "deserved" it. Similarly, Maryland's loss in the NIT doesn't prove that they didn't deserve an NCAA bid. Manhattan's players must have been thrilled to play a postseason game against a recent NCAA champion. On the other hand, Maryland's mere presence in the NIT rendered the Terps' season a failure.
3.19.2006 12:51pm
frankcross (mail):
Now this is an important topic. I have never been able to figure out how a 7th place team in a conference, no matter how good they are, belongs in the tourney. Go midmajors! They're not going to win the whole thing, but neither will the 7th place team. And they certainly make it more exciting.
3.19.2006 1:02pm
DNL (mail):

That's all true, but Maryland was hardly good against mid-majors this year, losing to Gonzaga, GW, and Temple. Their only win against a non-ACC team that made they tourney was a game against Arkansas at a neutral site (Maui). That's 2-8 against tourney teams, 16-6 otherwise.

Why does Williams think his team deserves a chance to demonstrate, again, that they can't beat good teams?
3.19.2006 1:06pm
John Thacker (mail):
Duke's success seems largely irrelevant as to whether Bradley or Maryland deserves an at-large berth.

Inconsistency alert.

In fact, he observes that the mid-major at-large teams were 4-4 in the first round, notwithstanding the fact that only two of them were higher seeded than their opponents.

Yes, and Wichita State's success seems largely irrelevant as to whether Air Force deserves an at-large berth. You can't compare the mid-majors that obviously deserved a bid to the marginal last teams in.

If the aggregate record of mid-major teams is significant for all mid-majors, then the aggregate 6-0 record of the ACC (including a win by a 10 seed over a 7) is significant for other ACC teams. If Duke et al.'s success doesn't matter for marginal ACC teams, then the record of better mid-major teams like Wichita State don't matter for determining whether marginal mid-majors get in.
3.19.2006 1:07pm
John Thacker (mail):
Why does Williams think his team deserves a chance to demonstrate, again, that they can't beat good teams?

In the case of Maryland, the team became much worse after losing Chris McCray to academic problems at the start of the spring semester. As a result of that, in no way did they deserve a bid. Maryland is a red herring because their higher power rankings mostly come from before injury. (The Temple game was afterwards; they routinely lose to local rival GW, even doing so their national championship year.)

Of course, Maryland's 2-8 versus tournament teams was better than Air Force or Utah State's record against such teams, easily.

You can't point to just the best mid-majors that obviously deserved bids. You have to look at the actual last couple at-large teams let in, not your Bucknells or Wichita States. Utah State and Air Force are the two mid-majors who in no way deserved their selections.

There are major conference at-large teams that I didn't think deserved it. I didn't think Seton Hall deserved to be in it anyway with their power rankings. Bradley had a very good Sagarin rating and deserved to be in. Missouri State deserved it over most of their MVC colleagues, but Southern Illinois winning the conference tournament hurt them.
3.19.2006 1:20pm
Steven Vickers:
John Thacker-

I'd agree with that, except Packer never seemed to complain too much about any specific (or "marginal" as you put it) mid-major teams getting at-large bids, just that there were "too many." From the article Zywicki cites: "During the interview, Packer said he disagreed with the committee's decision to give at-large invitations to Wichita State, Bradley, Northern Iowa and George Mason." Note that that's all three MVC at-large bids.

So, I think Zywicki is in the right using the results of all the at-large teams from the MVC, though I am a bit hesitant to say that tournament success validates an inclusion decision.
3.19.2006 1:21pm
John Thacker,
The point that Professor Zywicki was making was that Packer has set up a straw man and then proceeded to knock it down. He argues that Maryland, Cincinnati, Florida State, Michigan should get in instead of additional teams from the CAA and the MVC. But then, as justification, he argues that the best teams in the Big East, ACC, and other big six conferences consistently are the best teams in the tournament. Yeah, no kidding - and no one is arguing that the MVC and CAA teams are the best in the country - that's why the highest seed any of them got was Wichita State's 7. But Maryland, Cincinnati, FSU and Michigan aren't the best teams from the power conferences (this year). The relevant comparison is the middle teams in the power conferences against the best teams from the mid-major conferences.
Second, Steven Vickers' response is dead-on. Jim Nantz and Billy Packer specifically called out the CAA and MVC at-large teams, not the two from the Mountain West and WAC. So Professor Zywicki is only responding to the argument as Packer and Nantz set it out. Sure, they would have a better case about AFU and USU - but it's not the point they made.
Finally, I was annoyed last Sunday about Packer and Nantz. But if Bradley can finish off Pitt this afternoon, those two guys did the MVC the biggest favor they could have asked for, since it heightened everyone's interest in how the MVC (and CAA) do.
3.19.2006 1:51pm
frankcross (mail):
John Thacker, there's a logical problem with your argument. Because the Committee has a hard time comparing among midmajors. Actually Air Force gave Illinois a good battle. But the point is that they have a hard time knowing if WSU or AF is the good one. By contrast, it's quite easy to know which major conference teams are the good ones -- they play each other. We can say with confidence that Duke is better than Maryland.
3.19.2006 2:30pm
Trent McBride (mail) (www):
I have a post comparing the two otypes of schools over the history of the tournament seed-for-seed:
3.19.2006 2:35pm
JLR (mail) (www):
Bradley beat Pittsburgh. Way to go Bradley! (But terrible for my NCAA Bracket -- Pitt was one of my upset specials to make the Final Four. Luckily, there's no money involved; all I will feel is the agony of defeat.)

One comment: regardless of Packer's fallacious reasoning, I think there is no question that Cincinnati deserved to be in the NCAA Tournament more than Northern Iowa did.

Northern Iowa beat Iowa and Bucknell. Cincinnati beat Syracuse and West Virginia, and lost to Villanova by only two points. Northern Iowa was clearly the fourth out of the four MVC teams chosen, and Cincy deserved the NCAA bid more in my opinion.

But as has been noted above, Packer specifically called out ALL the MVC at-large teams (as well as George Mason). Bradley and Wichita State have both gone on to earn spots in the Sweet 16, thus refuting Packer's bias against the MVC.

But it must be said that just because Bradley (an MVC team) beat Pittsburgh (a Big East team) does not mean that Northern Iowa (an MVC team) deserved to be in the NCAA tournament over Cincinnati (a Big East team).

Nevertheless, it is clear that Wichita State and Bradley (and George Mason) clearly proved that they deserved to be chosen for the NCAA tournament.
3.19.2006 2:50pm
frankcross (mail):
Cincy was 0-2 against Pitt and Seton Hall.
Northern Iowa was 3-1 against Bradley and Wichita State.
3.19.2006 3:08pm
Excellent post at your site - I was hoping someone less lazy than I would do this work. Two points - I agree about the non-BCS schools seeded five and higher and that's precisely why I say in my comment that it is clearly the case that the very best teams in college basketball are almost always from the power conferences. I'm glad Gonzaga managed to play to seed this year, but they haven't in the recent past, and Nevada was seeded too high this year.
Second, I think there is a more obvious explanation for the difference in the records of 13 seeds (and provides a good reason to omit the records from your overall data points). The vast majority of non-BCS 13 seeds are from conference tourney winners in one bid conferences. All of the Big Six 13 seeds are at-large teams. So in order to make the relevant comparison, you would have to weed out the at-large non-BCS 13 seeds and compare them to the BCS 13 seeds. This year, there were two automatic qualifier 13 seeds and two at-large 13s (all 4 were non-BCS). The at larges went 1-1; the automatic qualifiers went 0-2.
3.19.2006 3:13pm
anonymous coward:
"Cincy was 0-2 against Pitt and Seton Hall.
Northern Iowa was 3-1 against Bradley and Wichita State."

I like the transitivity game. Wait, I got one: Duke lost to Georgetown. South Florida beat Georgetown. Yer mom could beat South Florida.
3.19.2006 3:51pm
Jim Hu:
Congrats to George Mason...what a great win over UNC.
3.19.2006 4:22pm
anonymous coward:
Wow, taking out the defending champs! Congrats GMU.

(To think I nearly had UNC in my boneheaded final four with Kansas...)
3.19.2006 4:24pm
John Thacker (mail):
John Thacker, there's a logical problem with your argument. Because the Committee has a hard time comparing among midmajors. Actually Air Force gave Illinois a good battle. But the point is that they have a hard time knowing if WSU or AF is the good one. By contrast, it's quite easy to know which major conference teams are the good ones -- they play each other. We can say with confidence that Duke is better than Maryland.

There are a tremendous number of computer rankings and various tools which do an excellent job comparing teams. You can choose different sorts of criteria, if you want-- ignore scoring margin, weight strength of schedule more or less, whatever.

It was incredibly obvious before the tournament that Wichita State was better than Air Force or Utah State, at least if you paid attention to any one of the many, many different ranking systems out there, or simply paid attention to mid-major basketball. It was equally obvious that Missouri State was much better than Air Force or Utah State. Missouri State clearly deserved a bid over those two. In case you didn't notice, I don't think Maryland deserved a bid, not without McCray. Florida State and Cincinnati had better cases; both easily deserved a bid over Seton Hall.

Congratulations to George Mason. As a Duke fan I have no trouble saying that. :)
3.19.2006 4:36pm
John Thacker (mail):
Here's a list of (pre-Tournament) power rankings of various teams, using 37 different systems. Wichita State is ranked mean 34.26, median 30.5 (their schedule strength hurts them in systems which ignore margin of victory), with a standard deviation of 9.6.

Air Force, on the other hand, was ranked mean 53.10, median 51.0, standard deviation of 12.84.

That's across 37 quite different systems, some of which count margin of error, some of which don't, and use all sorts of different formulae.

Seton Hall (55.73 mean, 54.0 median) and Utah State (64.84, 61) were even worse.
3.19.2006 4:43pm
Eric Jablow (mail):
The problem with major/mid-major/minor conference comparisons is that during the regular season they rarely play on even terms. Major conference teams schedule 'guarantee' games with minor conference team as tune-ups, true. But they are always at-home, in situations where they are nearly guaranteed to win. It's the same with NCAA football and the BCS. They stack the deck, and then they whine that the outsiders try to unstack it.

Do we see Georgetown/GMU in the regular season? Villanova tried to cripple the Big Five, the Philadelphia City Series, in the 90s; for a couple of years, they didn't play a full round-robin. Also in the early 90s, Villanova and Seton Hall canceled their long-standing series with Princeton; they learned that even beating them could be most embarassing.

It isn't just cowardice; it's restraint of trade and mercantilism. I know: this is supposed to be an amateur sport—for the players—but it isn't at the university level. The majors have the power, and they want to keep it. So, they use their economic power to legislate their continued success, and analysts like Billy Packer cooperate.

Frankly, I think that professional sports leagues in the US would work better with a promotion/relegation system, for similar reasons. That would have prevented the ridiculous spectacle of the Expos/Nationals disputes. It would require the teams acting like capitalists, though, and the rent-seeking owners just can't have that.
3.19.2006 5:06pm
frankcross (mail):
Great site, John Thacker, but it shows how inaccurate these things are at predicting tournament performance. Nos. 6, 8, and 10 are already gone
3.19.2006 7:09pm
I don't think the rankings are menat to predict the torney winners, frankcross. Tourneys are small-sample size events, by definition, and (just like in when they pick NBA lottery places) the team that's better or has the best "chance" to win doesn't always do so.
3.19.2006 8:09pm
Tennessean (mail):
Without looking at any numbers, I am amused that every year is proclaimed by someone the year of the mid-major and the year that mid-majors break through, as if this year's performance is all that anomolous.

(See Richmond however long ago doing what Mason is doing this year.)
3.19.2006 9:49pm
cdow (mail):
When was the last time a "mid-major" won the title? I think it was Marquette in 1977 (they were independent at the time. Ever since then, the title has been won by medmbers of power conferences--the Big East, ACC, Big 10, Big 12, or the Pac 10. The truth is, the power conferences still dominate, although the gap is narrowing.
3.19.2006 10:22pm
cdow, UNLV in 1990. Before that and after Marquette, Louisville won in 1980 and 1986 when they were a member of the Metro Conference, which was composed of a number of teams (including Louisville, obviously) that are now in big six conferences (as is Marquette).
3.19.2006 11:52pm
The Metro Conference also had some members that aren't in power conferences now, including Tulane, Saint Louis, Southern Miss, among others.
3.19.2006 11:59pm
Let's make the first 3/4 of the season round robin, 10 losses and you are out of the tournament. A win by more than 15 points counts as a loss.

All the teams who have 9 or fewer losses are in the tournament. The sections are set by geography only - northeast teams play northeast teams, se teams play se teams, etc. In each geographical section, the teams are "lined up" in alphabetical order by the school name, eg. blue devils, hoyas etc. The first plays the last, aa vs zz etc.

Each round is home/home/home(3rd if necessary). For each pairing, a coin flip determines which of the first two games determines the home team in a third game - coin flip before the first game, heads the winner of the first game gets the 3rd game at home, tails the winner of the second game gets it.

Each subsequent round, the teams are realigned by their mascots and play best 2 of 3.

The final four teams play best 4 out of seven.

That might last as long as the NBA playoffs.
3.20.2006 12:46am
AF -- another one (mail):
3.20.2006 1:55pm
Justin (mail):
Slightly off topic, but I'm of a mixed mind over whether "wait for it" is a useful literary device or not.
3.20.2006 3:40pm
cdow (mail):

"cdow, UNLV in 1990. Before that and after Marquette, Louisville won in 1980 and 1986 when they were a member of the Metro Conference, which was composed of a number of teams (including Louisville, obviously) that are now in big six conferences (as is Marquette).

I'm not sure I would call UNLV or Louisville "mid-majors" in the sense of a George Mason. Certainly, they were hoop powerhouses for a long time. However, your point is well taken
3.20.2006 9:33pm
John Thacker (mail):
Great site, John Thacker, but it shows how inaccurate these things are at predicting tournament performance. Nos. 6, 8, and 10 are already gone.

Except that they're as accurate as anything, of course. Basketball has a strong random outcome. If you enter these things in pools, you'll do very well, at least by expected value of total points. (Expected chance of winning is a little different, and you presumably in a contest want to maximize expected payout, which rewards going for somewhat unusual things.) Just as you'll do extremely well by playing the team that the line favors each time.

You consistently can't do better than the betting markets, and the best of the power rankings are very close to the betting markets. Sure, some people beat the market, but I wouldn't count on it.

UNLV was certainly as much of a mid-major as Gonzaga is now. The Big West was not good when they dominated it.
3.21.2006 1:31am
Obviously there are some really, really good mid-majors. But its a conference designation, so UNLV has to qualify. If every good mid-major is then considered a major, it is a self-fulfilling prophecy to say that the mid-majors aren't good. Regardless, they clearly aren't as good as the power conferences at the top. But the top of the mids are close to the middle of the big 6. We'll see if that trend is slowed by the NBA's new age requirement which will force some high school players to attend at least a year of college.
Interestingly, Memphis is technically a mid-major too although no one mentions it. CUSA was a bad conference this year past the top 4 and figures to get worse.
3.21.2006 1:41pm
dfslk (mail):
I'm a Florida State fan, and I was furious that they were passed up - by Air Force and Utah State. As much complaining as I did after the selection, there isn't a person in the world that can claim I ever once said "FSU deserved to get in ahead of George Mason, Bradley, or Wichita State." (And I think most of the complaining from others about George Mason came either because their selection ahead of Hofstra - another mid-major - was questionable, and because George Mason's AD was a member of the committee.) I've spoken to a number of FSU fans (yes, this is as anecdotal as anything else we're discussing). Our outrage was never focused on any of the teams currently in the sweet sixteen.

If performance in the NCAA tournament validates or invalidates entry, then Air Force and Utah State didn't deserve to get in, right? Oh, no, I guess the argument changes then to "well, they weren't supposed to win, because they got bad seeds." And if FSU, Cincinnati, or Missouri State win a couple games in the NIT, that proves that they deserved an at-large bid, right, since Maryland losing in the first round of the NIT apparently proved that they did not? Performance matters if it proves your point, but not if it doesn't.

I always believed that the MVC deserved 4 teams in the NCAA tourney. I also wouldn't have been quite as mad if Missouri State made it in above FSU - at least they had a reason for inclusion. Other than wanting to get two teams in from the MVC and WAC (an impermissible reason according to the rules), there was no justification that the committee could use to get Air Force and Utah State in that wasn't full of holes.

The problem I had with the selection was that the committee came out and talked out of both sides of their mouths. They simply did not provide a plausible narrative for why they selected or excluded certain teams. On one side non-conference schedule mattered - unless it came to Utah State, which had a non-con sked on par with FSU's - and Utah State lost some of their weak non-conference games. On one side a close loss to a highly seeded team (Utah State to Nevada) mattered (source: Andy Katz citing an anonymous committeeperson trying to defend the selection), unless it came to FSU, when the committee chairperson said that FSU's close losses to Duke, UNC, and BC couldn't be used to help it (by a combined total of 5 points). On one side - Air Force's side - beating a mediocre ACC team (Miami, without one of their best players) and a bottom-of-the ACC team (GA Tech), were counted as quality victories, but when it came time to look at FSU, their 9 ACC victories were downplayed as insignificant because 8 of them came against the middle and bottom of the ACC (such as Miami and GA Tech). FSU's early conference tourney loss to Wake Forest was seen as fatal. Air Force's early conference tourney loss to sub-.500 Wyoming? Eh, not so much.

My outrage at the committee came from their hypocrisy, not the inclusion of mid-majors. If they really, truly believed that Air Force was a better team than Cincinnati, FSU, Maryland, or Missouri State, then fine, but they haven't provided a consistent argument to make me believe that was the case. Letting the committee off the hook for their questionable picks, their unwillingness to adequately address conflicts of interest both in the committee makeup and in their ownership of the NIT, and their preposterous explanations for why they took some and excluded some bubble teams should not be the result of some mid-majors making it to the sweet sixteen.

For the record, I also have never bought into Billy Packer's notion that past tournament performance should be factored into selection and seeding. It is a silly idea that is bound to go nowhere.
3.21.2006 6:38pm
Jim Copland (www):
Missouri State and Hofstra should have gotten bids over Utah State and Air Force. Conference politics stopped it. Despite the rhetoric about conferences not mattering, it was easy to give MVC and CAA 4 and 2 instead of 5 and 3, merits notwithstanding. Air Force and Utah State had conference committee representation, so their conferences got 2 each, despite being undeserving (Utah State supposedly impressed the committee with a close game against Nevada, which was substantially overrated by the flawed RPI).

Maryland did not deserve a bid, period. Florida State had 2 good games against Duke, but they have only themselves to blame for scheduling complete patsies for their non-conference schedule. Michigan completely collapsed. Cincy should have gotten in over conference rival Seton Hall.
3.23.2006 5:12pm
Jim Copland (www):
P.S.: Gary Williams has reason to be self-serving; Packer sometimes makes good in-game analysis, but in this instance he didn't know what he was talking about. The committee's politics were terrible this year, but he focused on absolutely the wrong things.
3.23.2006 5:14pm