How to do NCAA expansion right: the 66-team At-Large Death Match Round

Continuing on with the discussion in my prior tweets and links and longer blog posts about the possibility of NCAA Tournament expansion, I want to expand on something I proposed in comments the other day.

If the NCAA is absolutely dead set on expansion, there’s a right way to do it, and a wrong way. The wrong way is to create a haphazard-feeling play-in round, with bizarre, uninteresting, and out-of-place matchups like #9 vs. #24 playing for the right to take on #8 (who cares?!), and #16 vs. #17 — both mid-level NIT teams in the present system — vying for the right to challenge #1. This setup would fail to reward teams on the top 4 seed lines for their excellent seasons because, instead of round-of-64 matchups against low-major auto bid teams that have little chance of winning, they’d get relatively solid, mid-60s RPI type teams, coming off big wins. At the same time, it would fail to give those low-major teams a chance at the truly magical, Hampton-esque moments that are the very best thing about the tournament.

Think about it: If, rather than being a #15 seed stunning a #2 with legit national-title ambitions, Hampton had been a #23 seed and had merely beaten a #10 seed — the sixth-place team in the Big 12, say, instead of its champion — who would remember it?

The right way to do expansion — the way to preserve the opportunity for unforgettably magical One Shining Moments like the one shown above, which are what make March Madness so special — is to do something like this:

Instead of expanding the field to 96 as such, replace the “opening round game” (i.e., the play-in game) with an actual opening round — a round that, crucially, isn’t considered part of the tournament, but rather is a final prerequisite for all potential at-large teams to make the tournament in the first place.

Since there are 31 auto bids, we need 33 at-larges for a 64-team tournament. So, have the committee select the 66 best teams that didn’t receive automatic bids — everyone from whichever two teams out of the Kansas-Texas-KState trifecta don’t win the Big 12 tourney, all the way down to BCS-conference mediocrities like Notre Dame and iffy mid-majors like Wichita State — seed those teams from #1 to #66, and match them up in 33 unbracketed opening-round games (1 vs. 66, 2 vs. 65, etc.). Then take the 33 winners, call them “NCAA Tournament at-large teams,” throw them into the field with the 31 auto bid winners, re-seed the entire field into a proper 64-team bracket (so a top-notch at-large team like Kansas/Texas/K-State could still get a #1 seed, and the auto-bid bottom feeders like Morehead State would still be #16 seeds), and go from there.

This, to my delight and Andrew’s chagrin, would preserve the frequent lopsidedness — but occasional absolute Hampton-style magic — of the games between #1-4 seeds and #13-16 seeds, since #s 13-16 would still be lowly auto-bid teams, as God intended. 🙂 Yet meanwhile, Andrew, and others who feel there aren’t enough competitive games in the early rounds, would still get their desired extra round of potential upsets, since the highest level teams (with the exception, basically, of the six auto-bid winners from the BCS conferences) would be playing in the opening round against teams whom they’d periodically lose to — and they’d have to win those games in order to call themselves “tournament teams,” and earn the right to play the Morehead State type teams in the first round of the actual tournament (i.e., the round of 64).

My scheme would somewhat (though not entirely) restore the old notion of the NCAA Tournament being a “tournament of champions.” Conference champs like, say, a Morehead State would actually get an advantage over a non-champion like, say, a Kansas team that loses to Texas or K-State in the Big 12 tourney. Morehead would automatically be in the tournament from the get-go, while the Jayhawks would have an extra hurdle to clear (i.e., winning an opening-round game over a Notre Dame-type team). If you’re a conference championship purist, this is a feature, not a bug. At-larges can join the party, but only when they’ve earned their way in — they don’t just get selected in by a committee, they’ve gotta win a game to get in (and it’s not that easy of a game).

Mid-major-haters and other non-tournament-of-champions-purists may say this isn’t “fair” to the Kansases of the world, teams from tough conferences like the Big 12 or Big East, for whom winning an auto bid is more challenging than it is for a team like Morehead State. To which I say: gimme a break. The Kansases and Kentuckys have more structural advantages over the Moreheads and Chattanoogas than you can possibly imagine. There’s nothing “unfair” about rewarding conference champions for being champions, and if this gives a tiny leg up to teams below the Red Line, it’s still like pissing into a hurricane in the grand scheme of things. If Kansas or Syracuse or Villanova or Kentucky is as good as you think they are, surely they can win one extra piddly little game to prove it. It’s simple: they gotta either win their conference tournament, or they gotta win an extra game against a team they should (but aren’t guaranteed to) beat. Once they do that, they’re in, they’re seeded properly, and they’ve got the same national championship shot as everyone else.

Having said that, I will happily concede — and again, I think this is a feature, not a bug — that my setup would probably increase somewhat (but not too much, IMHO) the number of upsets in the eventual #1-4 vs. #13-16 games during the round-of-64, for two reasons: (1) the quality of teams on the #1-4 lines would be slightly diluted, since you’d always lose a few teams in opening-round upsets that were otherwise destined for high seeds; and (2) perhaps more importantly, those #1-4 seeds who were at-larges (and thus had to survive an opening-round game) would be playing on short rest as compared to #13-16s, who would almost always be auto bids. Don’t think Morehead has a shot against Kansas? Neither do I — but if the Jayhawks, regular-season but not tournament champions of the Big 12, are running on fumes while their lowly opponents are well-rested, the odds increase slightly, certainly of a competitive game, and occasionally of a stunning upset. Which would be great for the tournament as a whole.

My system would also solve a major objection to expansion: the specter of mediocre-to-bad teams *cough*FireMikeBrey*cough* being able to rest on their flimsy laurels and claim they had a quasi-successful season by saying, in the height of lameness, “we made the tournament” by being in the field of 96. No, you didn’t: you only “make the tournament” if you either: 1) win your conference tournament, or 2) win your opening-round game.

Heck, if desired, in order to further emphasize that the opening round is separate and distinct from The Dance, the opening-round losers can be shunted over to the NIT. This would have the side effect of making the NIT more interesting to casual fans because, instead of being purely made of mediocrities, it’d have the occasional Kansas or Kentucky, a really good team that loses in its conference tournament and then loses its opening round game, and thus misses the NCAAs entirely, but is still super awesome and talented and fun to watch.

Another very improtant benefit: this would create a huge incentive to win your conference tournament, for everyone. Right now, many teams are guaranteed a bid (and in some cases a high seed) regardless of how they do in their conference tournaments. As a result, teams sometimes arguably “mail it in” a bit. In this new system, on the other hand, winning your conference tournament eliminates a potentially difficult opening-round game, protects you from the potential of a dangerous round-of-64 showdown with a perhaps inferior but well-rested opponent (indeed it allows you to be the well-rested one, if such a matchup occurs), and means you only have to win 6 games, instead of 7, to be the national champion. So it would make all conference tournaments more compelling, across the board, because everyone is fighting for something concrete (instead of the current situation, where a lot of BCS conference teams are just “playing for seeding,” which is an awfully amorphous goal, since nobody knows for sure what their status-quo seed is, or what their if-they-win seed is).

This conference-tournament enhancement issue goes from being merely desirable to being absolutely essential if we’re going to give automatic bids to conferences’ regular-season champs AND tournament champs, as some have suggested. If auto bids are given to both types of champions, yet the tournament champs aren’t given some inherent leg up in the tournament’s structure, then all regular-season champions — even in mid-major and low-major conferences — would have little or no incentive to play hard in their conference tourney. Indeed, low-major conferences in particular would have perverse incentives, since the whole conference benefits financially (and national profile-wise) if it’s a two-bid league, and the only way that could happen in a low-major conference would be if someone other than the regular-season champ wins the league tournament, thus garnering an “extra” automatic bid. Can you imagine the referee-conspiracy talk? The allegations of teams being bribed or otherwise incentivized to mail it in? Giving auto bids to both types of champions, without offering an extra reward for winning the tourney, would be just inviting the worst sort of conspiracy-theory nonsense. You’ve gotta give the regular-season champs some reason to care. My system would.

At the same time, Andrew’s goal of “everyone with a pulse” having a shot is fulfilled under my system. The only teams excluded from a shot at the Big Dance would be CBI-type teams (and below). Teams like, say, Hofstra in 2006, which was utterly jobbed out of a bid, would be solidly in the opening round, and we’d find out on the court whether they make the Dance or not. No one with a legitimate argument that they’ve played well enough in the regular season to deserve a a national-title shot would be able to reasonably complain about being excluded from the tournament.

Now, I don’t want to overstate this point, because admittedly, if the field expands (in whatever format) to include 60-something at-large teams, we inevitably will start to care about those final at-large teams (or potential at-large teams) who get in, and more to the point, the ones who just barely miss out. It may seem absurd now to worry about the #67 at-large team being excluded, but once it becomes reality, we’ll start to care more than we presently think we’ll care. Debating the merits of a 6-12 Big East team vs. a 5-11 ACC team vs. the third-place team in a “down” Missouri Valley vs. the fifth-place team in an “up” CAA will become the “new normal.” Before you know it, we’ll have Billy Packer bemoaning the Big East team’s exclusion, Dick Vitale singing the praises of the little guy, coaches bitching up a storm about their teams being jobbed, and Brendan Loy blogging furiously about how it’s an OUTRAGE that Pepperdine was excluded in favor of N.C. State, when the Waves, what with their close game at Gonzaga and only four WCC losses overall and third-place finish in the conference, plus a signature non-conference win over DePaul, clearly deserved the bid. (I’m making all that up, obviously, but you get the gist.)

But, although we will end up (to our surprise) subjectively caring about these things, 1) we still won’t subjectively care about them as much as we do about the current debates over #34 vs. #35 for at-large inclusion, and 2) however much we subjectively care, these debates will be objectively far less important, because our hypothetical Pepperdine and our our hypothetical N.C. State BOTH SUCK.

Meanwhile, office pools would be free to ignore the opening round, and still start with 64. Indeed, they’d basically have no choice, because of the reseeding — though I suppose people could run separate “opening-round pools” if they want. (I probably would!) Either way, there’d be no need to stock up on legal-length paper. 🙂

Oh yeah, and the entire opening round could be played on a single day, if so desired, to avoid teams having unfair advantages based on timing. This would become a totally awesome day of hoops for casual fans — 33 play-in games between teams with pulses, everything on the line, and nary a Morehead State in sight (unless we start giving auto bids to regular season league champs, in which case there’d be a few Moreheads, but not too many). Win and you’re in; lose and you’re done — or in the NIT or whatever. Brilliant.

Alternatively, we could go with the Jazz plan, adding a whole ‘nother weekend of sudden-death fun:

Will I be interested in #9 UNLV v. #24 Arkansas State and #16 Va. Tech vs. #17 Seton Hall [in a more traditional 96-team expansion format] to the extent that they are like the first week’s equivalent of the NFL wild card games? … [T]he wild card weekend is fun, and I’m thinking that UNLV v. Arkansas State doesn’t look fun. In not being fun, I won’t be interested, I won’t bother to find out which 8 seed that game feeds into, and I’m pretty sure I won’t spend my entire Tuesday keeping tabs on UNLV/Arkansas State, et.al.

By contrast, if the NCAA adds another weekend…and devotes the first weekend to 33 death matches, that looks like a lot of fun. There’s no need to recall which bracket the death matches feed into, as they won’t be determined until the smoke clears Sunday night. Each game would be like a life-or-death featurette of its own, which seems pretty fun.

(I realize that each game in the NCAA tournament is life-or-death, I just don’t think we fans process it as such. When 2-seeded Gonzaga is losing in the 2nd half to 10-seeded Nevada, that’s troubling for Bulldog fans, but a significant part of your attention is on what happened in the 3-seed minibracket, in case Gonzaga pulls it out, and who’s left in the rest of the region, in case they pull it out and pick up momentum. You’re rarely watching only one game without thinking about what’s going on in the rest of the region and the implications for the teams involved).

Death match weekend – it would be pure and simple, win and you’re in, and we’ll talk later about where you go, don’t worry about that now.

Either way, on a single day or in a Death Match Weekend, I think it’d work really well, and certainly way better than the awkward alternative of a bunch of clunky, UNLV-Arkansas State type games opening the tournament.

Another minor side-benefit of my plan: if, or rather when, a new splinter conference forms, or a Division II conference moves up to Division I, or whatever, thus increasing the number of auto bids from 31 to 32 to 33, etc., the solution would simply be to shrink the Death Match Round by 2, or by 4, etc., teams. This is structurally very, very simple, and feels like much less of a big deal than, say, reducing the number of at-larges from 34 to 33 would’ve been, back when the number of auto bids increased from 30 to 31 — which is precisely the dilemma that resulted in the creation of the farcical play-in game. The Gods of college basketball, in their infinite wisdom, decided that cutting out the 34th at-large team was too painful to contemplate, so they instead pitted two #16-seeded auto bid winners against one another (thus cheapening the very concept of an auto bid). If/when the same issue arises again in the future, under my system, I’m thinking that cutting out potential at-large teams #65 and #66 would not be quite so painful. And doing so wouldn’t affect the structure of the tournament at all.

The same is also true in reverse: if, because of Kyle Whelliston’s “Sports Bubble” or whatever other reason, a conference up and disappears, thereby shrinking the number of auto bids, we simply add 2 more teams to the Death Match Round. Again, unlike now, such changes would have no structural effect on the shape of the tournament, and would only impact teams we don’t really care about anyway (though see my earlier caveat about we’ll inevitably start to care somewhat, but still not as much as we care now about teams near the cut line).

So… that’s my proposal for expansion. I still prefer a 64-team tournament — eliminating the godforsaken play-in game, reducing the number of at-larges to 33 (the teams below #30 or so are almost always horribly flawed anyway), and going from there. I still think any 96-team setup is a bad idea on balance, not least because it will further devalue the regular season to such an extreme extent. However, if the NCAA is going to do expansion anyway, this is how they should do it. I even think it’d be pretty fun.