At times, the NCAA Tournament bracket makes sense mainly as a plan to maximize frequent flyer miles for some teams while allowing others with similar records to enjoy all the comforts of home.
At times, the NCAA Tournament bracket makes sense mainly as a plan to maximize frequent flyer miles for some teams while allowing others with similar records to enjoy all the comforts of home.

Bracket reaction 2014: "We're going where?!?!" "And playing whom?!?!"

Publisher
March 18, 2014 - 6:45pm

LOS ANGELES -- You've seen it all before: Selection Monday. Teams, coaches, fans and family members gather in front of big screen TVs to await the announcement of the NCAA Division I Women's Basketball Tournament bracket. Conference tournament winners sport championship hats, t-shirts and related paraphernalia, enjoying their chips and buffalo wings, at ease in the knowledge that they're in, and it all boils down to how they are seeded and where, and against whom, they'll be playing their first few games. With the exception of conference champs who are new to that role, and thus likely to go nuts just to hear their team's name mentioned, the champions pretty much take the whole event in stride.

Down a tier, but not by much, are the teams from the "Big Five" conferences who have assembled winning records and knocked off a "big dog" or two; they might lack the conference championship regalia, but they're nearly equally relaxed, secure that they've locked up their tournament bids and that for them, as well, it's all just a matter of seeding and placement. For these teams, too, the reaction to bracket announcements is predictably happy but muted.

And then there are the teams who make the hour-long Selection Shows watchable: The "bubble teams," whose good but less than wow-ing records, leave them hunched forward toward the screen, holding their collective breath as they nervously wait to see whether their team's name will be called at all. If, when, that moment arrives, these viewing rooms will erupt in back flips, hugs and high fives, the players and coaches caring little in that moment just how they are seeded or where they'll be headed, elated simply to know that they'll be going dancing this post-season after all. (Mercifully, the TV cameras tend to spare us much footage of "the agony of defeat" on the part of teams who sit through the hour fruitlessly, then head off to their dorms while coaches await word from the WNIT.)

But another tradition, equally entrenched, that goes hand-in-hand with Selection Monday, often tipping off even before the credits roll on the Selection Show itself -- and that's the carping, the griping, the second-guessing, most often from fans dissatisfied with how the Selection Committee has done a nearly impossible job. There's virtually no way in this process to make everybody happy. But to be fair, the Selection Committee brings a considerable amount of the naysaying on itself, when it attempts to justify anomolous results with appeals to a litany of ostensibly objective selection, seeding and placement procedures they profess to regard as close equivalents to tablets come down from Sinai, rather than forthrightly admitting the high degree of subjectivity involved in attempting to put the best 64 women's college teams in America on the floor.

There are, indeed, a fairly extensive set of written rules and procedures governing the selection and bracketing process, and by all evidence, the Committee members strive to adhere to them. Beyond that, their process is guided by thousands of hours of insight gleaned from watching the candidate teams in action -- whether in person or on film -- and by access to a database of statistical information that would leave us stats enthusiasts salivating.

The problem is that the rules themselves are neither intuitive nor self-evident. Left to begin anew from scratch, a reasonable person might well decide to adopt a rule that would avoid any rematch before the Final Four between teams that have met more than once in the regular season, regardless of conference. Or to always avoid requiring a higher-seed to play on a lower seed's home court, regardless of geographic considerations or the possibility of premature intraconference rematches. But, for better or worse, the rules are those that the NCAA, acting primarily through the aegis of conference administrators, has chosen to govern its annual tournament, whether or not the fan base, or even the players, would necessarily agree with those rules, either in the abstract or as applied in a particular case.

Making matters worse, the NCAA's selection principles and procedures will almost inevitably come into conflict multiple times even within the same selection weekend. When that happens, committee members can and do differ on the weight to be applied to each of the guidelines in conflict. Which matters more, for example, avoiding a meeting between two teams from the same conference prior to the regional final? Or placing a team in the most geographically proximate venue, in strict seed order? The committee is supposed to accomplish both, which cannot always be done, and has led to a form of gerrymandering in which the committee is allowed to move a team across a seed line, and/or up or down a line, when needed to make the rest of the principles jive. In other words, performance will take a back seat to logistics -- the committee might agree that University A should be a 3-seed, based on its record, and that Z College should be seeded 4, but Z could end up as anything from a three to a five-seed, and A anywhere from a two to a four, if the committee finds that necessary in order to avoid intraconference rematches between completely unrelated Teams C, D and E, while still allowing Teams Q, R and S, who are also unrelated to A or Z but are hosting, to play on their own home floors. But when the results are announced on national TV, there's no footnote explaining any of that, and fans are left to wonder why A, with its 25-5 record is a four-seed (below teams with much worse records), while Z, who squeaked in at 21-9, and a lousy strength of schedule ranking, was seeded third.

Then there's the fact that there really aren't any rules governing the inherently subjective process of drilling down into the "nitty gritty" to evaluate a team's performance for purposes of either seeding or selection. What counts more, for example, the number of a team's wins and losses, or the strength of a team's schedule? Which weighs most heavily -- a team's RPI, its strength of schedule, the number of victories it has pulled off over Top-50 opponents or that single uncharacteristic loss to one of the NCAA's bottom feeders? Should a team be penalized in seeding if it has lost a key player or two to injury late in the season, but has nevertheless acquitted itself reasonably well? How about the team that has a solid overall record, but has been tanking down the stretch? Most will tell you you need to take account of all of these factors, but you'll have nearly as many approaches to how they should count as you have committee members.

Finally, toss into the mix the widespread and well-justified perception that the manner in which the Selection Committee applies its procedures and principles tends to vary wildly from year to year and from one circumstance to the next and you've got the perfect recipe for confusion and discontent. Case in point? Wasn't it just three years ago that Baylor's Kim Mulkey got her hand-slapped by the NCAA for having the temerity to question the Selection Committee's decision to put her Lady Bears in the same bracket as Texas A&M, at the time a Big 12 conference rival whom Baylor had already faced three times over the course of regular-season and Big 12-tournament play? Didn't the committee that year explain its decision as a form of rigorous adherence to the geographic principle and the committee's bracket placement rules?

How then does one explain the decision this year to place No. 1-of-No. 1 seeds, UConn, in the Lincoln Regional -- a section of the bracket almost universally seen as offering the easiest path to Nashville -- as opposed to the Louisville Regional, which is more than 500 miles closer to Storrs, Conn., than is Lincoln, Neb.? True, ESPN's broadcast teams have been whining for weeks about the prospect of having the Huskies face conference rival and last season's national runner up for a fourth time in the regional final -- but that immensely unpopular result is exactly what the NCAA contended its rules required for Baylor and A&M in 2011. It might be different had Louisville earned a one-seed, and a case could certainly be made for such a seeding. In that situation, the Committee could point to the need to keep Louisville at home, under the hosting rules, and therefore to send UConn to the next closest site, which would obviously be Lincoln. But that was not the case this year. The Cardinals wound up at home with a three-seed and could readily have been kept at home, while still complying with the geographical principle and sending the Huskies where they belong -- i.e., Louisville.

And that is the nature of much of this year's moaning and groaning in the wake of Monday's bracket announcement. Unlike many seasons, where the gravamen of the complaint is addressed to teams believed to have been unfairly passed over, this season the bulk of the discontent is with the seeding and placement. The reason for this is plain to see: For just the second time in the history of the tournament, two teams -- Connecticut and Notre Dame -- will be entering the tournament with perfect records. Sure, you can kvetch, if you choose about UConn having had an easier path to its undefeated season with its American conference season having been played in the suddenly soft remnants of the old Big East. (The new American Athletic Conference had an RPI of six, behind all five of the "major-leaguers," and even that was thanks largely to the presence of Connecticut and Louisville, therein. Just wait until next season when Louisville and Rutgers depart the league for new conference homes.)

But whoever it is that handles Connecticut's scheduling had the foresight to address that potential criticism with the preconference slate, and with wins over Stanford, Maryland, Penn State, Duke, Cal and Baylor all on the Huskies' résumé alongside three wins apiece over conference-mates Rutgers and Louisville, who is it, exactly, who can point to the schedule and argue it was too "soft" to warrant the No. 1 overall seed? It's a universe of one -- that being Notre Dame -- which sports a marginally better RPI (No. 1, to Connecticut's 2) and a significantly better strength of schedule (4, to UConn's 19). But the Irish got the other undisputed one-seed, they're playing their regional at home in South Bend, and they're on the opposite side of the bracket and won't have to meet Connecticut before the national title game, assuming both teams go the distance. So really there's nothing for Notre Dame to complain about (and no one has heard a peep out of them anyway).

From that point on, however, there are an easy half-dozen potential one-seeds, and only two more of them to pass out. Each of the next six teams in the RPI rankings -- Stanford, Duke, Tennessee, South Carolina, Baylor and Louisville -- boasts a strong record, a respectable SOS ranking, several marquee wins ... and a couple of warts. And the decision to award the final two one-seeds to two teams from the SEC -- Tennessee and South Carolina -- over the other contenders is what much of this season's woofing is about.

For the most part, you could say it doesn't really matter all that much. With the lone exception of Louisville, who drew a three-seed (we'll come back to that), none of the six contenders listed above wound up with less than a two-seed. That means that in order to make it to the Final Four, the losers are going to have to get past the teams that won one-seeds, which in some shape or form, they would have had to do anyway had it been the "losers" who drew the one-seeds and the winners who wound up with twos. The difference between facing the 16th-seed versus the 15th in the opening round is fairly negligible, and though the stakes rise a bit at each step along the way, in the end, had the Cardinal been the top seed in the the Stanford Regional and the Gamecocks were sent there as the No. 2, the final match-up would have been identical.

The same, of course, cannot be said for Duke, which found itself sent as the two-seed to Lincoln, with another date with its nemesis Connecticut as its route to Nashville, nor of Baylor, the two-seed in South Bend, where it will need to go through the home-standing and undefeated Irish. Obviously, either of those teams would have much preferred to have been playing in Stanford or Knoxville, which would have been the case had they wound up as No. 1 seeds. Still, those seedings are certainly defensible. Putting aside whether it is fair to penalize Duke, a team that played the nation's toughest schedule and finished second only to the undefeated Irish in both regular-season and conference tournament play, simply because two key players suffered late-season injuries, Duke had the worst win-loss record (27-6 overall, 14-5 in ACC) of any of the six teams thought to have been in the mix for top seeds. All of the Blue Devils' losses were to top-drawer opponents -- to (in order of RPI) No. 1 Notre Dame (three times), No. 2 UConn, No. 21 North Carolina (twice) -- and Duke has plenty of Top-50 wins to point to, but still, there were six of those loses, more than anyone else.

Regrettably, however, when discussing the apparent snub to Duke with the media, Selection Committee Chair Carolayne Henry did not discuss any of those factors; instead, she pinned the decision almost entirely on its loss of key players to injury:

"Duke was in the conversation," Henry stated. "Obviously, they're on the two-line. When we watched numerous games over the season, we watched Duke quite a bit.  We thought Duke did quite well, even though they had losses of the two key players.  That factored in.

"When we're looking at the teams, obviously in terms of getting the team into the tournament, we look at what those teams do over the entire season," Henry continued. "Then when we get to the seeding process, we really look at who the team is today. Today in looking at Duke, while they've done quite well with those injuries to the two key players, who they are today is different than who they've been throughout the season."

The availability of key players is something the rules allow the Committee to look at, but there's certainly no rule that requires them to down-seed a team that's been bitten by the injury bug. That kind of explanation can't help but sound a lot like adding insult to injury (no pun intended), especially for Blue Devils' fans who may well feel that if anything, the team deserves more credit, rather than less, for grinding out an outstanding record, against the nation's toughest schedule, despite the adversity.

Baylor's case is stronger than Duke's in some respects, but worse in others. In the plus column, the Lady Bears (29-4) suffered only four losses all season, and that's fewer than any of the other contenders but Stanford (29-3), who lost only three, and Louisville who finished at 30-4. Plus, Baylor won its conference tournament and tied with West Virginia for the regular-season championship in the Big 12 (a league that might not have been quite as strong this season as it has been in years past, but still fields plenty of top teams). Two of the losses were to Kentucky (RPI 10) in a 133-130 quadruple-overtime instant classic and to UConn (55-66) in one of the closest games the Huskies played all season. There was no shame in the late-season loss to West Virginia (RPI 12), either, but that 60-76 defeat at Kansas (RPI 127), on the other hand, was the kind of loss that bites you right in the behind. Also on the down side: Baylor played a powderpuff out-of-conference schedule, and its SOS ranking (29) is one of the worst of the six contenders, besting only Louisville's SOS of 51.

Beyond that, it is tough for a school to contend that it was somehow shortchanged by being given a seed consistent with what its RPI would have warranted. Certainly, a team can argue that there are extraneous factors that should be considered, and at times the Committee will indeed give a team more than its RPI alone would have earned. But if your RPI puts you on the two-line, and you're seeded as a two, it's hard to make a case of grave injustice. And that's the case with Baylor, whose RPI of seven puts it near the low end of the two line.

Stanford, however, has grounds for complaint in having not just one, but two, teams from the SEC, with poorer records, lower RPIs and, at least in the case of South Carolina, significantly weaker schedules, seeded over it. Stanford (29-3) suffered only three losses all season. South Carolina (27-4) dropped four games; Tennessee (27-5) lost five. Stanford's RPI (3) is better than both Tennessee's (5) and South Carolina's (6). Tennessee boasts a moderately better strength of schedule (SOS 5 to Stanford's 9), but South Carolina's only noteworthy preconference game was against North Carolina (RPI 21), resulting in a far SOS of 21 for the Gamecocks. Tennessee won its conference tournament, which Stanford failed to do, but the Cardinal won the regular-season conference championship, while the Lady Vols did not. Conversely, the Gamecocks, like Stanford, won its league's regular-season crown, but fell in the tournament semifinal. So why is South Carolina the one-seed in the Stanford Regional, while the Cardinal are pegged as the No. 2?

This time, Henry explained the decision in terms of the RPIs of the teams to whom Stanford lost:

"Again, when we looked at Stanford and we looked at distinguishing the teams that were on the first line, those teams did not have any losses above the 50 [RPI]," said Henry. "[W]hen you look at South Carolina, they had no losses lower than 21 in the RPI. That's where that loss for Stanford came [in], to Washington, at [RPI] 89. They also lost to USC at 34."

The quality of both wins and losses is a factor that has long been part of the decisionmaking process. The problem here is that it appears the committee selected one particular factor -- the RPI of the opponents to whom a team lost -- and elevated that over all else (e.g., overall record, RPI, strength of schedule, and for that matter, the RPIs of the teams you defeated), something the rules don't prohibit, but neither do they require. Moreover, since RPI already encompasses schedule strength, this procedure seems to be putting a thumb on the scale, by weighing the same thing twice.

What South Carolina's lower SOS ranking is supposed to communicate is that across the board, Stanford played more high RPI opponents than the Gamecocks. What Stanford's higher RPI, combined with SOS, is supposed to communicate is that on top of playing the tougher schedule, Stanford beat more of those higher-ranked opponents than did South Carolina. Now Henry tells us, in effect, none of that matters, because in deciding on who gets the one seeds, we care more about the RPI of the one or two teams to whom you lost, than we do about the RPIs of all of the teams you defeated. That's a defensible rule, though I personally might not agree with it, but if that is how seeding is going to be decided, why not add it to the principles so everyone knows the rules of the game in advance?

Still, why does it matter, if Stanford is going to have to beat South Carolina in order to get to the Final Four regardless of who holds the one and who holds the two seed? Primarily because seeding dictates where and against whom you're going to be required to play in the opening rounds. As opening round hosts, Connecticut was going to be opening the tournament in Storrs, and Tennessee in Knoxville, regardless of whether they were seeded one or 16. Since South Carolina and Notre Dame weren't hosting the opening rounds -- and for that matter neither was Stanford -- all three teams were going to have to be sent somewhere else.

But the No. 1 seeds got a level of protection not accorded to those who weighed in at No. 2. As a one-seed, Notre Dame was sent to the nearest neutral opening-round site, Toledo, just two hours and 25 minutes (157.1 miles) away from home.

Had Stanford gotten the one-seed, it would likely have been sent either to Los Angeles or Seattle, both of them neutral sites, as the hosts are not playing, and both of them reasonably close to Palo Alto - L.A. is roughly a five-and-a-half hour drive (355.3 miles); Seattle is farther at nearly 13 hours (836 miles), but both are serviced by frequent, relatively low-cost express flights, and many Cardinal fans had already booked tickets to both sites, Tara VanDerveer told the media Monday evening after the brackets were announced.

Instead, as a two-seed, Stanford was shipped off to Ames, Iowa, and at 1,862 miles away, few fans if any are likely to be making the 27-hour drive. With flights currently averaging in the mid-700-dollar range, it's unlikely that many will be winging their way to the Cornbelt either. Perhaps more importantly, however, as a two-seed, Stanford was not afforded the protection of playing at a neutral site. Assuming chalk holds, Stanford will be facing seventh-seeded Iowa State (20-10, RPI 39, SOS 35), playing on its home court, in the second round.

With its one-seed, South Carolina did get a neutral site, where after racking up their frequent flier miles (Seattle is 2,828 miles from Columbus, S.C.), the highest-seeded team the Gamecocks could face in either of the first two rounds would be eighth-seeded Middle Tennessee, who will be far from home as well. (That eight-seed is another bone of contention, for a team with a 29-4 record and an RPI of 17, albeit an SOS of 79.)

That's what got the typically unflappable VanDerveer a bit hot under the color in yesterday's press conference. "I'm part Irish," she told reporters, "but I don't feel lucky."

VanDerveer was less concerned with the seeding itself -- "A No. 1 seed didn't do us any good last year," she stated. "That's the committee's opinion, that's their prerogative." -- as she was with the placement in hostile territory on the far side of the country:

"I was a little less surprised about [the two-seed] than the fact that we are going to a [host] site. All I hear from the NCAA is about the fans, the fans, the fans. Well, they just cut out 3,000 of ours. ... I think we would have helped the draw at UCLA or Seattle. So the people who are punished most are our fans."

West Virginia, another two-seed, faced a similar fate, shuttled off to the Louisville Region where they face the prospect of confronting a hosting LSU squad defending home turf in Baton Rouge. But the Mountaineers never had that strong a case for a top seed: Despite their solid 29-4 record, and a tie with Baylor for the Big 12's regular-season championship, their RPI (12) and SOS (56) fall well below those of the leading contenders.

The other two-seeds, Duke and Baylor, are both hosting opening-round sites.

Henry described the situation as an unfortunate corollary to the procedures and principles that govern the process:

"One of the things that we do as a priority is to try to protect the number one seeds. In doing that, as you correctly noted, we have Duke and Baylor both hosting first- and second-round games. In order to protect our one-seeds in the first and second-round games so they don't have to play on a lower-seeded team's court, we placed South Carolina in Seattle and Notre Dame in Toledo. In doing that, in protecting those one-seeds, we could not protect the number-two seeds. ... That left Stanford and West Virginia having to play on a lower-seeded team's court. Unfortunately, the way that our procedures are, and our principles, we cannot avoid that."

Of course, for many of the top seeds, the situation will be reversed come the regional rounds if they manage to make it that far: South Carolina, already a continent away from home, would be on course to meet Stanford on its home court; UConn could face third-seeded Nebraska in Lincoln; Tennessee might draw Louisville, also a three-seed, in the regionals; leaving Notre Dame as the only "safe" one-seed, when it returns home to host the regionals.

At least that part of the problem could be resolved by next year when the NCAA Tournament returns to neutral sites, at least for the regionals. There's also a plan afoot to have all of the opening rounds hosted by the highest seed, which should eliminate the prospect of a higher-seed having to play on a lower-seed's floor at any stage of the tournament.

Of course, even though this year's greatest friction surrounded seeding and placement, as in every season, there were defenders of the teams that got left out. The two most often mentioned as having been snubbed were Minnesota, who finished its Big Ten season with an overall record of 20-12 (9-9 in conference play) and South Florida, also left out with a record of 19-12, 14-6.

The question whenever a team is said to have been wrongly left out of the tournament is who got in who should not have. Several teams -- such as Florida (19-12), LSU (19-12), and Oklahoma (18-14) and Vanderbilt (18-12) received at-large bids despite records that were as bad or worse than the Gophers and the Bulls, leading some to contend that the SEC and the Big 12 were shown favoritism over better teams from smaller leagues.

It is true, that the SEC, the ACC and the Big 12 got a lot of love from the Committee this year, with the SEC and the ACC receiving eight bids each, while the Big 12 captured six. Making matters worse, some of those big-league teams got bids despite have finished their own conference seasons in the red, e.g., Vanderbilt (7-10, SEC), Georgia (7-9 SEC), LSU (7-9 SEC), see also, e.g.,  Florida (9-9 SEC), Oklahoma (9-9).

Still, of those teams, the only ones whose RPIs were lower than Minnesota's (RPI 47) were Oklahoma (RPI 59) and Florida (RPI 65),  and South Florida's RPI of 60 was worse than any of the at-large teams other than Florida's. Henry was able to point to objective criteria that warranted the selection of other teams in preference to South Florida, which she described as a "challenge."

"One of the things that we noted about South Florida," said Henry, "is they had no wins in the top 50, then they one win above 60 in the RPI. They were 0‑8 in the top 50.  Those were pretty significant factors when you get down do the nitty‑gritty determining who is going to be in, not going to be in."

As for Minnesota, despite its higher RPI, "When we were looking at the last four in, one of the things we noted was that the last four in had multiple wins over top-60 teams. While there were some losses in those columns, most of them had more top-60wins than those teams that were left out. When looking at Minnesota, they lost all of their top-25 games. They lost three games in the 51 to 100. They also lost a game above 100 at 109."

At the end of the day, said Henry, this year's proces was "as difficult, if not more difficult, than prior years." That was not because of the late-season and tournament conference change-ups, where,  for example, both South Caroilna and Staford were early outs in their tourneys. Instead, she said, "What makes it difficult -- this is actually a great problem to have -- our teams are getting better and better.  It''s becoming more and more difficult to diffentiate between the teams, who should be in, who should be out. If we continue to have that problem, that will be that will be great because we'll continue to put basketball teams on the floor."

 


Related: