March Madness is officially upon us! Today, I’m looking back at the field of 68 and seeing what to take away from the committee with respect to the teams who barely made and barely missed the cut. It was certainly an interesting year, and I’m overall not a fan with the job the committee did that oozed with inconsistencies. That’s enough complaining from me, let’s dive right in:
No Business Being Here
Quite frankly, I think the committee set an awful precedent with the inclusion of North Carolina and Xavier. I’m not saying that they can’t compete; in fact, one of these teams could easily make the Sweet 16. I’m simply stating that their body of work has historically not been good enough to warrant selection. For years, they have stated that you have to win games against good teams to get in the tournament, denying teams with predictive metrics in the 30s because they did not have the wins to support those figures. I believed that the committee wanted to put these teams in, but going 1-12 and 1-9 against Q1, respectively, was going to be a non-starter.
With the similarities between these two resumes, I was not surprised that Xavier got in once North Carolina was shockingly revealed on Selection Sunday. Contrary to consensus, however, I actually see more of an argument for the Tar Heels than the Musketeers; while both of these teams performed awfully in quad 1 and almost perfectly outside, UNC was perfect in Q2 while Xavier dropped a couple of quad 2 matches. Even with a Q3 defeat to Stanford, if “who you beat matters more than who beats you” rang true, this should be advantage Carolina despite Xavier’s home wins over UConn and Creighton. UNC won non-home games against UCLA, Dayton, and Wake Forest, all of which were at least in vague tournament conversations, while Xavier only has a win at Marquette to tout.
It’s evident that these resumes are neck-and-neck, similarly uninspiring but relatively clean. From the committee’s perspective, this should have been an appropriate spot to look at the metrics. Interestingly enough, though, North Carolina performs better than Xavier in EVERY single metric, whether it’s results-based, predictive, or NET. I didn’t see the argument for Xavier over North Carolina, which made it all the more intriguing when we saw the Musketeers as the third to last team while the ‘Heels were the final at-large selection. All in all, it seems that the committee valued Q2 victories and playing teams close quite highly. I would disagree with this method of evaluation, which makes it all the more interesting when you consider…
Opposites Attract
Texas also managed to crack the field. This isn’t crazy on its own, for I had Texas in my field, but not UNC and Xavier. I’m just shocked that the committee analyzed these two incredibly different profile archetypes and decided they were both good enough to warrant inclusion. I thought the Longhorns were screwed after they seemingly revealed their preference for the teams who beat teams they should beat, but here comes Texas, at 3-5 in quad 2, not only in, but the fourth to last team in the bracket. I have no problem with this personally, but such a dichotomy, especially considering what the inclusion of these three teams entailed, made getting the bubble teams correct almost impossible to predict.
And Now We’ve Hit Weird
So, we’ve determined that a team can get in with one Q1 victory to however many losses if they more or less clean house outside of the top quadrant. We’ve also determined that a team can be viewed slightly more favorably, even if they drop five games in Q2, if they can win a lot of their games against quality competition. What about a team who strikes the middle of these profiles? Sure, 4-13 is a far cry from 7-10 in Q1, but it’s significantly better than 1-9 or 1-12, and they have no losses outside of Q1. Not two. Not one. ZERO. At worst, surely this profile is only below the team with seven quad 1 victories, right?
Allow me to introduce you to the second team out of this tournament, the Indiana Hoosiers. Sure, two of those Q1 victories are at Penn State and Ohio State, but the Buckeyes were a first four out team. And there is no argument that Indiana’s top two wins (at Michigan State, vs Purdue) are less impressive than UNC’s and Xavier’s. No, actually, who is Carolina’s second best win? Wake in the ACC tourney? Dayton? By the committee’s ranking, IU’s win at OSU should clearly be viewed more favorably than those two as well.
Let’s analyze what they could use to dock the Hoosiers relative to the ‘Heels and Musketeers: their combined Q1/2 record, non-tournament team Q1 losses, and generally worse metrics. Technically, their 9-13 mark against the top two quads is worse than UNC (9-12) and Xavier (9-10), but Texas, at 10-15, managed to find their way above these two teams by way of their Q1 victories, while Indiana was below. They also lost at Northwestern, Penn State, and Nebraska, none of whom were in the tournament. While Carolina and Xavier only have two non-tournament Q1 defeats, the Tar Heels fell at home to Stanford while the Musketeers lost at TCU and Georgetown, meaning both of these teams lost at least as many games to teams outside the tournament without the quality wins to make up for it.
Lastly, you could make the argument that North Carolina outperformed IU in every metric aside from KPI, and Xavier ranked higher in all but KPI and WAB. I understand that the numbers play an ambiguously-sized role in tournament seeding, and perhaps they were uncomfortable putting the Hoosiers over these teams if only one or two of the seven favored them. However, if Xavier could be placed above UNC despite being outperformed in all seven metrics, this is not a viable argument for IU’s exclusion in the slightest. The inconsistency here is frankly egregious.
To wrap up, I don’t have a problem with Indiana missing. They were my last team in for a reason and would have been knocked out of my bracket had UAB (or even George Mason) won. It’s just so baffling that Texas would be the fourth to last team in while Indiana is the second team out, with UNC and Xavier sandwiched between them. With the committee valuing Texas’ solid Q1 record, the only other team with an argument for that last spot in my bracket was Boise State. With respectable 3-6 and 8-8 marks in Q1 and Q1/2, respectively, there’s more of a case for this team to be slotted directly behind the Longhorns. I expected the Mountain West stigma, a truly bad loss to Boston College, and mediocre overall metrics to thwart the Broncos’ chances despite six wins against tournament teams, including neutral site victories against Saint Mary’s and New Mexico. If the committee wanted to be consistent, this could have been the team that gets slotted right behind Texas if not Indiana.
Who Saw This Coming?
Wow. Just wow. I’m not entirely surprised that they considered the Tucker DeVries injury given that he was a big-name transfer and they got critical victories over Gonzaga and Arizona in Maui. I factored this into my projection and expected them to be in Dayton rather than the 10-line. Clearly, I did not factor it in nearly enough. With this bubble landscape, 6-10 in Q1 and 10-13 in Q1/2, should have easily been enough to crack the tournament with no additional losses. Even if you take out the games DeVries played, the Mountaineers are still 4-8 against the top quadrant and an identical 4-3 in Q2. In fairness, two of their Q1 victories were at Utah and Cincinnati, who were not terribly close to the field, but North Carolina and Xavier failed to win their similar contests (at Wake and Pitt for UNC, at Cincy and Villanova for Xavier). It also goes without saying that wins over Iowa State and at Kansas are better than the best pair of victories for either of the aforementioned teams who will be dancing.
I want to further analyze what the committee said with regards to WVU’s exclusion: obviously the DeVries injury, but also their 13-11 record following their injury. I’m first going to discuss that record. Yes, two games above .500 is not an at-large team, but this record is more or less a function of their performance in conference. It’s almost like the committee conveniently ignored that most power conference teams on the bubble are .500 in conference play (West Virginia went 10-10), if not worse.
For some reason, I haven’t heard anyone point out the double standard of factoring the Mountaineers’ injury, but not Josh Oijanwuna’s to fellow Big 12 foe, Baylor. Although he doesn’t have the recognition of being a top player on the team like Tucker DeVries, he was an integral part of their physicality and one of the team’s leading rebounders. In games without Oijanwuna, the Bears have gone 4-6 overall and 0-5 in quad 1. From a pure numbers standpoint, this is worse than what West Virginia has done without DeVries, making it appalling that Baylor cracked the 9-line while WVU was out altogether. And, while Baylor’s predictives are better across the board, we have already discussed that having team sheet metrics exclusively better than another team (West Virginia is only better in KPI) does not guarantee that said team will be placed above them. Only the Bears’ 7-1 Q2 record clearly stands out compared to WVU, with their Q3 home loss to TCU being a blemish that the Mountaineers don’t have, and a 5-12 Q1 mark is not as pretty as 6-10. Baylor doesn’t even have the argument of “better” Q1 wins either, for they scored victories at Utah and Arizona State, and they are a buzzer beater against St. John’s away from being 0-12 against the top half of Q1. Straight up, both of these profiles have pros and cons that the other team does not while having a similar injury situation. I would understand giving Baylor the edge due to their prettier KenPom and NET, but ignoring their injury while effectively placing them three seed lines above WVU is an outrageous disparity.
I was pretty confident that the bubble would come down to two spots to be claimed by Texas, Indiana, Boise State, North Carolina and Xavier, with UC Irvine and Ohio State having to rely on a favorable read that I didn’t see a path for. However, if there was to be a shocker for a team who missed the NCAA Tournament, I honestly would have said Baylor. Never in a million years would I have projected such a seeding discrepancy between these two Big 12 teams.
So, what actually happened?
After scrutinizing these different resumes, with Texas above Xavier and UNC, who were above West Virginia and Indiana, it’s nigh impossible to see a path where the committee ranks this group of teams in such an order. However, the three teams in the field perform significantly better in the NET and predictive metrics than the two teams that missed. Texas, Xavier, and UNC had respective NETs of 39, 45, and 36, respectively, while WVU and IU were ranked 51st and 54th, respectively. As for the predictive averages, North Carolina leads with 31.3, followed by Xavier at 39.7, Texas at 42, West Virginia at 46, and Indiana at 51.7. It seems like the committee let the advanced analytics determine who gets in around the cutline before looking at how frequently a team capitalized on its tournament-level opportunities to determine who was the fourth to least team in, rather than the last team in. It also seems like the only metric that truly punished teams for losing, even KPI, was almost ignored, while BPI was considered despite it maintaining a pre-season bias to this date.
Unfortunately, this does not answer every question. Ohio State has incredibly similar metric averages and resume composition to Texas. The reasonable explanation for their evaluative difference is that OSU’s victories aren’t nearly as good as Texas’, and, more importantly, 17-15 was viewed as a non-starter for an at-large team. I think this is fair. Boise State also had a noticeably bubbly resume, but nothing popped off the screen. With metrics hovering around 50 all around and losses to Washington State and Boston College, I can see how they were the 4th team out, even if I thought their Q1 and Q2 record would be rewarded more.
With predictive metrics setting the tone for these final at-large teams, there’s two teams left to discuss with very similar predictives (and results-based metrics) to Indiana. First, I am going to talk about Vanderbilt. Their averages of 49 and 41.7 (NET 48) seem extremely in line with IU’s 51.7 and 44.3 (NET 54), but the committee really seemed to value their five Q1 victories that were all against 6-seeds or better. That, understandably, puts them ahead of Indiana, while not being as bad against Q2 propelled them above Texas. It seems like the committee used genuine nuance when placing Vandy on the 10-line, and I think their position is nearly spot on. It’s a shame that the committee didn’t apply the same logic to the rest of the bubble.
It’s a miracle that I have gone this far without even mentioning the second to last team in the tournament, but with predictive and resume metrics averages of 51.3 and 43.3 (NET 52), San Diego State appears remarkably similar to the Hoosiers. They have three losses outside of Q1, granted two are to UNLV and the other is to Utah State, another tournament team. So, what got them in over Indiana? Even with the two Q2 losses and Q3 defeat, they were able to rack up critical non-conference wins away from home against Creighton and Houston. Yes, the same Houston Cougars that cruised through the Big 12 en route to a 17-1 conference record, a Big 12 championship, and a 1-seed in the tournament. Nobody on the bubble has a win like that, and IU’s victory at East Lansing is the only one that comes close. However, Indiana’s 4-13 Q1 is not only worse than SDSU’s 3-6, but they also lack the depth of tournament wins that the Aztecs possess. San Diego State’s Q2 victories are, quite frankly, higher quality than Indiana’s, for they have wins over New Mexico, UC San Diego, Colorado State, and Boise State, while the Hoosiers merely Have the Ohio State Q2 victory to compare while their other wins are a step back in quality. This seems to be the crux of SDSU’s inclusion and Indiana’s exclusion, and I felt this was a justifiable landing spot for the Aztecs
Last Words
After a thorough analysis, I can sort of wrap my head around the logic of how certain teams were placed around the bubble relative to each other. That does not, however, mean I agree with such logic, nor do I think this is necessarily predictable. What was relevant and irrelevant seems extremely cherry picked, inconsistent, and does not reward teams for winning. Even though I didn’t predict or necessarily agree with the Mountain West’s underseeding last year, at least the committee demonstrated a level of consistency with the decision, and one could easily point to their poor tournament history and overreliance on mid-tier conference opportunities to build their resumes. I hope the committee returns to some semblance of putting teams in based on who deserves it through winning. If not, why even play the games?
Comments
Post a Comment