There are some interesting puzzles posed by the results so far in March Madness. Before I get to them, here are a few preliminaries. First, rankings are of two varieties - ordinal (the seedings provide ordinal rankings of teams) and cardinal (the Sagarin ratings claim to provide cardinal rankings of teams). Ordinal rankings can be used to determine which team is the favorite and which team is the underdog. But ordinal rankings are insufficient to determine the odds in a head-to-head match-up. Cardinal rankings can do that, so can suggest whether the game is close to a you-pick-em, with the odds near 50-50, or if one of the teams is an overwhelming favorite. If you think of seeding versus team quality as plotted to form a "quality hill," then the ordinal rankings themselves can't distinguish from a gradually rising hill, more like a plateau, from a steeply sloped hill. The gradually rising hill will likely produce more "upsets" since the teams are pretty evenly matched. The steeply sloped version should produce more lopsided wins by the favorites. I think it's not a bad inference to say that this year there was no super team, but rather a bunch of good teams that were competitive. The gradually rising hill metaphor seems to hold in 2023.
These rankings depend on the history of play during the season. Pretty much everyone agrees with that, though I will raise a criticism about that experience in a bit. Here are a few issues that have been with us for some time. Should the experience simply measure wins versus losses (and the quality of the opponent played) or should the entire scoring history during the game matter. Take a look at this, from the Illinois at Purdue game, the last game of the regular season.
ESPN provides a game flow diagram for each game. Purdue had the lead for much of the game. Illinois developed a reputation during the season as a come-from-behind team. The score was actually tied with 1:17 left in the game. Purdue won out, but the game was close at the end. Purdue became a number 1 seed in the Tournament. Illinois was a 9 seed.
One game does not a season maketh or, statistically speaking, be wary of inferences from small samples. Purdue did win its prior game against Wisconsin, but it was a nail-biter. Before that Purdue lost to Indiana and a couple of games before that there was a bad loss to Maryland. Does that look like the play of a number 1 seed? Purdue did win the Big Ten Tournament. Winning the regular season and the tournament likely were enough to give Purdue that number 1 seed.
But there might be something else that explains things, which is how Purdue played in December. Purdue went into the season unrated, outside of the top 25 in both the AP and UPI polls. By week 6 they were rated number 1, after impressive wins over Gonzaga and Duke. Plus, Zach Edey seemed the dominant player in college basketball at the time. That impression endured, even if Purdue's late season performance didn't quite support it. So, this brings about another question. Should games in December count as much as games in late February and early March? One response might be that it depends on who the opponent is, as those December games are played out of conference.
It is my understanding that the conference itself sets the schedule for conference games, with perhaps some input from the individual schools about holidays, final exams, and other possible information that might make the date a no go. But the individual schools put together their non-conference schedules. Of course, money matters here. It matters a lot. I'm not current about revenue sharing agreements within conferences, but I assume that schools get to keep a greater share of the revenues from the non-conference games they play. In the old days, before there were multiple cable TV channels that would carry college basketball games as part of their normal fare, there were perhaps three or four national games on the weekend, aired on the major networks, or there was local TV only. Given that, name brand teams had incentive to schedule a lot of home games against cream puffs, teams that wouldn't demand a home and home playing agreement, would take a modest amount of the gate revenue, because they would welcome the exposure given by playing a big school. Now there is incentive to play other name brand teams, quite possibly at neutral sites, to get the better TV coverage. In addition, the NCAA has strength of schedule as one of the factors that matter in getting into the tournament and in seeding the teams that do get in.
This plays against the mid-major schools that have good teams. I find it amazing that even today Sagarin has Illinois ranked ahead of FAU (though just barely) while Illinois was one-and-done and FAU is in the Final Four! This seems like a measurement error by Sagarin. It can be "explained" by looking at FAU's rather low strength of schedule rating. Which power team would schedule them in the preseason? While they might be a fan favorite now as the underdog that made it to the Final Four, nobody knew that would happen back in December. They were simply a no-name at that point. So they scheduled the games they could get. Ditto for Fairleigh Dickinson (which upset Purdue and then lost in a reasonably close game to FAU).
If these schools show consistent excellence in men's basketball in the upcoming years, they will develop a reputation and be able to get around this issue of pre-conference scheduling. Gonzaga has shown that can be done. But the more typical pattern is for the coach to leave for greener pastures and then the team will revert to its past mediocrity. Thus, when there is a very big fish school playing in a small pond conference for the first time, it is difficult to know just how big the fish is. This particular NCAA Tournament has made the issue acutely evident. I'm not sure there is a good solution to it. But I want to make one other point that might matter in getting it resolved.
Being a college basketball player for a team that has a chance to get to the Tournament is a full-time commitment and for many college players it may be seen as a path to their career - playing pro ball, or coaching at some level, or becoming a commentator or announcer. There will be some players who see their career path as elsewhere, say in business, medicine, or teaching. Then, part of the choice of which school to attend will depend on how they perceive it to prepare them for that path. But for those in the first category, having the chance to play immediately and showcase what they can do is likely much more important to them than to chose a school based on academics. Further, now with the Transfer Portal, if they do well playing-wise early on they can then transfer to a big-time school after that, assuming the latter will give them more exposure and more chances to play against better competition, giving them other options than declaring for the NBA draft immediately.
Further, talented younger coaches, whom players might relate to better than their more senior counterparts, are likely to start their careers at mid-major schools. That's been with us right along. I'm just including it here now to make the case that another FAU in the near future is more likely than it was in the past. If I'm right about that, the NCAA might want to think through how to address the big fish in a small pond issue. The power conferences clearly would prefer that it doesn't happen, for it seems to question their legitimacy. I don't see an answer here, but I do see that a lot of hair pulling will be likely.
Now let me turn to a different issue. It's always been true that a team in December is not the same as the team in March. The hope is that the players grow, individually and as teammates, so teams are better in the spring than they were the previous fall. Of course, injuries can make that go in the other direction. And so can emotional/psychological issues within the team that the fans aren't usually aware of. But let's face it. Many if not most college students have been having such emotional issues, as a consequence of life under Covid, the nature of college education today, and student loan debt exerting pressure of various sorts. I'm ignorant about how NIL is playing out in reality, but I can guess that it might cause a lot of jealousy and discontent within a team, as will the coach's decisions on who plays and when, which indirectly impacts this money flow. Those money issues were not on the table before. (Other money issues were. I'm not trying to argue that NIL is bad, only that coaches likely haven't fully anticipated its consequences.) Further, the huge amount of money that is in what the head coach gets paid, and what the university makes from its NCAA sports activities, might really bum out the players if they feel they aren't getting a fair shake. Without going further down this path, I'm just saying there are reasons why now is the winter of their discontent.
For assessing how a team will do in the tournament, you need an accurate assessment of how the team is doing now. This is an argument that recent games matter more, as they should be incorporated in that assessment in a different way than early games. But if the early schedule matters not as much, then the NCAA is put in a bind of incentivizing schools to schedule high quality opponents in the preseason. Why are they doing this? Why don't they instead try to encourage power conference schools to play mid-major schools? But it also might be an argument that the division between preseason and conference games gets changed, with some out-of-conference games scheduled in February and early March. In the Big Ten this past season, where Purdue had a lock on the regular season championship, that might have actually increased fan interest. In a different conference, where the conference champion wasn't determined till the last week of the season, such games might seem only a distraction.
* * * * *
Now I want to turn my attention to the conference tournaments. I can remember back to when the Big Ten actually had ten teams, and the conference schedule for Big Ten Men's Basketball had a home and home round-robin where each team played every other team twice, once at home and once away. Bob Knight famously argued against the tournament, as this regular season schedule offered the fairest way to select the conference champion. Of course, since then the Big Ten has increased its membership, a round-robin with a balanced schedule is no longer feasible. A tournament might be fairer now, though we still announce the regular season champion as well. But that is not the reason why it was argued that the Big Ten needs a tournament. Rather, it was said that a conference tournament would better prepare schools for the NCAA Tournament, increasing the chances that Big Ten schools would perform better in the latter. That's the issue I want to take on here.
I don't want to deny that the effect might have once existed. But conferences, at least some of the major ones, have since grown larger. Further, over-scheduling has since become the norm. The Illinois Coach, Brad Underwood, has complained about too intensive a schedule on multiple occasions, and not just this year. It's from listening to him that I'm getting the hypothesis. If, as I said earlier, teams can get a lot of attention now on TV by playing other teams with whom they are competitive, then why not schedule those games? But the players on those teams are young men who are not yet professional athletes and who don't have any say in the frequency at which games are played.
Regarding how this over scheduling during the regular season impacts play in the NCAA Tournament, it is worth noting that Michigan State was the only Big Ten team this year to make it to the Sweet Sixteen. They actually played one game less during the regular season than other teams, because their opponent had too few players due to Covid and they couldn't get the game rescheduled in time. They also lost their first game in the Big Ten Tournament. Contrast how they did to Purdue's performance.
Similarly, in the Big East, UConn lost early in their conference tournament, but has made it to the Final Four. Similarly Creighton lost in a nail biter to to UCSD in the Elite Eight, but lost by a big margin to Xavier in a semifinal of Big East Tournament, and that happened the Friday before so gave them ample time to recover for the NCAA Tournament.
So, here's the message. With the increase in conference size among the power conferences and the over scheduling in general that has resulted from the profitability of games on TV, good teams are exhausted at the end of the season. Those that know they will make the NCAA Tournament have incentive to sandbag the conference tournament, giving them more time to recover. Just to give one other example, Miami, which lost in the semifinals of the ACC Tournament, is in the Final Four. Miami lost to Duke, the ultimate winner of the ACC Tournament. In turn, Duke lost fairly handily to Tennessee in the second round of the NCAA Tournament. So, what purpose do the conference tournaments serve nowadays, other than to get more revenue from that fans? And if this analysis is right, this purpose can no longer be served, once the fans come to realize what is going on.
In a conference where there will only be one member that makes it to the NCAA Tournament, and that member is determined by the conference tournament, the conference tournament will retain meaning. But in the other case where schools will lose their conference tournaments but will get placed into the NCAA Tournament due to their excellence in play during the rest of the season, they will make their preparations for the NCAA Tournament so as to ignore, as much as possible, their conference tournament. This seems evident. Is there anything to do about it?
While it's easy to imagine that Brad Painter would complain about his team's fatigue before it played FDU, would Jim Larrañaga do likewise, now that Miami is going to the Final Four? If not, this might seem more like sour grapes than a real issue. Until you hear the post mortems from those who have benefited from the current system, I don't think this is a problem that can be solved. But then again, those beneficiaries have little incentive to raise these issues.
A feasible change that might help, especially once a true round-robin regular season can no longer happen because there are too many teams in the conference, would be to shorten the regular season by a couple of games and then hold the conference tournament over two weeks, with winning teams playing either Thursday-Saturday or Friday-Sunday, at least during the first week. This would lessen the role of the conference tournament as its own endurance contest and more closely emulate how it is done in the NCAA Tournament. Would such a change eliminate the sandbagging by teams that are evidently already being selected for the NCAA Tournament? I don't know, but I think it's worth knocking that idea around some.
Watching the games this past weekend, it seemed to me that at least a couple were decided by injuries incurred during the game to key guards that were a consequence of some violent contact with an opposing player. We treat that as if it's part of the game, and so it is. The coach can't do much about it then, other than hope that the player can gut it out during the rest of the game. But when injuries to important players are sustained in practice and/or in earlier games, the coach will look to lengthen the downtime, if at all possible, to better improve the chances of recovery. Fatigue and injury are not the same, especially if the former is considered purely from a physical standpoint. If, however, you take into account burnout and mental health, the two might start to look quite similar. Care about the players along these lines might motivate my suggested change.
On the other hand, teams tend not to disclose non-obvious injuries to players, and having a mental health professional as part of the support group for the team might be a no-no. These are matters outside my scope of knowledge.
So, what might be done, purely from a fan standpoint, is to track performance in the conference tournament versus performance in the NCAA Tournament among power conference teams, and do so for the last 5 or 10 years. This year could be an outlier that way. I'm thinking it's the new normal, but who knows?