Between the Lines: Selection Edition
by Adam Wodon/Managing Editor
Could it be that we actually have nothing to complain about in the NCAA tournament selection and seeding process?
With the way the PairWise numbers broke, it made it very simple for the committee to place the teams in mechanical fashion without much controversy. They were very fortunate in this regard, for a change. It seems the criticisms we have this year amount to nothing more than minor quibbles, which won't stop us from making them.
On the other hand, there remains aspects of the selection criteria in need of change. According to the system in place, the committee did everything right. But the system is still in need of a tune-up.
Mainly, what we see in the second year of the 16-team tournament is the committee's adherence to a very strict and literal interpretation of the Pairwise Rankings. In other words, it went by the list and seeded the teams 1-16, according to the numbers. It then matched them, 1 vs. 16, 2 vs. 15, 3 vs. 14, 4. vs. 13, etc... down the line. With potential second-round matchups featuring 1 vs. 8, 2 vs. 7, etc... As a result, for example, 1-16-8-9 would be in one bracket ... 2-15-7-10 in the next ... 3-14-6-11 in the next ... and 4-13-5-12 in another.
The committee tried this last year, but it was much more tricky in the first round. As it broke down, there was a lot of need to switch things up to satisfy the two sacred rules: avoid first-round intra-conference matchups, and the host schools must be in that region. But the committee did adhere to the 1-8, 2-7, 3-6, 4-5 format for the second round.
Since it was the first year of the 16-team tournament, no one knew the committee would try so hard to do that. This year, we were more prepared. And it's exactly what we got, excepting one: Switching the 8-9 pairing with the 7-10 pairing. That is it. Everything else went by the book. And the only reason those pairings were changed is because UNH (9) had to be in Manchester.
The committee just replaced UNH/9 with the next No. 3 seed — Miami/10. But just doing that would've created an all-CCHA matchup in the West Region of Michigan/8 vs. Miami/10. So, this gave the committee a perfectly valid excuse, beyond the attendance considerations, of also switching Denver/7 with Michigan/8. This had the added benefit of not only putting Denver in Colorado Springs to help attendance, but also keeping the 7 vs. 10, 8 vs. 9 integrity intact. The only deviation could come in the second round, if it goes 1 vs. 7 and 2 vs. 8, but that's pretty minor as these things go, and when we start quibbling over something like that, we know we've reached the depths of Pairwise esoterica.
Some other thoughts:
* For all the conspiracy theorists out there, if they ever needed proof that conspiracies don't exist, it was blatantly evident this year. When the bonus point system was introduced, many fans claimed this was a fudge factor that the committee would use to get the teams they really wanted into the tournament, or keep other teams out. Since the committee didn't make the bonus values public, it certainly opened themselves up to this kind of thinking. But this year, they had a blatant opportunity to take advantage of such a fudge, and didn't use it. They went strictly by the numbers — even if not all the numbers are public. Without the bonus, Colorado College's RPI was better than Notre Dame's, and CC would've gotten in. Colorado College is the host school in the West. If ever there was a moment when the committee would really, really want a bubble team in the tournament, it was now. But, no. The committee applied its bonus criteria, and, as a result, Notre Dame surpassed CC in the RPI. Notre Dame was in the tournament. ... We still don't know the exact number, but we do have further evidence that the committee doesn't scheme wildly to favor one team or another. Not that we needed that evidence.
* I still don't like the bonus points, and for all the same reasons I didn't like them last year. The numbers are arbitrary, the cutoff point for being a quality opponent (Top 15 of RPI) is arbitrary, and all you're doing is slapping on a number with no basis in any mathematical principle onto another number (RPI) also with no basis in any mathematical principle. Further, why should a team get rewarded for defeating tough opponents, while not getting penalized for losing to really bad ones? Of course, if you reward and penalize, then you're right back where you started and you didn't need any bonus/penalty system. Fine by us. ... But if Notre Dame had all these great wins, then don't they also have a number of bad losses? I mean, if they had great wins and no bad losses, then they'd be a lot higher on the PWR and the whole thing would be moot. Why is it better to defeat Boston College and lose to Lake Superior than the other way around? ... Further, it's been said the bonus is good because it encourages teams to play tough opponents. The problem is, this idea is at odds with the committee's other aims, which is to strengthen college hockey as a whole. Automatic bids are there in order to help smaller conferences. The committee's desire is to encourage more balance in the brackets across conferences. But if teams all start trying to schedule the strongest opponents, who will the Atlantic Hockey teams play? Another reason the bonus is not helpful.
* There was a thought Michigan could have been sent to Albany to help attendance reasons. But the pull of keeping the 8-9 matchup and the 6-11 matchup (Ohio State-Wisconsin) together was apparently too strong for the committee. And, since Maine and Wisconsin are in Albany anyway, it seems as if the attendance should be helped enough with those two teams. But this, and, even moreso, the fact that the committee didn't keep Michigan in Grand Rapids, shows yet again that bracket integrity is fast becoming the most important issue. Which is good, in general. Though we'd still like to see the committee afford itself the opportunity to be flexible. The Pairwise is not precise enough to rely on it too religiously, particularly in seeding.
* The committee could have tried to avoid the Minnesota vs. Minnesota-Duluth potential second-round matchup. This would be the 4 vs. 5 game. Last year, the committee had a similar situation, with BU and New Hampshire slated for the second round, even though the teams just played in the Hockey East tournament finals. But last year, the committee began to show that it believed preserving these pristine matchups was more important, even though they had easy ways to switch it around. This year, avoiding Gophers vs. Bulldogs in the second round would have been a bit more tricky, so it is probably just as well to leave it that way. But the philosophy seems to be, at this point, that the committee just doesn't care about avoiding such second-round matchups. In the past, coaches wanted them to because they didn't think it was right for their teams to have to try to defeat a team again that it just defeated in its conference playoffs.
* There are some serendipitous matchups in the first round. Though the intriguing storylines are nothing more than a fortunate happenstance for the committee, we'll take it. There's actually a lot to be said for purposefully creating these kinds of things. Perhaps the committee should consider it. They certainly do it in the basketball tournament, that's for sure. The sense of intrigue and newsmaking on the basketball committee is finely tuned. Not so much in hockey. Nevertheless, we did get this: Miami, coached by Enrico Blasi, vs. Denver, coached by George Gwozdecky. Blasi played for Gwozdecky at Miami, and then coached with him at Denver until getting the job back at Miami when Mark Mazzoleni left. ... Notre Dame, coached by Dave Poulin, vs. Minnesota, coached by Don Lucia. Poulin and Lucia were teammates at Notre Dame. ... Harvard vs. Maine. This one has made the Murphy family both happy and torn at the same time. Harvard sophomore Dan Murphy will be pitted against older brother Ben Murphy, the Maine junior forward who scored the game-winning goal in triple overtime against UMass at the Hockey East Championship. Their parents will now be able to see both kids play at the same time, but who to root for, that's the hard part.
* I'd still like to see some sort of "down the stretch" criteria. In the past, this was "record in last 20 games." Then it was 16 games. Then it was eliminated altogether because of strength of schedule issues. The question is, does the committee not want a "down the stretch" criteria on philosophical reasons — i.e. it doesn't think games at the end of the season should count more than games at the beginning — or does it not want it because of the strength of schedule inequity? If it's the latter, then that can be solved through advanced methods of correcting for strength of schedule, though it would be trickier to convince the committee of adopting that. If it's because of philosophical reasons, then that's fine, but I'd still like to see something in there that accounts for play in the conference tournament. Maybe not for selection purposes, but at least for seeding. For example, Boston College could've slipped below Maine and Minnesota in the seedings based on its 1-5-0 finish this year.
* New Hampshire is a quasi-home team this weekend, but it's different than it often is for, say, Michigan playing in Ann Arbor. The arena in Manchester is NHL-sized, not the Olympic-sized arena that the Wildcats are used to playing in at home in Durham. That, plus the proximity of so many other D-I schools nearby, whose fans will come to the games, doesn't make the advantage quite as dramatic as it does in other cases.
* Harvard just played, and won a championship, in Albany, now gets to return to face Maine. Does this matter? It can't hurt.
On Another Note
Getting to work with ESPN's John Buccigross this weekend was fun — especially because it's always good to work with a fellow crusader for increasing the size of the nets — 4x6 inches, two square feet, introduced gradually over a period of years. He'll continue to advocate on ESPN.com to 10 million people, and I'll talk to 10,000 here, and maybe together, we'll make a dent.
In response to my last column on this topic, a reader sent his thoughts in return: "I always thought only a complete MORON! would say the nets should be bigger. Since I know you're not a complete moron, I can't understand it. Increasing the size of the nets is too drastic a change."
I think there was some faint praise in there somewhere, but the thing is, NOT increasing the size of the nets is the drastic change. Yes, that's right. Say you have a 4x6-foot net. Standing in front of it is a 3x5-foot pylon. The object of this game is to shoot a vulcanized rubber disc into the 4x6-foot net. Now, one day, someone decides to stick a 3 1/2-foot by 5 1/2 foot pylon in front of the net. Don't you think whomever made this decision just drastically changed the nature of the game? Don't you think it will be a lot harder to shoot this rubber disc past this bigger pylon? What if you were told this bigger pylon is now also a more athletic human being that moves around?
Fact is, the goalie is bigger, and there is no offsetting factor. If the defense gets bigger, the offensive players can get bigger. In basketball, if the average human gets taller, the average defender and offender will both be taller. Scoring isn't changed because of this. Could you imagine increasing the size of the ball while the hoop stays the same? In hockey, the goalie gets bigger, then obviously the nature of the game has changed dramatically. There is no offsetting factor. The net is the same size, the puck is the same size.
Increasing the size of the net is the only way to keep things THE SAME.
Now, this gentleman went on to say that scoring is not the problem, it's scoring chances. On this score, I basically agree. There are plenty of other things that need fixing in order to increase the scoring opportunities. Even though goalies were getting bigger and better over the last 20 years, scoring chances were still pretty good, and games were still pretty exciting. It's only over the last five years or so that even chances have been dramatically reduced. But that doesn't mean we should ignore the size of the net as a factor. With an increased size of the net, perhaps you'll see more shots taken from more places, shots which are futile nowadays.