2017 NCAA Hockey Tournament Bracket Analysis

There were a lot of different ways the NCAA tournament committee could have gone in assembling this year’s tournament field. Late last night, when all the people who do think hard about this stuff were putting together their best guess, we all came up with wildly different opinions on what the field would look like, and I had a sneaking suspicion we all would be wrong.

It turns out, everybody did have it wrong. I’m not sure I saw anybody with the exact bracket predicted last night. I probably came the closest(#humblebrag), with the only change being that the committee flipped Air Force and Cornell, a change that kind of came out of left field.

Regardless, the fun of the exercise is all about looking at the critical decisions the tournament committee had to make in creating a field and which direction they chose to go.

The first big decision was what to do with Denver: either put them in Fargo or in Cincinnati. I opted to put them in Cincinnati in my Bracketology and ultimately the committee agreed.

Their stated reasoning was that they matched up teams by ranking before looking at regionals, and with the 7-10 match-up locked into Fargo, it made sense to put the second overall seed there. I’m not sure that logic would hold up every year, but it certainly made the most sense this year.

The more thought I put into it, the more I agree with the decision. There is a natural inclination when putting these brackets together to start by putting the first overall seed in their closest regional. I think that is the right idea, but a little off. The idea should be to put the #1 overall seed in the best regional for them. Nine times out of ten, that is going to be the regional with the least amount of travel for a team. But in Denver’s case, they were flying either way, and weren’t going to have any sort of fan presence anywhere. I’m not sure the fact that Fargo is a home game for North Dakota should count(after all, the NCAA insists it’s not a home game if a team doesn’t play there regularly, and that their student-athletes “enjoy the atmosphere”). But technically, being placed in a region with 8-9 is fairer to them than being with 7-10.

After the Denver thing, most of the discussion is about decisions the committee didn’t make.

A few people discussed switching #16 Michigan Tech with #15 Ohio State to put OSU in Cincinnati. To me that’s a non-starter when #15 is an at-large and #16 is an autobid ranked 27th nationally. Even though I do think Tech is lower than they should be due to some flaws in college hockey’s scheduling and the way RPI works, and that they are probably pretty close to Ohio State in terms of ability, the numbers are the numbers and it isn’t fair to the #1 overall to make that switch.

Same goes for a bunch of other attendance swaps that ended up not being made. On one hand, the theory is that all these teams are so close that making a bunch of swaps within seed bands isn’t a big deal, and to an extent that is true. Every year we see how much one result can completely change the seeding order. Especially in a single-elimination tournament, there isn’t much discernible difference between teams. At the same time, you have to draw a line somewhere, and I’m fine with drawing it the day after the season ends.

A couple projected brackets I saw had people making a couple flips to get #6 Western Michigan to Cincinnati playing #9 Penn State, which resulted in #8 Union playing #12 Air Force. To me, that was too big of a leap in bracket integrity when it was fully possible to have something much purer. As a reminder, you want the overall seeds of the two teams meeting in the first round to add up to 17; anything lower is a strong bracket, anything higher is weak. Creating one match-up that adds up 15, and another that is 20, that’s a huge difference. I know Western Michigan would have greatly preferred to play in Cincinnati, and likely wouldn’t have complained about the theoretically tougher match-up, but like the Pairwise itself, the only to do it is to go strictly by the numbers.

Some people had Providence moving to the New Hampshire regional because it wasn’t fair to put Providence in Providence(again). Betting on the NCAA to do what they “should” do over what is profitable(sorry, provides “atmosphere”), is a great way to go broke. If the committee put Providence in Providence two years ago when they had every reason not to, I couldn’t see them passing on the opportunity this year when they had tons of plausible deniability in terms of it making sense from a bracket integrity perspective.

The one change the committee did make was flipping #11 Cornell and #12 Air Force, which, okay. Air Force is always kind of a wildcard because they can go anywhere(One of my many favorite Frank Serratore quotes: “We’re not called Bus Force”). I guess Cornell brings a little more attendance money(oops, “atmosphere” again) to New Hampshire and it avoids a potential rematch against conference mate Harvard. But it seems like a weird switch.

Who knows, maybe they just flipped it to make sure nobody had a correct bracket, and to keep us all guessing, speculating, and discussing for another year.