In review.
I could see Harvard was in based on their quality wins. Then they seemed to really discount Duke for the bad losses. So one could see OSU getting in for having a win over Harvard. OSU also had a 14-11 head-to-head win over Notre Dame.
The choice of UVa over ND is a little hard to fathom. They apparently discounted the Duke wins so much, that the went to the UVa-ND head-to-head. Wow. Of course, Duke beat UVa.
In the past, the committees have talked about playing a tough schedule. Notre Dame appeared to try to do this by adding Georgetown and Maryland, while UVa only added Maryland.
It all comes down to bubble composition. I spent basically an hour + making on a show laying out various things to eventually arrive at this point.
This talk from the committee of “it was 6 teams for 4 spots” - that’s not the ultimate bubble. The bubble isn’t half the at-larges, I’m sorry. At some point you have to zoom in and make difficult decisions for one or two spots.
So… from a data analysis perspective… not a what I ultimately thought they were going to do perspective… I thought the final bubble was OSU, Harvard, and UND for 2 spots. I saw Duke as second team out but being clearly off the back of this group, not close enough to be in the final discussion.
OSU has the head to head over both of the others and are an easy choice to remove (in) from this discussion. Then it’s Harvard/UND in a close call for the last spot. I thought Harvard had the slightly better case in this comparison, but the data was there to support either and I could live with it.
I do not think that UVA was seriously ever on the ultimate bubble. I say this mostly only based on how the committee had previously evaluated UVA (and Rutgers). They were both being consistently evaluated more positively than the actual selection criteria data suggested they should be. Now had UVA been in this zoomed in bubble, maybe they lose the comparison to whom they are being compared for this final spot - but I don’t think they were ever in this position. If anything, my main problem with how UVA was evaluated in the final bracket is the gulf between where they and Rutgers were placed. I do not think their evaluation of these two were internally consistent. Either over-evaluate both (6/7) or neither (8/9), and I'm fine with it. However, my theory here is that they split them by the Ivy League pair primarily to "make the bracket work." I had been on the record for quite some time saying that 8/9 almost certainly had to be Ivy vs. ACC to make the bracket work in a sensible way with a representative matchup in that game/QF it fed into. And with UVA as the only ACC team in the field, well...
What ultimately screwed up 2 of the 3 bubble calls (mine and IL) is that we were treating Duke as a bubble team because we expected that was how the committee was going to choose to treat them. With them, it’s 4 teams for 2 spots and there was only one way to present the order within this final bubble that made any sense (ND>Duke // OSU>Harvard) to ultimately put both ACC bubble teams in the field - which is what all 3 calls felt they were going to do. (Stevens is a... different... case of logic that I still do not know 100% what to make of. )
I was screaming at the top of my lungs that I thought this was wrong. I did not think that the data when you really sat down and looked at it in conjunction with their stated criteria supported Duke as a team on the ultimate bubble. I am not a professional journalist, and I have more freedom and latitude to say what I want without worrying about the consequences. IL was more pragmatic “it’s close” - and that’s totally fair. Stevens… oh where do I start.
How out of touch is this? "Thinking it's a safe bet there will be more consternation about the seeding rather than the selection of the 18-team tournament tomorrow night." He STILL had Duke seeded. His ultimate bubble call without Duke being in the mix there to muddle things of UND (in) // OSU&Havard (out) just blows my mind. Looking at those 3 teams as an ultimate comparison for 1 spot and coming up with anything other than OSU as the one in... I do not see how you do it unless you think there is NO ULTIMATE BUBBLE. He did not think it was even close for the last spot, I guess? This call of 3 ACC teams in was presented as the *only* call - there was no room for debate.
So... a lot of words to say, it comes down to the composition of teams for the last spot or two as to what the call is. Some teams do well in close comparisons with certain teams, but not others. For example, OSU has a head-to-head win over UND and Harvard that helps them in that comparison. But OSU vs. Duke is less conclusive if you think that's the comparison for the final spot (as my projection/IL's projection thought the committee would eventually arrive and and it seems ESPN was trying to present it as).
Woodruff’s words to me suggest that Duke was removed from consideration before they arrived at at their ultimate bubble, which I think went as I presented above of 3 teams for 2 spots. Then remove OSU (in) and it’s Harvard/UND for the last spot.
But this is only my opinion of how it all went down from a) my having assessed these teams seriously for over a month now and b) reading and interpreting the quotes that we were given from the committee. There is a chance that I am totally wrong and just as out of touch with the committee's line of thinking as I think Stevens was with his call.
