After an eventful couple of days in college hoops, my Bracketology is now updated.
I found the debate on Twitter earlier in this week over the proper method to use when selecting the 36-team at-large field rather interesting. Essentially, the participants in the debate were those in favor of utilizing an analytic measuring the quality of resume versus those who argue resume analytics are but a tool in the larger “whole body of work” one is to judge when choosing the field.
Ken Pomeroy wrote an interesting article in The Athletic discussing this very subject.
Here’s where I stand on the matter. I believe that the Committee’s mandate is to select the best 36 at-large teams. We can quantify the best teams using Wins Above Bubble (or Parcells if we go all the way). The best teams are those that have performed the best relative to their schedule – and relative to what an average bubble team would be expected to do in the same scenario. In a perfectly rational world (without humans to muddy the waters), reliance on WAB/Parcells as the sole metric for selecting the field just makes sense. It’s the literal definition of the Committee’s mandate. As others have pointed out, predictive metrics can be better utilized for seeding purposes – they’re not designed to measure how good a team has been but rather how good they’re likely to be going forward.
Now, there are a few elements at play that make this likely to never be the way the Committee selects the field. The first and most important factor as to why a singular resume-quantifying metric is unlikely to dictate field selection is this: why have a Committee then? Why have a group of “experts” and “soothsayers” congregate to watch basketball, debate the finer points of Quad 1 and 2 wins, and use their combined wisdom to pick, with a precision only this chosen few can muster, the most perfect field of 36 at-large college basketball teams conceivable? There would be no need. And there would be no Committee.
Second, a bunch of talking heads would have much less to talk about on the major sports networks if we replaced the penumbra of gray that is the method by which the Committee executes its mandate at present with a crystal-clear black and white metric. With WAB, you’re either in or you’re not. With the current system, you might be in, you might not be in, and we can talk for hours about which side is more right. Who has more Quad 1 road wins? Who has fewer Quad 4 losses on Tuesdays in Nevada? Well, resume metrics already factor those in. It’s right there in WAB (maybe not the Tuesdays in Nevada part – but I’m not fully versed in the math).
The debate might rage on because frankly most people are skeptical of math and analytics and are motivated by their inherent biases. “I know it when I see it” is fine for a Supreme Court Justice determining what is or is not obscene, but it really shouldn’t be the way to determine the best college basketball teams in a given season at a point in time. That part is pretty easily quantifiable. Yet, there will always be those who deem the “eye test” and their subjective opinions as to “the whole body of work” to be paramount to stats. I don’t see that changing any time soon.
Here’s the upside: that means I can still have this blog and I can still be a bracketologist! Really, bracketology isn’t the selection of the 36 best teams. I don’t believe I engage in that endeavor, and I’d hope most bracketologists aren’t trying to make their own judgments as to the 36 best at-large teams. The distinction of most importance is that bracketology is the application of all known information as to the Committee’s selection process to determine to the best extent one can what the output of that process will be. I’m not picking the best teams – I don’t care about that! (Of course I do, but not in this exercise).
My exercise in bracketology is to see if I can see what the Committee sees. See? And until such time as the Committee implodes itself for the love of the stats, I’ll keep on bracketologizing.
Bracket and Bubble Notes:
Finally, to the brackets we go. No change at the top. Baylor (1 seed, #1) and Kansas (1 seed, #2) will do battle in Waco on Saturday, Gonzaga (1 seed, #3) hosts the Dons of San Fran (NR) tonight and travels to BYU (7 seed, #27) on Saturday for a tough test, and San Diego St. (1 seed, #4) hosts UNLV (NR) on Saturday.
Duke (2 seed, #8) got annihilated in Raleigh last night, and also lost a basketball game to (Welcome to the Field!) NC State (Last Four In Play In, #47). Not quite a seed line drop, but Duke falls a few spots on the seed list to the final 2 seed. Dayton (2 seed, #5), Maryland (2 seed, #6), and Louisville (2 seed, #7) remain on the 2 line.
Creighton (3 seed, #10) keeps on winning and is putting together an impressive season as we move it to the 3 line behind steady-Eddie Florida St. (3 seed, #9). Villanova (3 seed, #11) also moves up, while Seton Hall (3 seed, #12), by virtue of stemming a recent 2-game skid with a tip-in win over Butler (5 seed, #18), remains on the 3 seed line. Penn St. (4 seed, #14) falls back a couple spots and 1 seed line after losing to Illinois (8 seed, #31).
Auburn (6 seed, #22) plummets a whopping 3 seed lines after its recent run of luck ran out and it took back to back losses to non-tournament teams in Missouri (NR) and Georgia (NR).
Richmond (Last Four In Play In, #46), winners of 5 in a row, replaces Purdue (First Four Out, #70) in the field due to the latter’s 3 losses in a row, and only being 1 game over .500, and being 7-11 in Quad 1 and 2 opportunities (and also, no doubt due to my biases overcompensating towards WAB and the resume). Georgetown (First Four Out, #71) also falls out of this iteration’s field.
Dubz on Hoops 