By RICHARD CHEEKS
I don’t know if UK could have beaten North Carolina or not, but I believed they had been playing at a level that made that not just possible but likely. Then the wheels came off against Indiana, exposing all of the team’s early season deficiencies all at once. That is what usually happens in this tournament to frauds when a good coach gets a week to prepare a plan to expose those deficiencies. That is what happened to UK against IU. If IU had not done it, I now think UNC would have.
However, as we see almost every year, when the Elite 8 arrives, the real tournament begins. For the most part the pretenders, or as Coach Calipari has remarked this year, the frauds are gone. The men of this game remain standing. Today, tomorrow, and next weekend are the most important games of every season of college basketball. Seven games to crown a true champion. If the NCAA wanted to cut to the chase and set this up for real, I think 6 of these 8 teams would have been picked for the real tournament. For completeness, the other 2 Elite 8 contestants are #28 (10 seed) Syracuse and #33 (6 seed) Notre Dame. The tournament has 67 games. 60 of those games are played to boil down the field into the condensed high value Elite 8, whereas I would have selected Michigan State and Kentucky for my Elite 8 on Selection Sunday.
The Elite 8 matchup between Villanova and Kansas probably should be a Final 4 type of matchup, not a Regional Final, and Villanova would have gotten my 4th #1 seed instead of Oregon. But that was a close call, and it seemed at selection time Oregon was getting tons of love from the talking heads, which is the way the subjective influences the seeding process.
Nevertheless, the teams left in the running at this elite 8 include 6 of the very elite teams from this season along with Notre Dame and Syracuse. For the majority of years, the Pomeroy #1 team wins this thing, and the #2 team in next in line with likelihood of winning. There is very little difference this year between #1 and #6. Finally, the Champion usually has offensive and defensive efficiency levels in the top 20, and those teams still standing from the top 6 all satisfy this test, only Michigan State and Kentucky did not, and they each fell early, which should not surprise any of us in hindsight.
At the end of the day, 2016 is again going to show that the teams that played the most efficiently throughout the entire season are the teams that win the post season tournament.
All of the fuss I made this season about the “change” that occurred with the UK team was interesting, and added some late season hope to a season marked by mediocre performance. But, I have again learned my lesson and will return to the full season basis for analysis with the next season, and abandon the segmentation of the season as a basis for the late season projections. I hope my desire to find some reason to hope for UK’s post season success does not lead me to such folly down the road.
Had I done that this year, I would have had UK falling out in the Sweet 16, with the recognition that the IU matchup was really a toss-up. But, based on the shift in basis I used for UK when projecting this tournament, I had UK winning the championship. All other tournament picks have been based on the conventional season long statistical basis. Through 60 of the 67 games, I lead the 247 Tournament Bracket Challenge again despite the bracket busting second round loss by my eventual champion. Part of that is because most of the UK fans in the 247 pool also had UK picked to go deep, but more significantly, my picks for the non-UK games have simply been more reliable again.
This system works, and is more effective than the vast majority of fan picks for simple winners and losers. The variation between predicted and actual winners occurs only because of human variations that introduce variance on the actual vs theoretical margins. The better a system projects a theoretical margin, the higher the percentage of selecting correct winners. That is the relationship that matters, not that the system predicts actual margins in a precise way for every game. When theoretical margins are legitimately small, the probability of the favorite winning approaches 50%. As the theoretical margin grows, the probability that the favorite wins also grows and approaches 100% as that theoretical margin exceeds 20 points.
One last point about this year’s tournament. Pre-tournament, the consensus was this would be a crazy tournament with upsets, upsets, upsets. How has that played out?
- First Round and First 4 Games: This Year 26 of 36 favorites won (72.2%) and 2012-2016 130 of 180 favorites won (72.2%). This year was right on target with prior years.
- Second Round games: This year 13 of 16 favorites won (81.3%) and 2012-2016 62 of 80 favorites won (77.5%). This season was very similar, with marginal fewer upsets than has typically occurred in prior years.
- Sweet 16: This year, 6 of 8 favorites won (75%) and 2012-2016 29-40 favorites won (72.5%) This season was very similar, with marginal fewer upsets than has typically occurred in prior years.
There have been fewer upsets in the first 60 games this year than in the 4 prior tournaments combined. Again, the conventional wisdom has not been so wise after all.
In the Elite 8, 2012-2015, the favorites have won 9 of 16 games (56.3%), and in the final 4 and championship games in 2012-2015, the favorite has won 6 or 12 games (50%). Thus the proof that this is the real tournament. Everything else that happens prior to this weekend of the tournament is for the pretenders to have a spot on the big stage, and a moment in the spot light. The real men of this game begin their work for this tournament today.