BIG BLUE FANS FOR
2014-15 Season Analytical Writings
My statistical analysis of UK and each UK opponent projects a margin for each game before it occurs. Over the years, this model has predicted the margin of games with an average variation between the predicted and actual margin has been within +/- 1 point with a standard deviation of the variation of about 10 points.
Not all games fall close to the predicted margin, but most do. For example, in 2014-15, eight of the 31 games ended within +/- 2.5 points of the predicted margin, and 10 more were within 2.5 to 7.5 points, 5 on each side of the center. That is 18 of the 31 games. Two games had final margins that were 17 ½ or more points higher than predicted, and two games had final margins that were 17 ½ point or more lower than the predicted margins. These four games out of 31 represent the extreme outliers that occur on each side of the ledger. They occur in about a 1 in 8 frequency for nearly all teams and seasons.
These Cats did not lose the two outliers that represent severe under performances because in each of those games, the Cats were so much better than their opponents that even such a poor game performance was not sufficient to overcome the inherent margin that exists between UK and each of those teams. In most seasons, these games would have been disappointing losses that tend to send the fan base into a tailspin wrought with questions about the coach's judgment, or the players' abilities. The remaining 9 games fall within the 7 ½ to 17 ½ ranges on either side of zero, 5 on the plus side and 4 on the negative side this season.
Over the course of the season, the Cats faced five opponents for which the pre-game projected margin was less than 10 points. In those five games, three ended within +/- 2.5 points of the projected margin, and 1 fell on each side of zero in the 2 ½ to 7 ½ point range. Those games, with the projected and final margins are @Louisville (4 ½, 8), Vanderbilt at Rupp (5 ½, 8), @Florida (8 ½, 7), @Georgia (9, 8), and @Texas A&M (9 ½, 6).
As the Cats enter March Madness, it should be clear to all that the model will project the Cats as favorites for each game of the tournament. When that projected margin exceeds 7 ½ points, the probability of a game performance poor enough to shift that game into a loss are quite low. However, there are seven teams that could stand against this UK team with projected pre-game margin between 2 ½ and 5 points. Any of these seven potential opponents may pose a legitimate risk to stop the Cats' run toward a perfect season and national championship. The seven teams in this category, and the projected margins in a game with UK are Wisconsin (2 ½ points), Virginia (3 points), Gonzaga (3 ½ points), Duke (4 points), Arizona (4 ½ points), Villanova (5 points), and Utah (5 points).
In my opinion, the Cats will probably have to beat 3 of these seven teams to win #9. Which three, and in what order will depend upon the tournament seeding. The only way the Cats would have to face more than 3 of these would be if the committee assigns one or more in the Cats' path as an 4 seed or higher on the Cats' side of the Regional bracket. The Cats could actually face fewer than 3 of these if they fail to advance in the tournament according to their seeding.
The March toward April In Search Of Perfection begins next week in Nashville at the SEC Tournament.
Submitted by Richard Cheeks
Submitted by Richard Cheeks