EXPLANATION OF DATA AND
Kentucky Completes Its Tune-up Games Against Weak Opponents. North Carolina Looms In the Wings
Coach Smith said before this season began that he preferred an easy, early schedule at Rupp Arena over a November filled with NCAA tournament quality opposition at the traditional early season neutral courts. His reasoning, he needs these games to integrate his yearlings into the College game. Tonight, Kentucky completed the four game sweep, and on Saturday we will learn whether Coach Smith's strategy has been effective.
In contrast to Kentucky's November, North Carolina will have played six games by Saturday against the likes of Santa Clara, BYU, Tennessee, Iowa, USC, and Indiana. Of the first six games, one was at Chapel Hill [USC], two on the road [Santa Clara and Indiana], and BYU, Tennessee and Iowa on neutral courts.
More about Saturday's game at UNC later.
Kentucky handled Tennessee Tech without any difficulty winning 92 - 63. Based on last season's NCAA RPI Strength of Schedule and season performance, I predicted a 36 point Kentucky margin, 95 - 59. The game performance was slightly weaker than expected on offense and defense. The defense forced 24 Tennessee Tech turnovers and forced Georgia State to shoot below 50%.
During the Tennessee Tech game, Kentucky had a total of 88 possessions, thirteen of which resulted from offensive rebounds. The average offensive efficiency for the first chance possessions for the game was 1.067 points per possession [ppp] while the average efficiency for the 13 second chance possessions was 0.929 ppp. For the first game this season, Kentucky had higher efficiency on first chance possessions than second chance possessions. However, for the third game in a row, Kentucky's opponent, was less efficient on second chance possessions [0.667 ppp] than first chance possessions [0.743 ppp].
North Carolina Looms!
On Saturday, two of the greatest college basketball programs collide once again. Kentucky has enjoyed recent success in this series that spans several decades. In recent years, since Dean Smith's retirement, the UNC teams have been below their usual standards, and Kentucky has dominated recently. Now, Coach Williams has UNC poised to regain its national prominence again.
Let's compare the teams
Win-Loss Record 4-0 5-1
Based on each team's early performance only, and no adjustment for strengthof schedule, I would make the following prediction: UK 76 UNC 72
However, this analysis ignores the differences in competition over the first 4 and 6 games. Is this difference significant? If it is, how will the difference impact the game's outcome?
I watched last night's UNC @ Indiana game, and was not impressed by either team. For the game, UNC had an offensive efficiency of 0.787 ppp and a defensive effectiveness of 0.670 ppp. In my opinion, Indiana's anemic point production speaks more about Indiana's offensive woes than it says about UNC's defensive prowess. However, what about UNC's poor offensive performance against Indiana. Indiana had six more possessions than UNC due to offensive rebounding margin. This could expose a weakness in the UNC team that the statistical comparisons obscure.
Based on each team's early performance and strength of schedule, I make the following prediction:
UNC 85 UK 66
Kentucky will taste defeat on Saturday against a North Carolina team that is hungry for a W against Kentucky, has been tested by several tournament tested teams. Kentucky's weak early schedule has produced a euphoria not justified by actual results.
My prediction methodology is based on average team performances for the season, to date, and the teams' respective RPI Strength of Schedule. Predictions have been tracked for about 40 games since late in the 2004 season, and the winner has been successfully predicted over 70 percent of the time. The average predicted score and actual scores are shown below together with the standard deviation. In this table, the last three columns show the mean and standard deviation for the differences between the predicted and actual scores for the home and visiting team, and the mean and standard deviation of the predicted v. actual scoring margin.
Of course, the system seldom predicts the actual score but over a period of about 40 game predictions, the predicted and actual scores track remarkably well. As the Point Differential data illustrates, the average difference between the home team's predicted and actual score for a game is less than 1 point, with a standard deviation of 9.2 points. Similar results for the visiting team and the game margin.
Tennessee Tech is the fifth game of this season. Kentucky's offensive and defensive efficiency for this season continue to emerge. Click below to view graphical presentation of the data for the first three games played this season. Companion Graphs for last season are also available so you can compare the progress of this season with last.
Copyright 2004 Richard Cheeks