BIG BLUE FANS FOR
2014-15 Season Analytical Writings
I realize that a game like we saw last night can be discouraging, but it really shouldn't, at least not by itself. Game to game performance levels vary for the Cats, and for all teams. There is such a strong tendency to judge teams by their worst or their best performances, and by their last performance. However, when team average performance levels established over time are the basis of judgments, the entire situation becomes more manageable and understandable. This is how I stay off the emotional roller coaster, and how I strive to avoid over valuing my team and under valuing our opponents.
This graph shows the amount of variation in adjusted Net Game Efficiency (NGE) by half, by full game, and the running average through the first 4 games. The changes, half to half are very large. If I showed you the changes that occur for each 4 minute segment of a game, the variations would be even larger. The variations for a half are less volatile than the game segment variations, and the game to game variations are less volatile that breaking games into 1/2 game segments. The game to game changes, when averaged show some noise through the first 3 games, but by game 4, the noise level is becoming damped, and by games 5 to 8, there must be a trend shift to affect the running average to any significant degree.
This is why I believe the character of a team can be defined after the first 5 to 8 games of a season, and the current year data can be relied upon without calming it with prior year data starting with game 5. We are there for UK, several opponents, and the other opponents are nearly there. The Cats' Adjusted NGE is 0.407 ppp. That is outstanding. Generally, NCAA Champions end with adjusted NGE values of 0.300 ppp or higher, but there has only been one champion (Kansas 08) with an adjusted NGE over 0.40 ppp. This team is starting out in very special territory.
I have updated the database for opponents. At the beginning, based on 2014 results, the projected record for UK was 28-3. The predicted spread for UK @ UL was UL by 6 or 8 points, based on last year's data. UL is 1 game away for 2014-15 data only, and UK is now there. UK's Adjusted NGE has climbed dramatically, and UL's Adjusted NGE has been slipping, and after their 4th game, I expect it will slip a little more. Today, at their place, this is a pick em. By the time UL plays a few more games, UK will go into the Yum Center next month as a favorite, and I don't see a single projected loss as of today. A similar shift has been occurring with the SEC portion of the schedule. UK is now favored in all SEC games, including at Florida.
Bottom Line: Projection of 31-0 regular season is now OFFICIAL!!! First time ever in my system. For those who follow this with me, you will recall that my system had the game at IU in December 2011 as a projected loss, and as disappointing as it was, it was not a surprise based on this analysis.
Does that mean that the Cats will actually win them all? Of course not, but look at these predicted margins. The Standard Deviation is about 10 points typically, and rarely does a team lose a game when the predicted margin exceeds 10 points. That means as of today the only non-conference game that remains in play is at Louisville. As evidence of this, look at the variation between predicted and actual margins for the first 4 games.
Grand Canyon +2 1/2, Buffalo -10, Kansas +11 1/2, and Boston -7 1/2.
There are only 2 SEC games still in play, at Florida and at TAMU. The rest have predicted margins outside the Standard Deviaiton. They can be lost, but that is very unlikely.
Predicted record: 31-0. Worst projected record 28-3. 30-1 or 29-2 not surprising. That should be more than ample to maintain the #1 ranking, and the #1 seed going into the tournament. First 2 rounds in Louisville at the Yum Center, and a team like Michigan State, UConn, VCU, UCLA, or UNC in the Sweet 16, and one like Wichita State, Texas, Gonzaga, or Kansas in the Elite 8. Final 4: UK v Wisconsin and Louisville v Duke, Championship: UK v Duke ( A True Battle of the Titans).
Remember, Upset do occur. Here is how:
Submitted by Richard Cheeks
Submitted by Richard Cheeks