Does Kentucky’s performance decline as opponent strength rises?

Kentucky coaches hope to get UK back on the winning track tonight at Ole Miss. (Vicky Graff Photo)

By RICHARD CHEEKS, Contributing Writer

People are asking an interesting question about whether individual players such as CJ Fredrick, perform more poorly against the better opponents than they do against weaker opponents.  Bluemonk presents data for Fredrick that suggests such a relationship.  I surmised that this type of relationship may be more prevalent than just Fredrick.  I cannot examine individual players, but I have tremendous data about the team’s performance.  

The following graph shows the raw net efficiency (RNE) for each game this season plotted against the opponents’ current Pomeroy efficiency rank.  The RNE is simply the final game margin divided by the number of possessions in that game.  As the rank of the opponent increases the raw net efficiency climbs.  This trend is one that everyone should expect.  The 7 losses are the 7 data points with negative values, and the loss to South Carolina sticks out like a sore thumb.  

As team performance rises or falls based on opponent strength, the collective individual performances will also rise and fall.  However, that would not necessarily be true of each player for each game, or even set of games, but generally speaking, a single player’s poor performance in a single game is not likely to explain the trend we see at the following link.

http://bigbluefans4uk.com/2022-23DataandWritings/SEASON_WRITINGS/RNE%20vs%20Opp%20Rank.png

I  prefer to look at adjusted efficiencies than raw efficiencies to evaluate game to game, season to season, etc.  The conversion from raw to adjusted requires factors for venue and the opponents’ strength.  The following graph shows the Adjusted Net Efficiency (ANE) vs opponents’ rank.  

First, it is not possible to pick out the wins from losses with this presentation because a team may perform reasonably well against a strong opponent yet lose the game.  For example, this team’s strongest performance in a loss occurred against UCLA with an ANE of +0.138 points per possession (ppp) and this team’s weakest performance in a loss was, no surprise here, against South Carolina with an ANE of -0.217 ppp.  

In a similar way, not all winning performances are created equal.  This team’s strongest winning performance was at Tennessee with an  ANE of +0.451 ppp and this team’s weakest performance with a win was an ANE of +0.004 ppp against LSU at Rupp.

When the ANE values are plotted against the opponents’ ranking, there is a very slight trend of rising ANE with increased opponent rank.  However, that trend is so slight, that I doubt that it is a significant trend, and I believe it is more likely that this team’s overall performance level is less related to opponent strength and mostly explained by the usual variations that all teams experience game to game over the course of a season. See this data at the following link.

http://bigbluefans4uk.com/2022-23DataandWritings/SEASON_WRITINGS/ANE%20vs%20Opp%20Rank.png