• Welcome to OGBoards 10.0, keep in mind that we will be making LOTS of changes to smooth out the experience here and make it as close as possible functionally to the old software, but feel free to drop suggestions or requests in the Tech Support subforum!

Ken Pomery/Tennessee Rank Question (NWT)

WakeHornet

Member
Joined
Apr 7, 2011
Messages
160
Reaction score
9
Can someone who understands Kenpom's ranking criteria well help explain Tennessee's #13 overall ranking to me? It is quite an outlier. How much weight does Kenpom put in their 35 point victory over UVA in December? Is this the sole reason for their high rank (e.g. higher than Michigan, Syracuse, Iowa State)?

http://kenpom.com/
 
Last edited:
prepare for a meaningful contribution
 
I do, which is why I asked the question. Can you help?

I'll do my best to answer.

29th best adjusted offensive efficiency and 16th best adjusted defensive efficiency

Yes, they lost 12 games, but their margin of victory in wins was pretty impressive. They have a decent amount of conference wins by 15+ points. Say what you want about the SEC, but that's very impressive. Their overall strength of schedule was 46th, which helps in their adjusted rankings.
 
prepare for a meaningful contribution

Lol.

KenPom doesn't weigh any one game more than any other. In fact, it just views games as a sample size X possessions against a team who has a specific (but dynamic) efficiency on both offense and defense. The concept is that games are arbitrary constructs of time, where possessions are actually a better unit to use in determining how good or bad a team is. Essentially it boils down to how many possessions you score on and how many times you prevent the opposing team from scoring. These numbers are garnered by taking points per possession and then adjusting each game for the strength of the opponent. The system accounts for both offensive and defensive efficiency, in other words if you are playing a fantastic offensive team then it may be a great defensive performance to hold them to just over one point per possession (which is considered to be roughly "average.")

So more specifically, Tennessee's case, when they win they win big (21 wins - 18 by double digits) and when they lose it's generally close (12 losses - 11 by single digits). Since the KenPom system doesn't account for one possession counting any more than another (i.e. if you're down 61-60 with 2 seconds left, it doesn't vastly impact you one way or the other over the course of the season whether or not you score on that possession and thus win or lose the game), wins and losses don't paint the whole picture but rather serve as a proxy for how good the team really is. Tennessee has the 29th best adjusted offensive efficiency and the 16th best adjusted defensive efficiency, which brings to an overall 13 on KenPom.

If you have any more questions, as Milhouse and Moonz will surely point out, I will be more than happy to answer.
 
nm, numbers beat me to it.
 
They were the 341st luckiest team (10th from the bottom) in the country so you'd expect their W-L to be much better if you replayed the season 10,000 times.
 
They lose TWELVE games. It's insane to give excuses. They LOSE games and LOTS of them.

You can play with hypotheticals all you like, but they still lose games.
 
Also, teams with very good KP rankings with low "luck" are generally underseeded by the committee (see: Tennessee, Oklahoma State, Kentucky, Iowa) while teams with poor KP rankings with high "luck" are generally overseeded (see: Colorado, Texas, N.C. State, Iowa State).
 
They lose TWELVE games. It's insane to give excuses. They LOSE games and LOTS of them.

You can play with hypotheticals all you like, but they still lose games.

Right, but these statements assume that there is an art to winning or losing close games. Good teams don't win close games at a higher rate than bad teams, they just play fewer close games.
 
Of their 12 losses

-3 were to Florida
-1 was at Wichita State
-1 was at Kentucky
-1 was at Xavier

No shame in those being half of a team's losses.
 
KP heavily weighs wins against juggernaut Wake Forest.
 
I'll also add that I enjoy any thread that gives rj a chance to argue against statistics.
 
It's total crap. They lost TWELVE games. It's irrelevant how badly they lost. At Top 15 team wouldn't lose all those close games. That's what makes them a Top 15 team.

They aren't at Top 40 team. They don't have the balls to win close games. Good team do.

They are the epitome of what's wrong with computer rankings. In REAL games, they don't win. It is irrational to say they are good when they lost on the court.

Plus they play in a very bad conference.
 
Example:

Duke played 12 games where the score was in single digits, they went 7-5 in those games.

Tennessee played 12 games where the score was in single digits, they went 2-10 in those games.

You could say Martin sucks as a coach, you could say Coach K is fantastic, or you could say Tenessee generally got a little bit unlucky since you would anticipate single digit games roughly turn out a 50/50 proposition over the long run.

On the other side is double digit games:

Duke played 22 of these games and went 19-3 (including a double digit loss to Wake)

Tennessee played 21 games and went 19-2
 
It's total crap. They lost TWELVE games. It's irrelevant how badly they lost. At Top 15 team wouldn't lose all those close games. That's what makes them a Top 15 team.

They aren't at Top 40 team. They don't have the balls to win close games. Good team do.

They are the epitome of what's wrong with computer rankings. In REAL games, they don't win. It is irrational to say they are good when they lost on the court.

Plus they play in a very bad conference.

So let's say team B plays team A 100 times and loses by 1 point in every game and also that team C plays team A 100 times but loses by 15 points in every game. Do you really not think that team B is better than team C?
 
Kansas lost NINE games. Should they be like an 8th seed by your standards?
 
Back
Top