Tennessee 2013 Pomeroy Projection

#1
golfballs

golfballs

Mostly Peaceful Poster
Joined
Oct 28, 2009
Messages
75,414
Likes
57,668
#1
kenpom.com subscription

KenPom currently ranks us 38th which puts us at 5th in the SEC behind Kentucky (overall #1), Florida (11), Missouri (29), and Alabama (35)

Our regular season record is projected to be 18-10 (11-7) in the SEC.

However, on a game-by-game basis, he's projecting the following losses:
@ Georgetown (33% chance of winning)
@ Virginia (48%)
@ Alabama (34%)
@ Kentucky (14%)
@ Ole Miss (50%)
@ Vanderbilt (49%)
vs. Kentucky (33%)
vs. Florida (49%)

Close wins include:
vs. Memphis (53%)
@ Arkansas (54%)
@ Gerogia (52%)

Projected Final SEC Standings:

1. Kentucky 15-3
2. Florida 13-5
T3. Missouri 12-6
T3. Alabama 12-6
5. Tennessee 11-7
T6. Vanderbilt 9-9
T6. Mississippi 9-9
T8. Georgia 8-10
T8. Arkansas 8-10
10. Texas A&M 7-11
T11. Louisiana St 6-12
T11. South Carolina 6-12
T13. Auburn 5-13
T13. Mississippi St. 5-13
 
#2
#2
I see no way we lose to Virginia or Vandy, I'd be shocked. (assuming maymon is healthy).

Throw those in as wins and I'd be close to agreeing.
 
#3
#3
Be interesting to compare it to 2011-2012.
I bumped it but here's what he predicted last year.

13-16 (5-11)

66% chance of beating austin Peay
 
Last edited:
#4
#4
bleedingTNorange said:
I see no way we lose to Virginia or Vandy, I'd be shocked. (assuming maymon is healthy).

Throw those in as wins and I'd be close to agreeing.
Click to expand...

Kem Pom has always given a sizeable advantage for home-court advantage.

Vandy, if any, is one school that definitely has a home-court advantage
 
#5
#5
I think Alabama could be good enough to finish 4th, but their freshman will have to be as good as advertised.
 
#8
#8
How can a computer formula make preseason predictions? He only has previous seasons' data, which includes players that are now gone, and arbitrary recruiting rankings to use. There's also no way to value coaching changes. KenPom's predictions are about as useful as the preseason coaches poll.
 
Last edited:
#9
#9
ukvols said:
How can a computer formula make preseason predictions? He only has previous seasons' data, which includes players that are now gone, and arbitrary recruiting rankings to use. There's also no way to value coaching changes. KenPom's predictions are about as useful as the preseason coaches poll.
Click to expand...

He has projections for first year players that incorporate data from high school.

I've used KenPom as one of the factors for making preseason bets for the past few years. And in that realm - it's not about being accurate, it's about being more accurate than all the joe blows. so far so good.
 
#16
#16
bleedingTNorange said:
That's what I started to think, but if he's projecting 10 losses but only lists 8....does that mean he's projecting we lose 2 in puerto Rico?
Click to expand...

On a game by game basis, his forumula projects 8 losses.

But when projecting over the course of a season, there is a distribution of odds over a range of outcomes. The most likely of which, his formula predicts is 10. Most of those tossup games are hard to predict effectively on a game-by-game basis. But over the course of a season, one can expect to win a certain number of them. His system is all about statistics. That's the point of any computer model and ranking - it takes out the biases of the human element.
 
#17
#17
golfballs said:
On a game by game basis, his forumula projects 8 losses.

But when projecting over the course of a season, there is a distribution of odds over a range of outcomes. The most likely of which, his formula predicts is 10. Most of those tossup games are hard to predict effectively on a game-by-game basis. But over the course of a season, one can expect to win a certain number of them. His system is all about statistics. That's the point of any computer model and ranking - it takes out the biases of the human element.
Click to expand...

Gotcha, makes sense.

Add 2 wins in puerto Rico and one or two of the toss up games and I'm right there with him at 22 wins ish.
 
#18
#18
bleedingTNorange said:
Gotcha, makes sense.

Add 2 wins in puerto Rico and one or two of the toss up games and I'm right there with him at 22 wins ish.
Click to expand...

It could be a 30 percent likelihood of 10 losses and a 29 percent likelihood of 9 losses, and the system will say 10. Of course, at those odds it's essentially a tossup. That's why it's not an exact science - particularly in projecting how just one team will fare. But when looking at the big picture - all D1 teams throughout the entire season - things tend to even out and his projections provide a decent overall picture (in my opinion). Of course, like all preseason rankings and projections - it should be taken with a big grain of salt.
 
#21
#21
I like KenPom's site midseason and leading up to Selection Sunday, however preseason it's a bit of a joke. I think UT easily does 3 games better than his preditions. UT could reasonably be favored @ Bama, OleMiss, & @ Vandy and that would put us even with UF @ 2nd in the SEC, which is where I think we finish.
 
#22
#22
bleedingTNorange said:
Gotcha, makes sense.

Add 2 wins in puerto Rico and one or two of the toss up games and I'm right there with him at 22 wins ish.
Click to expand...

For someone that follows basketball as much as you do, you need to set aside a few hours and $20, get a year subscription to kenpom.com, and look over the numbers.

It's astounding what those numbers can tell you about teams, conferences, etc.

Probably the single most useful tool in analyzing the college game.
 
#23
#23
MemphisCanes said:
For someone that follows basketball as much as you do, you need to set aside a few hours and $20, get a year subscription to kenpom.com, and look over the numbers.

It's astounding what those numbers can tell you about teams, conferences, etc.

Probably the single most useful tool in analyzing the college game.
Click to expand...

Yeah, I just put down my $20 today. His site speaks for itself. When analysts from all over the college basketball landscape go to Kenpom for data you know the guy is on top of his game. Come tournament time the 2 rankings you hear talked about the most are RPI (because the committee uses it) and Kenpom.com (because many consider it the best for statistical info).
 
#25
#25
MemphisCanes said:
For someone that follows basketball as much as you do, you need to set aside a few hours and $20, get a year subscription to kenpom.com, and look over the numbers.

It's astounding what those numbers can tell you about teams, conferences, etc.

Probably the single most useful tool in analyzing the college game.
Click to expand...

You've just set up my evening for me thanks, I've only heard great things, probably time to give it a shot.
 
New posts
You must log in or register to reply here.

VN Store



Back
Top