Sunday, February 05, 2012

W-5 BP68

We're now more than halfway through conference play, and five weeks from Selection Sunday. What this means is that brackets now come out twice weekly. In addition to the regular bracket after the completion of Saturday's games, I'll also have a bracket out after Wednesday's games. So stay tuned for the W-4 BP68 some time after Wednesday night's games end.

At the top of the bracket I've moved Ohio State into the #1 overall spot. I don't think there's much doubt at this point that Ohio State is the best team in the country. I was holding them back because their level of competition to earn the Big Ten tournament title was a lot higher than the competition that either Kentucky or Syracuse will face. But Ohio State's done enough to convince me that they'll be able to overcome even that. Certainly if they sweep the Big Ten regular season and tournament titles, they should end up as the #1 overall seed (not that being the #1 overall seed really means anything - the last two years the #1 overall seed ended up with a brutal Tournament draw compared to some of the other 1 seeds).

Meanwhile, I talked earlier this season about how thin the bubble was. It's not that the bubble is bad (for the umpteenth consecutive year, television analysts are whining that the bubble is terrible, and for the umpteenth consecutive year these same people will whine on Selection Sunday about the "deserving teams" that got left out of the Tournament), but just that it seemed thin. I was only choosing between a few teams for the final few spots.

I think that's changed over the last week. The final few teams in the Field of 68 struggled, while the best of the teams outside the Field of 68 picked up some nice wins. So the bubble is fatter than it was last week. That said, I still struggled to fill the final few spots. All of the teams I have holding the final four or five at-large spots are teams that I think have a below-50% chance of an actual at-large bid, I just liked the other bubble teams even less. The one change to the field was that I moved in Notre Dame, and booted out Dayton to make room.

Among the automatic bids, there wasn't much of a shift. The one change was in the NEC, where I moved in Wagner in the place of Robert Morris.

I eliminated six teams from at-large contention: Boise State, Fordham, Georgia Tech, South Carolina, Utah State and Wisconsin-Milwaukee. That leaves 68 teams not in my current bracket that still have a chance at an at-large bid.

The following are my typical disclaimers:

If I projected your favorite team below where you think it deserves to be, it's because I hate your favorite team. If I projected a team above where you think it deserves to be, it's because I secretly love them and have an incredibly blind bias in their favor.

On a more serious note, this is a projection of the final bracket on Selection Sunday, and not a listing of how I think teams would be seeded if the season ended now. There's a difference.

Here we go:


1. OHIO STATE (BIG TEN)
1. KENTUCKY (SEC)
1. SYRACUSE (BIG EAST)
1. NORTH CAROLINA (ACC)

2. KANSAS (BIG 12)
2. Florida
2. Duke
2. Georgetown

3. UNLV (MWC)
3. Missouri
3. Michigan State
3. Baylor

4. Marquette
4. Virginia
4. Wisconsin
4. TEMPLE (ATLANTIC TEN)

5. Indiana
5. San Diego State
5. ST. MARY'S (WCC)
5. Michigan

6. Vanderbilt
6. Louisville
6. Florida State
6. Illinois

7. CREIGHTON (MVC)
7. Texas
7. UConn
7. Alabama

8. West Virginia
8. Gonzaga
8. Southern Miss
8. Kansas State

9. Saint Louis
9. Wichita State
9. MURRAY STATE (OVC)
9. New Mexico

10. Purdue
10. CALIFORNIA (PAC-12)
10. Iowa State
10. MEMPHIS (CONFERENCE USA)

11. HARVARD (IVY)
11. BYU
11. Mississippi State
11. Seton Hall

12. Northwestern
12. LONG BEACH STATE (BIG WEST)
12. Notre Dame
12. Cincinnati
12. NC State

13. Xavier
13. Stanford
13. IONA (MAAC)
13. BELMONT (ATLANTIC SUN)
13. MIDDLE TENNESSEE ST (SUN BELT)

14. ORAL ROBERTS (SUMMIT)
14. OHIO (MAC)
14. GEORGE MASON (COLONIAL)
14. DAVIDSON (SOUTHERN)

15. NEW MEXICO STATE (WAC)
15. BUTLER (HORIZON)
15. BUCKNELL (PATRIOT)
15. WAGNER (NEC)

16. MONTANA (BIG SKY)
16. COASTAL CAROLINA (BIG SOUTH)
16. NORFOLK STATE (MEAC)
16. STONY BROOK (AMERICA EAST)
16. TEXAS-ARLINGTON (SOUTHLAND)
16. MISSISSIPPI VALLEY STATE (SWAC)

Teams seriously considered that just missed the cut:
Miami (Fl), Dayton, St. Joseph's, Minnesota, VCU, Marshall, Cleveland State, Northern Iowa, Colorado State, Arizona, Washington

Decent resumes, but not good enough:
Virginia Tech, La Salle, UMass, Oklahoma, Drexel, Central Florida, Wyoming, Colorado, Oregon, Arkansas, LSU, Mississippi, Nevada

Long shots, but still in the at-large discussion:
Maryland, St. Bonaventure, Pittsburgh, South Florida, Villanova, Iowa, Nebraska, Oklahoma State, Tulsa, Akron, Drake, Illinois State, Missouri State, UCLA, Oregon State, South Dakota State

Still alive, but pretty much need a miracle:
Clemson, Wake Forest, Charlotte, Duquesne, Richmond, DePaul, Providence, Rutgers, St. John's, Weber State, Penn State, Texas A&M, Georgia State, Old Dominion, UTEP, Tulane, Valparaiso, Loyola-Maryland, Buffalo, Kent State, Indiana State, Air Force, TCU, Washington State, Auburn, Georgia, Tennessee, Denver

29 comments:

Rainmaker203 said...

Seton Hall still in the field after losing 6 in a row?

I don't care how weak the bubble is, that's a bold prediction.

Jeff said...

I think that Seton Hall will be favored in 5 of their final 7 games, and if they do go 5-2 in that stretch and then win a game or two in the Big East tournament then they'll be in pretty good shape.


Not that I feel that confident that Seton Hall will make the Tournament. I just don't feel that any of the teams behind them are in a better situation. Look at the resumes for NC State or Stanford or Xavier. They're all pretty mediocre at this point.

Anonymous said...

Look...

Mizzou has beaten ku and Baylor. They're 21-2 with excellent peripheral numbers as well.

I'd like to hear you justify why they're not better than a 3, especially with their remaining schedule (other than at ku, every other "tough" game is at home).

How is Mizzou a 3 and ku is the highest 2? okay, whatever, your personal bias and what not.

Also, switch Iowa State with Texas and I can buy it. But even at a 7, you are buying Texas waaaaaaaay too high. Your prerogative, but your reputation as a college basketball "expert" on the line.

Jeff said...

Kansas came into the Missouri game rated higher in every computer rating that measures team quality, and they only lost by three on the road despite some blown ref calls going against them (in normal cases, home court was worth four points, but it was obviously worth more here). So, I don't see how anybody can come out of that game objectively thinking Missouri is better than Kansas.

I understand that you're a Missouri fan, and everybody at a certain Missouri website website has decided that I'm either the dumbest person in the history of mankind, or I'm some secret Longhorn Network plant.

I'm just going to warn all you Missouri fans again that I will not hesitate to delete comments if they get personal. I try very hard to keep the level of discourse here very high. If you want to engage in typical message board trash talking, take it to another website.

Rainmaker203 said...

I really don't think Texas at a 7 is "way too high". Aside from playing Kansas in Lawrence (March 3), every game left on the schedule is winnable. Their toughest game before that is @Oklahoma, so they could easily go on a winning streak between now and the last game.

The 10-3 OOC with a win over Temple (who is currently all alone in first in the A-10 and projected as a 4 seed in this bracket) is not going to hurt the Longhorns in the at-large process.

I agree that Iowa State will definitely be in the NCAA tournament though.

Rainmaker203 said...
This comment has been removed by the author.
David Mann said...

I always enjoy your in depth analysis, and as a die-hard Missouri-fan, I definitely don't think you're the dumbest person in the world...I doubt if folks would actually take the time to read your blog on a regular basis if they really thought that.

Okay, but now I have to ask something. Assuming we're able to take down Oklahoma on the road (which might be a trap game, given they are improving and we may be a bit overconfident heading there after the huge win against them at home earlier), and we beat Baylor on Saturday, then in your mind do we get into the discussion for a 1 or 2 seed? It may very well be the case that Kansas is a better team, I don't want to argue that. But with them at 5 losses already, if we're able to beat Baylor next weekend, don't you think by then we'll at least have the stronger resume? Of course us beating both Oklahoma and Baylor is a pretty big "if" but I'm just curious what you think.

Jeff said...

Thanks for the question, David. I think that for Missouri to earn a 1 seed they need to earn at least a share of the Big 12 regular season title and then need to win the Big 12 tournament.

There are a few ways for them to get to a 2 seed. Winning the Big 12 regular season title and then losing in the Big 12 tournament semis or finals would probably do it. Or else finishing a game or two back in the Big 12 standings and then winning the Big 12 tournament. I think either of those scenarios would earn them a 2 seed.

David Mann said...

Okay, that seems pretty believable. Thanks.

Anonymous said...

I'm glad a lot of Missouri fans called this guy on his stupidity. Instead of backing off the handful of terrible predictions he made, he has only become more entrenched in those terrible predictions.

Anonymous said...

Iowa State > Texas

Jeff said...

Thanks for that constructive criticism.

By the way, can you find any objective computer rating measuring team quality that puts Iowa State ahead of Texas? I can't.

Tom said...

Jeff, FWIW here's how Real Time RPI (they use Sagarin, right?) projects the Pac-12 to finish at this point:

Arizona ..... 13-5
UW ............ 12-6
Cal ............ 12-6
CU ............. 12-6
UO ............ 11-7
Stanford .... 11-7

schedule very favorable for Arizona with the exception of going to Washington. Schedule difficult for UW and Cal

Anonymous said...

I've got something better than objective computer analysis...

results.

Iowa State is 7-3 in one of the best conferences in the country. Texas is 4-6.

Jeff said...

So Anonymous, you think that won/loss records over small sample sizes and mis-matched schedules are proof of which team is superior? And I'm the idiot?

Jeff said...

Tom, I'm not sure what realtimerpi uses, actually. I'm not a paid member of that website.

If you want to see Sagarin projections, you can use rpiforecast.com.

Right now, both Sagarin and Pomeroy are projecting the same finish: Washington & Cal at 13-5. Arizona & Colorado at 12-6. It is true that Arizona has an easier schedule remaining than the other top Pac-12 teams, but asking them to go 6-1 is asking a bit much.

If they can somehow pull it off, it will definitely put them in the Field of 68, and they'd probably only need one more win in the Pac-12 tournament to stay in.

Ken Miller said...

I'm a bit surprised that you are sticking with Butler in the Horizon. Considering where Pomeroy has the teams in that conference plus the league standings, it seems to me about time to slot CSU in at about a 13 seed.

As for Texas, although I think their youth will keep them from performing quite to the level that the statistics say they should, they are improving in time to salvage the season. How far does Texas need to go in the conference tournament to pull off the 7 seed?

Anonymous said...

We are anonymous.

Jeff said...

Ken, I know that Cleveland State has been playing better than Butler, but that was true last season also. Butler seems to just not show any urgency during the regular season and they turn it on when it counts most. They have had a bunch of games this year where they played excellent, and a bunch more where they were awful, it's been a mixed bag.

If Cleveland State really destroys Butler on Saturday then I might make the switch. I also might move CSU in as an at-large team - they were probably the first team out this time around. Their lack of quality wins is really just what's holding me back.

And if Texas gets to 10-8 then I think just one win in the Big 12 tournament will get them a 7 seed. If they only go 9-9 then they probably need a pair of wins.

Anonymous said...

Sample sizes aren't small any more! More than half of the conference season has passed! At some point, you are who you are. While I don't think that is absolutely true yet, a team that is 7-3 in conference deserves to be a) considered legitimate, and b) more likely to finish with a better resume at the end of the season than a team that is currently 5-6.

How can you be so stubborn on this?????

Jeff said...

Anonymous, I'm trying to explain basic Sabermetrics/APBRmetrics here. We don't have to change years of research in scoring efficiency because you don't like the answers.

Judging a team by their raw won/loss record is like judging a major league pitcher on their won/loss record, or an NBA player on how many points he scores per game. It's an extremely rough measure that involves a whole lot of stuff outside the control of the players.

Anonymous said...

I understand that. You might not believe it, but I am well-versed in understanding sabermetric data and quantitative points of view.

All I am saying: you can have a prediction for how the final 7 or 8 games of a team's season are going to go. If you are going to have a prediction, it's wise to use efficiency ratings. In fact, using anything else would be silly.

But those efficiency ratings cannot change what already has happened. Iowa State is 17-6 overall, and 7-3 in the second best conference in the country with victories over Kansas, Texas, and Kansas State, and their efficiency rating is a respectable 35th. They are PREDICTED to go 11-7 in the Big 12 by Pomeroy. That would make them 21-10 going into the Big 12 tourney.

Texas is 15-9, 5-6. Even the most optimistic prediction would put them at 11-7, but I'll go with KenPom and say they'll be 9-9. (KenPom is one of your preferred tools of analysis, right?).

So, using data from the guy whose predictive stats are referenced by every knowledgeable college basketball mind, Iowa State is projected to be 21-10, 11-7, and Texas 19-12, 9-9. But you're saying that Texas will be a 7 seed and ISU a 10.

All we're saying is that if you flip the two teams, it might make sense. Even the most optimistic end to the season for UT will make them (at minimum) a 10 loss team and no better than a 6 seed.

The committee won't seed by efficiency ratings at the end of the year. They seed on results.

You should, by now, be starting to see why your version of "predictive analysis" does not jive with everyone else's. More than anything, the persistent overvaluing of Texas, when both results AND predictive data flies in the face of your theory, is irrational.

Jeff said...

I asked you to find an objective computer metric that rated Iowa State as better than Texas and you responded with won/loss records. You can say that you're "well-versed" in sabermetrics, but your posting doesn't reflect that. It would be like responding to somebody using OPS+ with RBIs.

I've explained myself on this stupid Texas issue about 100 times to Missouri fans. You can re-read what I wrote or not, I don't care. I'm not going to spend the next month of my life fighting this same argument over and over again.

Anonymous said...

I never claimed that ISU has a better computer metric anywhere than UT does. Not once did I say that.

But a good objective computer metric cannot even buy you a cup of coffee if you don't WIN, a la 2011-2012 Texas.

Sugarcoat the facts all you want.

Jeff said...

You said Iowa State was better than Texas. I asked for any objective computer rating that agreed with you. And your response was:

"I've got something better than objective computer analysis...

results.

Iowa State is 7-3 in one of the best conferences in the country. Texas is 4-6."

I'm sorry, but that's the "I don't care how many more strikeouts the other guy has, this guy has more wins!" argument. And I understand that the attitude you have is going to stay mainstream for the foreseeable future because most people can't calculate 15% tips, let alone understand statistical analysis. But that doesn't mean I have to lower the level of my website to make the lowest common denominator happy.

Anonymous said...

You said "I don't think there's much doubt at this point that Ohio State is the best team in the country."

That is a bit of an overstatement. Ohio State is a great team and the argument can certainly be made that they deserve the #1 overall seed. But I disagree that they are certainly better than Kentucky and Syracuse.

I think you rely too heavily on computer rankings, and particularly the Pomeroy rankings, which I believe are inherently flawed if they rank a team like Wisconsin as the #4 team in the country as they currently do. Wisconsin lost at home to a mediocre Iowa team and to Marquette without their starting PG Junior Cadougan, and on the road by 18 points to Michigan. There is no way they are better than Syracuse as they are currently ranked in Pomeroy.

Say what you will about computer rankings but Ohio State has already lost 3 times this year.

Kentucky is the hottest team in the country right now, blowing out its opponents practically every game.

Syracuse is a different team without Fab Melo, and they could very well be undefeated if he didn't miss any time.

I'm not saying Kentucky or Syracuse is certainly better than Ohio State. I'm just saying you can't say for sure that Ohio State is better than both of those teams.

I enjoy your blog. Keep it up.

Jeff said...

Surely you know that I don't just regurgitate Pomeroy ratings. And I've written a couple of posts recently where I go into the reasons why OSU is the clear best team.

The reason why Kentucky and Syracuse are rated higher than Ohio State in the human polls is because they have fewer losses, but that has to do with schedule strength more than anything else.

For example, Ohio State is outscoring opponents by 0.27 PPP in conference play. Kentucky is outscoring opponents by 0.29 in conference play, and Syracuse is outscoring opponents by 0.19. Considering how much stronger the Big Ten is than either the SEC or Big East and the fact that Kentucky's hardest conference games are still to come (they haven't played either of their games against Florida or Vanderbilt yet), Ohio State's conference performance has clearly been the most impressive of any team.

Ohio State is blowing away their performance from last season, and it's not even that close. 2011-12 OSU would be a several point favorite against 2010-11 OSU, and 2010-11 OSU was the #1 overall seed last year.

By the way, I don't think Wisconsin is that far off from accurate in KenPom. They're not the 4th best team in the country, but they're probably in the 10-15 range. Their shooting against Iowa and Michigan was incredibly awful - a statistical fluke that shouldn't be taken too seriously. No team is going to win too many games when they hit 10% of their threes.

Phil said...

I think your analogy on won/loss record for a basketball team to won/loss record for a baseball pitcher, or strikeouts for a pitcher, is flawed. A better analogy might be W/L for a baseball team (as compared to Pythagorean W/L). Just because, say, the Brewers are 70-50 at the 3/4 mark and the Cardinals are 65-55 doesn't mean the Brewers are the better team ... schedules may vary in difficulty, luck plays a role, etc. etc.

But I think Anonymous has a point. At SOME point in the season, W/L record overtakes sabermetrics in predictive importance, at least if we're trying to predict which teams will make the tournament (or playoffs) and how they'll do once they get there. When the Brewers are 90-65 and the Cardinals are 82-73 (and thereby mathematically eliminated), we can definitively say the Brewers were the better team, regardless of what any other statistics show. Likewise with college basketball, if Iowa State gets to 11-7 and Texas only gets to 9-9, seems to me we can definitively say Iowa State was the better team for the season, no matter what the computers say, because their record tells us so. That's not to say Iowa State would always beat Texas, obviously, but ON THE WHOLE, Iowa State was better.

So it seems to me that you are correct that we're not quite to that point in the season yet, and that objective metrics still have a lot to tell us ... but Anonymous is also correct that W/L is BECOMING more and more an indicator of "good" teams as games are played and , at some point, W/L overwhelms any other metric.

But I was a political science major 20 years ago, so what do I know! Pick apart as you see fit. It's an interesting discussion.

Jeff said...

Well, the difference with baseball is that you play 162 games. And even in baseball, the "luck" factor can cost a team +/- 5 runs relative to their "Pythagorean record".

If you take two teams that are equal and deserve to go 10-8, only one has really good luck in winning a game and one has really bad luck in losing a game, you've made those teams 11-7 and 9-9 already. And obviously each team in the Big 12 will end up playing 6-7 games where luck will be a significant factor.

Now, in the Big 12 we have a true round robin, so everybody's schedule evens out. So if we look at the PPP differential at the end of the season, after a true round-robin, then we will see pretty clearly who is better than who. If one team is +.10 per possession and another team is -.02, then we can say pretty definitively that the first team was better. Those two teams could end up with the same records because of luck, of course. But point differential is much better to look at than wins/losses.


Now, you can argue against me here and point out that Iowa State is +0.07 PPP so far and Texas is +0.02. That's not a "definitive" margin after 11 games, but it's still a significant stat. But my response is two-fold. First, I think Iowa State has played an easier schedule so far and will have a more difficult schedule the rest of the way. And second, I think Iowa State had a few games there where everything was clicking, and they're going to have a couple of "reality check" clunkers the rest of the way. Meanwhile, in my personal opinion, Texas is a team that will keep getting better throughout the season.

Of course, that is just my opinion, and you're free to disagree.