Sunday, February 23, 2014

W-3 BP68

Three weeks from Selection Sunday... is the bubble growing or shrinking? I suppose the answer depends on what you mean by the questions. There appear to be a few more teams that are within a single win or loss of realistically being in the Field of 68. At the same time, the number of teams within vicinity of the bubble is reducing, with the herd being culled daily.

In this bracket, two at-large teams at the bottom changed places. Baylor and BYU joined the projected Field of 68, while Minnesota and Missouri dropped out. Remember again that this doesn't mean I think that Baylor and BYU are making the NCAA Tournament (I'd still put their odds below 50%). I just think their odds are better than any team that is currently out of the bracket.

At the top of the bracket the 1 seeds remain very confusing. The two teams that control their own 1 seed destiny are Florida and Wichita State - they'll earn a 1 seed if they win out. The Big Ten, Big 12 and ACC are all at risk of losing a 1 seed simply because the top teams are beating each other too much, which opens the door for Arizona, though we need to see more than one great Arizona performance before believing that they've solved the problem of the Brandon Ashley injury.

One other issue starting to arrive is that the bracket is getting clogged around the 2, 3 and 4 seeds. I've talked before about the parity this season and just how little gap there is between, say, the 6th and 14th best teams in the country. And the result is that there just is so little gap between the teams fighting for the last 2 seed and the teams fighting for 3 and 4 seeds. Unless we get a lot more clarity in the next 3 weeks, those seed lines are going to be very hard to predict.

Four teams were eliminated from at-large contention since the last bracket: Auburn, Georgia Tech, Mississippi State and Utah State. That leaves 41 teams not in my bracket that still have a chance at an at-large bid. Around 30 of those teams will be eliminated over the next three weeks.

Please remember that the following is a projection of Selection Sunday and not a list of where teams would be seeded if the season ended now. The teams in all-caps are the projected automatic bid winners for their respective conference.

1. KANSAS (BIG 12)

2. Syracuse

3. Michigan
3. Villanova

4. Virginia
4. Wisconsin
4. Iowa

5. Cincinnati
5. North Carolina
5. Iowa State
5. Ohio State

6. Kentucky
6. Texas
6. UConn

7. New Mexico
7. Memphis
7. Arizona State
7. UMass

8. Oklahoma
8. SMU
8. VCU
8. Pittsburgh

9. Kansas State
9. George Washington
9. Stanford

10. Xavier
10. California
10. Saint Joseph's

11. Oregon
11. St. John's
11. Tennessee
11. Colorado

12. Georgetown
12. Clemson
12. Baylor
12. BYU





Teams seriously considered that just missed the cut:
Florida State, Marquette, Providence, Minnesota, Nebraska, Oklahoma State, West Virginia, Southern Miss, Missouri

Decent resumes, but not good enough:
NC State, Dayton, Richmond, Indiana State, Boise State, Utah, Arkansas, LSU, Saint Mary's

Long shots, but still in the at-large discussion:
Maryland, Indiana, Middle Tennessee, UTEP, Ohio, UNLV, Wyoming, Mississippi

Still alive, but pretty much need a miracle:
Miami (Fl), Notre Dame, Wake Forest, St. Bonaventure, Seton Hall, Illinois, Penn State, Purdue, Texas Tech, Akron, Oregon State, Washington, Georgia, Texas A&M, Vanderbilt


DMoore said...

Does today's Michigan State loss change your prediction of MSU as a #1 seed? I have trouble seeing how any team out of the Big Ten can earn a 1 seed. They will simply have too many losses, and every team is a long shot to win the conference tourney because there are so many contenders.

Is Arizona listed as a 2 seed because of the uncertainty post-Ashley? Comparably, they have only 2 losses against a very slightly stronger schedule.

Jeff said...

Yeah, I think if I was doing a bracket right at this moment, I'd drop Michigan State to a 2 and make Arizona a 1. Going to be awfully hard now for Michigan State to earn a share of the Big Ten regular season title.

My concern with Arizona is that not including the game against Colorado last night they've been awfully mediocre in their games since losing Ashley. I want to see them do that again before I believe that game wasn't just a one-off fluke.

Daniel Tappa said...

Going to continue the Wisconsin discussion we started in your previous bracket projection. I understand that their final 4 games aren't going to help their RPI, SOS, etc. But, it also shouldn't really hurt their RPI as long as they win those 4 games. Again, Wisconsin not being at least a 3 seed in your projection is a total joke. They had 1 minor slip up for 2.5 weeks, which seems to be long gone now. They arguably have the most impressive resume in the country, with 5 wins against current top 20 RPI teams (Florida, @ Virginia, (N)St. Louis, Michigan State, @Michigan, @Iowa). They are 10-2 in road and neutral games (W's vs St. Johns, St. Louis, West Virginia, Virginia, Northwestern, Purdue, Illinois, Michigan, Iowa and L's @ Minnesota and @ Indiana during their funk when they forgot how to play defense).

Your recap of the Wisconsin @ Iowa game was pretty interesting. You made it sound like Iowa was in control and Wisconsin stole it. After the first few minutes of the game, Wisconsin took control and had the lead for the rest of the first half (leading by 5-10 for most of the half), and Iowa didn't have the lead again until 10:55 left in 2nd half, then the 2 teams played very close until Wisconsin finished them off.

Derek Kramer said...

You may need to think about switching around Iowa and Iowa State. In what universe does Iowa's resume look better than Iowa State's? Plus Iowa St. beat them, and soon to be Big 10 champs, Michigan. And while you're at it, Michigan St. is not currently a 1 seed either.

Jeff said...

Nobody is saying that Wisconsin won't be helped if they win all four of their remaining games, I'm saying that the slight reward isn't worth the huge risk, which is that the odds are that they'll lose at least one of those games (i.e. a "bad" loss). It's just awfully hard to get through a schedule like that unscathed.

You can argue that Iowa State's resume is slightly better than Iowa's at the moment (though the head-to-head result is irrelevant - the Selection Committee never seems to take that into account), but Iowa is the significantly better team, so it's not unreasonable to think that Iowa's finish will be better than Iowa State's finish.

David Mann said...

Thanks for taking Mizzou out. Honestly ,I think most of our fans think it's a relief to finally give up hope of an at-large. What a painful season.

Derek Kramer said...

The argument between what team is better than another is mostly subjective. Your assessment that Iowa is the "significantly" better team is your opinion, but an opinion in the minority. Bracket Matrix had Iowa St. as a 3.65 seed average, Iowa at 5.22. Instead of doing a subjective test, let's do a objective test, just like the committee will do. So bear with me as this may run on a apologies.
According to ESPN, Iowa States RPI is 12th with a 8-4 record against the RPI top 50. Iowa is 34th and 4-7, respectively. The only teams with more top 50 wins than Iowa St. are Kansas and Arizona. Iowa St. SOS is 35 while Iowa's is 54. Now, looking at KenPom and Sagarin ratings, Iowa does have an advantage in those. KenPom is much more of a rating of efficiency, and Iowa is quite efficient thanks to their transition scoring. Iowa St. is downgraded due to their questionable defense at times, but is partly offset due to its high scoring. Sagarin mostly measures how bad you beat your opponent by. Iowa had a soft Non-con schedule (233) where they were whopping said opponents. Iowa St. had a little more difficult Non-con (107) and didn't wax near as many teams. The arguement over which the Big 12 or B1G conference is better, is virtually a tie (with a slight edge to the Big 12 according to Conf. RPI). I may be a bit bias when it comes to Iowa St., but these numbers have no bias.

Jeff said...

You're confusing "better team" with "better resume at the moment". All of your statistics had to do with who has the better resume at the moment (things like their RPI ratings).

If Iowa and Iowa State played on a neutral court tomorrow, Iowa would be a 2-3 point favorite in Vegas. That's what I mean by "who is the better team".

Anonymous said...

Complete hypothetical. If MD were to win out in the ACC regular season finishing 11-7, which would include wins over Syracuse, @ Clemson, and UVA, would they be on the bubble heading into the ACC tourney?

Jeff said...

Sure. If Maryland was 11-7 in ACC play (and with a corresponding RPI somewhere around 45th) there's no question that they'd be on the bubble. They might even be in the Field of 68 outright.

But the odds of Maryland winning their final 4 games are pretty low. Depending on which computer rating you prefer, the odds of Maryland winning their last four games are around 4-5%.

Anonymous said...

Thanks for the quick response. Just think how many close games they've lost this season that cost them a tourney bid. UConn, Pitt, @ N.C. St., @ UVA, and @ Duke. Mark Turgeon has not proven to be able to dial up the right plays down the stretch to get this team over the hump.

Jeff said...

Yeah, there's no question that Maryland is good enough that they could have been a Tournament team if they'd had better luck in close games.

Sburke said...

If NE wins at IL, at IN, and v NW but loses the finale home v Wisc, what do you rate their at large chances? I know a lot of moving pieces go into that, so assume no (or 1) stolen bids.

Jeff said...

Well, it's hard to assume no stolen bids, since there are always stolen bids. But an 11-7 Nebraska team will be right on the bubble. It'll probably come down to their Big Ten tourney performance. A run to the semifinals would probably be enough to get them in. A one-and-done performance would probably knock them to the NIT.

Derek Kramer said...

Okay, using this logic most of us could agree that Witchia St. wouldn't be favored on a neutral court against top teams in the power conferences. But they will probably be the 2nd overall 1seed, thanks to a resume that has no losses. You're a "bracketologist", not a bookmaker. Point is teams get their seed thanks to resumes, not potential point spreads.

Jeff said...

Derek, you're mixing up different things. I'm very clear about (and talk all the time about) the fact that the Selection Committee doesn't care how good teams are.

The fact that Wichita State probably isn't one of the nation's 10 best teams doesn't change the fact that they'll be a 1 seed if they win out.

But I'm talking about using team quality to project future performance. Who is more likely to play better down the stretch? Who is more likely to make a run in their conference tournament? To me, the answer to both of those questions is Iowa.

Derek Kramer said...

Let me say, I do respect you and your blog. I agree with most of your opinions and where you have teams placed, but obviously disagree with others. And I completely agree with how the selection committee doesn't care how good you are. They care about resumes, I've been saying this all along. You compared who's going to play better down the stretch to a head to head match up, which are different.

Iowa will probably be favored to win 4/5 remaining games, 3 on the road, against 2 tourney teams most likely. Iowa St. will probably be favored in 2-3/4 games, 2 on the road, against all bubble teams. Iowa's road is bit easier IMO, so they could go 4-1. Iowa St. may go 3-1 or 2-2. Is that a big enough of a difference to determine who deserves a better seed? Maybe, maybe not. Then there needs to be a major difference in their conference tourneys, and I give both teams equal chances in their conference tourneys due to match ups that are hard to determine now.
I do respect you and your blog. I agree with most of your opinions and where you have teams placed, but obviously disagree with others.

corey said...

I think my biggest issue lies with using the efficiency ratings to not only predict the future outcome of games, but then project seedings based upon those outcome probabilities. It seems some teams within Kenpom, who are highly rated, get a pass when they lose a game that they shouldn't...simply based on their high efficiency ratings. The reasoning is then passed off that "they had a bad shooting game" or their opponent "shot the ball above their average"...while not taking into account why they were beaten. Their seeding then may not be influenced like it should because they still have a high efficiency rating.

Kenpom seems to have "lagging indicators" or a delayed response to actual results. I use Oklahoma State as an example. They are still rated in the top 30 despite losing 7 of 8 (4 of those with Marcus Smart). At some point, they eye test needs to taken into consideration to a degree and these advanced efficiency/metric rankings need to be taken with a grain of salt. And this isn't a diatribe to belittle Oklahoma State, it is more an issue that I have with these rankings and the weaknesses associated with them. I see a team four spots behind them (Oklahoma) who has outperformed them as a whole over the season....and also beaten Oklahoma State twice in the process. Still, according to Kenpom, this would be around a pick'em game on a neutral floor. Therein lies the problem...a team with a better resume (Oklahoma) and had also beaten the aforementioned (Oklahoma State) team twice would still not be favored against them on a neutral floor...according to these rankings. Sagarin incorporates results/scores into his rankings so they are a bit more accurate, yet far from perfect.

My point is that if you go by computers only, then the "x" factor of a "team's innate ability to win a game" gets diminished. This is a tough aspect to account for quantitatively, but it certainly needs to be part of the formula. Just my two cents. I enjoy your blog.

Jeff said...

I don't believe teams have magical "just know how to win the game" powers, Corey.

If a team loses or wins because of a really fluky shooting day or some other weird quirk, I will tend to mention it, but there's no bias on whether they are highly rated teams or not. I'll mention it when I see it.