Sports News

Loading...

Tuesday, March 15, 2011

Odds and thoughts on the 2011 NCAA Tournament

Last year I composed a series of convoluted posts detailing each team's odds of winning their respective conference tournaments, then composed a full set of probabilities for each NCAA Tournament team to make the Final Four. Despite a little rust on the prob/stat end due to my theatre exploits, I jogged my memory and quickly whipped out a full set of probabilities, adjusting for the new wonky 68 team format.

The new at-large play-in format doesn't adversely affect those teams' already dim chances of making the Sweet 16 or Final Four. We're talking about 11 and 12 seeds here, after all. These teams are already facing long odds of getting to the Final Four. At worst, playing in cuts those chances by 70%, and if so that team's probably not good enough to project getting far anyway. More than likely, a worthy Cinderella candidate only sees their chances drop by 1/3... for example a 12% shot to make the Final Four becomes a 9% shot... not a huge difference.

In the distant past I've gone over the formula for determining each team's odds of winning individual games based on Sagarin ratings, and determining their odds of reaching certain points in the tournament based on Markov formulas. I won't bore you now with the details but rest assured this isn't just a bunch of numbers and ideas I pulled out of my ass.

So now, here are the four regions and odds for each team in the region:

EAST

The only region with two play in games has a lot of wonky stuff going on aside from the bonus action, but this is pretty much Ohio State's region for the taking.

Dark horse: Washington (7). I'd say Kentucky at #4 but they'll run into consensus #1 team Ohio State in the Sweet 16, easily the team to beat this year, and Orange would be a 2 to 1 dog in that matchup: I don't foresee an upset. UW, however, is a deserving but dangerous 7 seed that, for all their ups and downs (and legal trouble) this year, can play with just about anyone in the country. If they faced overseeded #2 UNC in round two I'd actually have the Huskies as the slight favorite, and think #3 Syracuse would give them a tougher but winnable challenge in the Sweet 16. Don't be too shocked if Ohio State ends up facing UW in the regional final. I won't say the Huskies would win... but they could certainly give the Buckeyes a scare at the least, and scoring the upset isn't impossible.

Overseeded: North Carolina (2) and Xavier (6). Reputation's pretty much carrying UNC in a down year: They're more like a 4 this year. Three teams below them (Syracuse, Kentucky, Washington) would be favorites to beat them on a neutral floor. Roy Williams' team would be fortunate to make the 2nd weekend, dad gummit. Xavier is more of a double digit seed, such a bad misseeding by the committee that they may have table-set a 1st round upset (more in a bit).

Underseeded: Clemson (12) and Marquette (11). Clemson did not have the best year but they should not be playing in when they're better than at least 4-5 other at-large teams, if not more. Marquette also got unduly punished, and probably deserved a single digit seed. 1st round opponent Xavier is similarly mismatched, and don't be surprised if Marquette sends Xavier packing.

No chance in hell: Neither 16 seed poses any sort of credible threat. Texas-San Antonio's odds at the Final Four are suitably long at 430,000 to 1, but the odds for Alabama State, easily the weakest team in this 68 team field, are so laughably long they make Powerball look like a better bet: 69,000,000 to 1. That is not an exaggeration. Their odds of getting to the round of 32 alone are 612 to 1 and their odds of making the 2nd weekend are roughly 23,800 to 1.

You have a better chance of... dying in a tsunami (condolences, Japan). In fact, you're more than ten times more likely to die randomly in a tsunami (615,488 to 1). Those odds obviously go up if you live on coastline in a fault zone, but still.

At least their odds are better than the odds of winning Powerball: 195,000,000 to 1.

Odds:

1. Ohio State. Sweet 16: 79.3%. Final Four: 1.4 to 1 (41.5%)
2. North Carolina. Sweet 16: 46.2%. Final Four: 9.5 to 1
3. Syracuse. Sweet 16: 59.4%. Final Four: 7.2 to 1
4. Kentucky. Sweet 16: 59.4%. Final Four: 6.0 to 1
5. West Virginia. Sweet 16: 26.4%. Final Four: 27.3 to 1
6. Xavier. Sweet 16: 14.6%. Final Four: 96.9 to 1
7. Washington. Sweet 16: 45.1%. Final Four: 8.3 to 1
8. George Mason. Sweet 16: 7.8%. Final Four: 96.7 to 1
9. Villanova. Sweet 16: 12.8%. Final Four: 42.4 to 1
10. Georgia. Sweet 16: 6.9%. Final Four: 271 to 1
11. Marquette. Sweet 16: 23.7%. Final Four: 38.9 to 1
12. UAB. Sweet 16: 2.8%. Final Four: 861 to 1
12. Clemson. Sweet 16: 8.8%. Final Four: 131 to 1
13. Princeton. Sweet 16: 2.6%. Final Four: 2488 to 1
14. Indiana State. Sweet 16: 2.3%. Final Four: 4125 to 1
15. Long Island. Sweet 16: 1.8%. Final Four: 4331 to 1
16. Texas San Antonio. Sweet 16: 0.1%. Final Four: 430,000 to 1
16. Alabama State. Sweet 16: LOL. Final Four: 69,000,000 to 1

WEST

A more conventional 16 team region also has a conventional favorite, as Duke like OSU is facing considerable odds (41%) of making the Final Four. Few teams here can pose a serious challenge for them.

Dark Horse: San Diego State (2) and Texas (4). Silly to cite two high seeds as dark horses, but these are pretty much the best teams in the field not named Duke and the only real challengers to the Blue Devils. Texas is only a 60-40 dog to Duke and if they pulled the upset they'd be the favorite against every possible opponent. SDSU would be a 2 to 1 dog vs Duke in the Elite Eight and a 56-44 dog to Texas if the Longhorns pulled the upset but would be the favorite against anyone else. The odds for either aren't terrific, around 9 to 2 or 5 to 1, but they have the most realistic chance of everyone else to pull it off.

Overseeded: Tennessee (9) and Memphis (12). The Volunteers probably deserved a double digit seed, while Memphis' reputational license has long since expired. They won their conference tourney to get in, but they pack little more punch than your run of the mill mid-major these days, and probably belongs in the 13-14 range. Don't count on a 5-12 upset when Memphis meets legit 5 seed Arizona, while Tennessee has a reasonable shot at a round one win but stands little chance against Duke in the 32-round.

Underseeded: Texas (4) and Missouri (11). Texas has the strength of a 2, and if they meet Duke in the Sweet 16 as expected they will likely pose the toughest challenge of the bracket to the Blue Devils. Missouri probably warrants a middle seed more than a bubble seed, and likely no one will notice because Cincinnati at 6 is pretty good and is a slight 54-46 favorite to dispatch the Tigers, who deserved better.

No chance in hell: 16 seed Hampton is already a 50 to 1 dog to beat Duke in round one. Their odds of going all the way to the Final Four? A paltry 100,000 to 1.

You have a better chance of... sinking a hole in one from 150 yards (80,000 to 1). Maybe Hampton should bag the tourney and join the golf team for a relaxing weekend that doesn't involve getting crushed by 30 points against a Duke team running at half speed.

Odds:

1. Duke. Sweet 16: 82.8%. Final Four: 1.4 to 1 (41.2%)
2. San Diego State. Sweet 16: 65.3%. Final Four: 4.8 to 1
3. Connecticut. Sweet 16: 45.4%. Final Four: 12.4 to 1
4. Texas. Sweet 16: 59.9%. Final Four: 4.6 to 1
5. Arizona. Sweet 16: 30.2%. Final Four: 21.9 to 1
6. Cincinnati. Sweet 16: 28.2%. Final Four: 24 to 1
7. Temple. Sweet 16: 18.6%. Final Four: 54.1 to 1
8. Michigan. Sweet 16: 9.7%. Final Four: 90.9 to 1
9. Tennessee. Sweet 16: 7.2%. Final Four: 153 to 1
10. Penn State. Sweet 16: 13.8%. Final Four: 93.7 to 1
11. Missouri. Sweet 16: 22.9%. Final Four: 35.5 to 1
12. Memphis. Sweet 16: 3.5%. Final Four: 1133 to 1
13. Oakland. Sweet 16: 6.4%. Final Four: 315 to 1
14. Bucknell. Sweet 16: 3.4%. Final Four: 1568 to 1
15. Northern Colorado. Sweet 16: 2.4%. Final Four: 2974 to 1
16. Hampton. Sweet 16: 0.3%. Final Four: 100,000 to 1

SOUTHWEST

The odds of someone other than the top seed get a little better here, but not by much,a s top seed Kansas has a 34.6% chance of making the Final Four. However, the challenge is more broad and general, with no specific dark horses posing a threat aside from....

Dark horse: Purdue (3). Purdue might be a touch underseeded but it makes little difference at 3. They are a do-able 62-38 dog against Kansas if they meet in the regional final, and are solid (though typically not dominant) favorites against anyone else in the field, even 2 seed Notre Dame. Their chances of making the Elite Eight are a solid 39.5%, and there's a 47% chance Kansas falls before that point, which would make the Boilermakers a favorite to make the Final Four.

Overseeded: Both 11 seeds in the region's play-in game: USC and VCU. Actually, "shouldn't be in the tournament at all" is a better label. Given the snubs (Colorado, St Mary's, New Mexico), and how low both of these teams rate overall, the fact that either of these teams are playing championship basketball at all is insulting. Neither objectively is close to being a bubble team, and yet here they are. USC is a 69-31 favorite in the game, and a 74-26 dog against a far superior #6 Georgetown team. 12 seed Richmond could wipe the floor with both these play-in jokes.

Underseeded: UNLV (8) and Illinois (9). Both these 1st round opponents should be a couple seeds higher. If not playing each other for the right to get force-fed to Kansas, both would have a very good shot at the 2nd weekend.

No chance in hell: Newcomer St Peter's (14) faces some fairly long odds at 11,000 to 1 thanks to drawing tough Purdue in round one, making them a 13 to 1 dog for the upset. They probably should have been a 15 but debating the low seeds is a quibble. Our 16 in this bracket, Boston U, is facing 53,600 to 1 odds, and 33 to 1 odds of getting past top seeded Kansas. Even given that, the Terriers have the best chances of any 16 seed in the field.

Odds:

1. Kansas. Sweet 16: 72.3%. Final Four: 1.9 to 1 (34.6%)
2. Notre Dame. Sweet 16: 61.8%. Final Four: 6.8 to 1
3. Purdue. Sweet 16: 62.1%. Final Four: 4.1 to 1
4. Louisville. Sweet 16: 56.9%. Final Four: 8.2 to 1
5. Vanderbilt. Sweet 16: 26.0%. Final Four: 33.8 to 1
6. Georgetown. Sweet 16: 23.0%. Final Four: 25.4 to 1
7. Texas A&M. Sweet 16: 18.1%. Final Four: 65.3 to 1
8. UNLV. Sweet 16: 12.9%. Final Four: 37.1 to 1
9. Illinois. Sweet 16: 14.6%. Final Four: 30.1 to 1
10. Florida State. Sweet 16: 18.2%. Final Four: 65.0 to 1
11. USC. Sweet 16: 7.5%. Final Four: 141 to 1
11. VCU. Sweet 16: 1.1%. Final Four: 2911 to 1
12. Richmond. Sweet 16: 14.3%. Final Four: 105 to 1
13. Morehead State. Sweet 16: 2.8%. Final Four: 2818 to 1
14. St Peter's. Sweet 16: 0.9%. Final Four: 11,000 to 1
15. Akron. Sweet 16: 1.9%. Final Four: 5832 to 1
16. Boston U. Sweet 16: 0.3%. Final Four: 53,600 to 1

SOUTHEAST

The weakest (1) seed in Pittsburgh has a somewhat serious challenger and perhaps a couple of dim challengers to their Final Four throne. This region is rather seed, chock full of mid majors, only some of which pose a serious threat one on one, and many are underseeded, forcing them to climb uphill for the right to try and knock off the Panthers. But that said, Pittsburgh's weak-favorite status means there's a 68% chance someone else will represent the Southeast region in the Final Four, with over half the teams having roulette-like odds at doing so.

Dark Horse: BYU (3). Even without dismissed Brandon Davies, the Mormon Cougars have a solid top ten ballclub, and are no worse than a slight underdog to anyone. A couple of breaks and they could sneak past Pitt into the Final Four. Only a lack of dominant strength makes them a somewhat distant shot at 20.7%. Wisconsin (4) is close in strength but is slated to run into Pittsburgh in the Sweet 16 if they get that far.

Overseeded: Florida (2) and UCLA (7). The weakest two seed in the field should be a 5 or even lower. A very weak sub-bracket (UCLA at 7 and a downish Michigan State at 10, and of course the Gators draw the easily beatable 15 seed UCSB in round one) gives them an even money chance to get to the 2nd weekend, but BYU would be a 2 to 1 favorite if they met in the Sweet 16, and even #6 St John's could give them a serious game.

I had UCLA on the bubble, but they got a 7 seed. Ridiculous. They're a 56-44 underdog to Michigan State in the 1st round.

Underseeded: Utah State (12) and Belmont (13). Look, I know mid majors play weak schedules and you have to bear that in mind when they, say, win 30 games comfortably and only lose to top 50 opponents. But weak scheds never stopped the NCAA in previous years from giving Tark's UNLV teams or old powerhouse Memphis teams high seeds. Hell, Princeton teams in the 90's got single digit seeds a few times, and they rarely played anybody during the regular season.

Oh, but they never go far in the tourney anyway, right? The reason you never see these Utah States and Belmonts go far in the tourney is because you keep giving them 12 and 13 seeds and they're forced to play a tough 4 or 5 seed in the first round instead of some middling power conference team that they could beat. Are the powers that be afraid of getting their conferences exposed as overrated in an NCAA Div-I with better parity? Are they afraid of Utah State pulling a Gonzaga and showing they could go deep and that maybe these perennially successful small schools can play with the big boys and deserve respect?

Anyway, watch out for the 5-12 upset special here, as (5) Kansas State is a slight 47-53 underdog to Utah State. Belmont is not so lucky at 13: (4) Wisconsin is the real deal and a 64-36 favorite.

Speaking of Gonzaga, they're not what they used to be, but they're a slight 52-48 favorite at (11) to topple (6) St John's. Expect many upsets in this region. But not from....

No chance in hell: Arkansas Little Rock is a gift... for Pittsburgh, that is, if they can overcome 2 to 1 favorite UNC Asheville in the play in game. UALR is a 173,000 to 1 shot to thread the needle and make the Final Four.

You have a better chance of: Being killed on the job if you're a teacher or nurse (143,000 to 1). Education majors at UALR can both take comfort and weep at these facts.

Odds:

1. Pittsburgh. Sweet 16: 75.6%. Final Four: 2.1 to 1
2. Florida. Sweet 16: 52.4%. Final Four: 9 to 1
3. BYU. Sweet 16: 58.7%. Final Four: 3.8 to 1
4. Wisconsin. Sweet 16: 41.3%. Final Four: 6.9 to 1
5. Kansas State. Sweet 16: 18.2%. Final Four: 33.2 to 1
6. St John's. Sweet 16: 18.1%. Final Four: 31.2 to 1
7. UCLA. Sweet 16: 18.3%. Final Four: 58.9 to 1
8. Butler. Sweet 16: 13.5%. Final Four: 61.7 to 1
9. Old Dominion. Sweet 16: 9.9%. Final Four: 108 to 1
10. Michigan State. Sweet 16: 25.8%. Final Four: 30.6 to 1
11. Gonzaga. Sweet 16: 20.1%. Final Four: 25.9 to 1
12. Utah State. Sweet 16: 22.2%. Final Four: 23.5 to 1
13. Belmont. Sweet 16: 18.2%. Final Four: 29.9 to 1
14. Wofford. Sweet 16: 3.1%. Final Four: 840 to 1
15. UC Santa Barbara. Sweet 16: 3.5%. Final Four: 1649 to 1
16. UNC Asheville. Sweet 16: 0.9%. Final Four: 7729 to 1
16. Ark Little Rock. Sweet 16: 0.1%. Final Four: 173,000 to 1

Stumble Upon Toolbar BallHype: hype it up!

Monday, March 14, 2011

More teams, more delusions: A terrible effort by the NCAA Tournament Selection Committee

The NCAA Tournament Selection Committee's made some questionable choices every year, but even granted the novelty of the expanded 68 team field, this has to be one of the worst efforts I've seen from them in over a decade... as if most of them didn't know anything about college basketball.

By Sagarin rating (to this day one of the best barometers for team strength), if you filter out all the automatic qualifiers then the at-large selections should go down to the 50th rated team. Among those in the top 50, here is who got snubbed:

Virginia Tech (40)
St. Mary's (43)
New Mexico (48)
Colorado (49)
Maryland (50)

Here are the teams rated below that cutoff line that did get in at-large:

Tennessee (51)
Georgia (52)
UAB (53)
USC (62)
VCU (77)

The top three on that last list might be explainable: Bubble decisions can come down to various, sometimes arbitrary factors. But letting USC and VCU in at-large was inexcusable. Those two teams might not even qualify for the top half of an NIT field.

This is some of the worst seeding I've seen in years.

- I'm not a Texas fan but they're worthy of a 2 or a 3, not the 4 seed they actually got, not-as-strong schedule or not.

- And if you're going to dock Texas for a lack of top-25 wins and holding serve on a padded schedule, then why give North Carolina a 2 seed despite a 4-seed caliber season and the same type of resume?

- Georgetown getting a 6 despite going 21-10 against arguably the toughest schedule in the nation (18 of their 31 games against top 50 teams, no less) is ridiculous. Give them at least a 5, if not a 4.

- Speaking of ridiculous, Florida at 2 despite a relatively easy schedule, only two top 25 wins and four losses versus teams outside the top 50. They're a 5.

- Call me a Vegas homer, but UNLV got ripped off with an 8 seed despite finishing 3rd in a tough Mountain West conference with two top 10 teams (BYU, San Diego State) and a shoulda-been-in New Mexico team while going 24-8 with one of the nation's tougher non-conference schedules. Give them a 6 if you don't like them but they could've warranted a 5... not an 8. An 8 is what you give a 19-13 power conference team or a decent power team that built a 23-26 win season on an easy schedule. It's about time the Mountain West's better teams started getting more than mid-major respect.

- Along those lines, when is the committee going to stop treating Utah State like a small school and giving them 12 seeds when they're clearly a top 25 team? They didn't even do that to the 80's-90's UNLV teams, who played in the weaksauce Big West... and the NCAA HATED Jerry Tarkanian! Utah State this year was a 6 for sure, even with their weak schedule.

- I'm more of a body-of-work guy than a reward-the-hot-finish guy, so I'm not a fan of Villanova at 9 even though they did kind of collapse to close the season. Thing is... even if you're hot or cold at season's end, the NCAA Tourney is a whole new environment, whole new opponents, whole different situation, and when you put 18-22 year old kids in a new situation like that, it can change the perspective of even the most resilient kid and change their game in an instant. Once the nominal season ends and we hit the do-or-die nationally-staged portion of the season, the hot and cold streaks don't really matter anymore because whatever environment you entrenched your groove or rut in is wiped out. Villanova had a decent season overall and keep in mind their cold run came during conference play in the nation's toughest conference (Big East). They still had eight top-50 wins and went 21-11 against one of the 20 toughest schedules in the country. Hit the reset button on their year and I bet things change. And I bet they're not only going to beat George Mason, but they could give #1 seed Ohio State problems this weekend. Yeah, the collapse still happened, but I would've given them a 6 or 7.

- Kansas State is overseeded at 5. Objectively they're maybe a 6, likely a 7. You could even make a case they're more of an 8, but I've give their tough schedule and 21-10 record the benefit of the doubt. And go figure their 12-seed opponent in the 64 round is none other than Utah State. Possible upset.

- Missouri and Marquette at 11 is kind of ridiculous. Both are top 30 teams... yes, even 20-14 Marquette. Missouri didn't fare well vs what top 25 talent they faced but they had four wins vs numbers 26-50 and pretty much held serve otherwise over a meaty schedule in a 22-10 season. Marquette faced 18 top-50 teams in their 34 game season, and beat five top 25 teams. Nearly all of their 14 losses came against the top 50, and bear in mind in this field the committee would have ideally let in everyone in the top 50. Sure, I can see punishing Marquette's 14 losses, and sure I can see punishing Missouri's lack of big wins. But I find that petty here: Both had tough schedules and there's no reason either one wouldn't warrant an 8 or a 9 in this field.

- One 13 loss team that should've gotten the 10-12 seed treatment is Michigan, who did not beat a single top 25 team and though they beat six teams rated 26-50, they took a dive versus four teams outside the top 50. I'd still give them a 10, but certainly not the 8 seed they got.

- Vanderbilt got their annual overseeding gift from the tournament committee. This year's prize: A 5 seed in an 8-seed worthy 23-10 season with a dodgy non-con schedule, only two top 25 wins, and four losses against teams outside the top 50. Seriously, I wonder if gifts from the Vanderbilt AD to the Tournament Selection Committee are part of Vanderbilt University's annual budget.

- Belmont at 13 is stupid. Look at their track record: 30 mostly convincing wins (yes, they play in the lowly Atlantic Sun, but several of those teams had good years), hanging tough in two losses to tourney team Tennessee and in another loss to tourney-bound Vanderbilt, their only real blip being a road loss to Lipscomb in late January. With a lack of convincing majors in the field this year, give them a 9 or a 10 and see if they can give one of those meh majors a scare. Or at least reward the strong year with an 11 or 12. Don't give them a 13 like they're some garbage auto-qualified small school with 8-10 losses as if they got blown out by every top 25 team they faced, or like they lost several games against crappy no-name schools.

- Xavier and their zero top-25 wins is so overseeded at 6 that the committee clearly table-set an upset loss to underseeded 11 seed Marquette. Conversely, Richmond kind of got the shaft at 12, though with a padded sched I kind of see it. Both should have been a 9 or 10 IMO. If you're going to shaft Richmond with a 12 you should have done the same for Xavier given they've had mostly identical seasons. Why unduly reward one while punishing the other? Because we recognize Xavier on a TV set more than we recognize Richmond? That's not how you're supposed to seed teams.

- Old Dominion at 9 isn't terrible but they're easily the weakest 9 in the field and if they weren't facing a similarly weak 8 in Butler, ODU would get crushed. Both are riding reputations from previous seasons: I'd have Butler at 12 this year (they went 22-9 off a pretty weak schedule), and while I might make ODU a 10, I could see seeding them as low as 12: They won 27 games off a somewhat weak schedule and didn't look as convincing as Belmont did with their weak sched.

- I don't like making Clemson play-in: I have eight tournament teams behind them, four of which incidentally got wrongfully snubbed. They're at least an 11. They didn't look too impressive against tougher teams but they still won 21 games against a decent schedule.

- If you're going to punish 13-14 loss teams at all, then why give Penn State a 10 despite 14 losses? Because they had a hot run in the Big Ten Tourney? Again, not a fan of rewarding hot streaks, because then you get 3-4 days off and have to play a new, tough opponent on a new, neutral, nationally televised court in a do or die game... which changes everything. Yes, I like their very tough schedule, but if you reward that for a team while ignoring 13-14 losses, then don't go punishing Marquette.

- UCLA at 7 makes little sense. Yes, three top 25 wins in a 22-10 season, but they had a pretty weak schedule for a Pac 10 team (only eight top-50 opponents in all), didn't look particularly convincing in many of their wins, and lost five winnable games to teams outside the top 50. I'd have them at 12 and would have had them play-in.

******

Bear in mind... I'm only hitting the clear, distinct mistakes. Poring through the teams and ratings, there was so much else to cover that, if I could see a couple of plausible reasons to overlook a seeding mistake, maybe a particularly strong/weak schedule or a lot more losses/wins than usual or something, I let it go if it was just 1-2 seeds over or under. This post could have been twice as long if I hit everything.

Well done, Selection Committee. Way to use three extra at-large bids as an excuse to turn completely stupid.

Stumble Upon Toolbar BallHype: hype it up!