2014 OFFICIAL NCAA Tournament Thread: Connecticut wins the title

[Matthean]

Hoosiers AD: 'We're Indiana. We don't play in the CBI' - CBSSports.com

[Butter]

Indiana has some of the best fans in college basketball, but also some of the most unrealistic. I can't remember an IU foul call in Assembly Hall that's gone un-booed, and some IU fans are already sick of Tom Crean after missing the tournament with a nearly all Freshman and Sophomore team after 2 consecutive Sweet 16's.

[Butter]

Looks like Bruce Pearl is about to latch back onto coaching at Auburn.

[Butter]

Since Pearl has a show-cause until August, how is that going to work?

[Chief Rum]

Originally Posted by Butter_of_69

Since Pearl has a show-cause until August, how is that going to work?

What does he need to do between now and August? He can hire assistant coaches who can recruit and then in August he's back on the job.

[Butter]

Seems like given his track record, people will be more than thrilled to turn him in if he slips up even a tiny bit.

[britrock88]

I thought his show-cause was perceived as an NCAA overreach. Wasn't this for making burgers for Aaron Craft, and then lying about it? He knows better than to lie about it now.

[nol]

Here's something if you want to have a fun bracket to cheer for. Take Buffett's Billion

Basically you'll be randomly assigned one of the 14,000 or so statistically most likely brackets and split the winnings if any of them happen to win.

[SportsDino]

Even if you could chalk it down to sweet 16 you would have 32,000+ combos which would all be reasonably likely. Predicting the upsets in the first two rounds blows the numbers out of reach... Although for fun I think I will calculate the pool for a 15 million bracket strategy which is a little more sporting.

[molson]

Two random thoughts on the Buffett thing.

We have 15+ years now of online brackets - does anybody know what the "best" documented bracket ever was? I remember looking at some close-to-perfect brackets over the years.

I know it's extremely unlikely, but imagine if someone had a perfect bracket on the eve of the championship game, or before the final four. If it were me, I'd either try to make a deal with Buffet to hedge my bets, or try to find some legal loophole to encourage assistance from the team I picked against. It'd be chaos. I mean really, both teams could be on the take in some manner, there would be a billion dollars at stake.

[nol]

Originally Posted by SportsDino

Even if you could chalk it down to sweet 16 you would have 32,000+ combos which would all be reasonably likely. Predicting the upsets in the first two rounds blows the numbers out of reach... Although for fun I think I will calculate the pool for a 15 million bracket strategy which is a little more sporting.

And if choosing from the ~20,000 most likely scenarios over the whole tournament narrows the odds down to 1 in 32,000 and everyone picks a different one of those 20,000, the odds of one of them winning is almost 50-50.

[sterlingice]

Originally Posted by nol

And if choosing from the ~20,000 most likely scenarios over the whole tournament narrows the odds down to 1 in 32,000 and everyone picks a different one of those 20,000, the odds of one of them winning is almost 50-50.

Except that only assumes chalk for the Sweet 16, which I don't think has ever happened.

SI

[nol]

Originally Posted by sterlingice

Except that only assumes chalk for the Sweet 16, which I don't think has ever happened.

SI

I just said 1 in 32,000 as a ballpark estimate for the cumulative probability of the 20,000 most likely unique scenarios for the entire tournament (e.g. if you had the advantage of getting 32,000 brackets to fill out right before the Sweet 16 begins and got an automatic 48/48 for the first two rounds, you wouldn't want to waste a bracket on one where the underdog wins each of the remaining 15 games, since chalk until the Sweet 16 followed by all upsets is probably less likely than a more "normal-looking" bracket with something crazy like UVA getting bounced in the first round thrown in.) to illustrate that if a large enough number of people worked together to try to cover all bases, it places the odds of at least somebody getting all of the games right in more imaginable terms, even if it's just in the 1-5% range. Much smaller version of the monkeys producing Shakespeare

[Butter]

Hooray for NC State!

[Logan]

Originally Posted by molson

Two random thoughts on the Buffett thing.

We have 15+ years now of online brackets - does anybody know what the "best" documented bracket ever was? I remember looking at some close-to-perfect brackets over the years.

I know it's extremely unlikely, but imagine if someone had a perfect bracket on the eve of the championship game, or before the final four. If it were me, I'd either try to make a deal with Buffet to hedge my bets, or try to find some legal loophole to encourage assistance from the team I picked against. It'd be chaos. I mean really, both teams could be on the take in some manner, there would be a billion dollars at stake.

Not sure on your first question, but I heard some discussion on Mike and Mike the other day about approaching it as a "deal or no deal" type scenario. If you were perfect through the Final Four, how much would you need in order to walk away from the chance of that money? You still need three more wins, but obviously those three wins were much easier to forecast than all the ones that came before them.

I believe they also played a clip from Buffet that said he would consider buying someone out if they got close.

[Butter]

$20 million? Fifty?

[SportsDino]

As a math problem I am calculating a 15 million pool strategy, basically eliminating combos from the pool to get a probability of the entire pool losing I could live with that has 15 million combos or less. The same concept would be applied to a 15,000 sample, the key is not ruling out upset combos that I think a pure probability model would fail at. Essentially following the probabilities under values early round upsets that could be the missing component of an otherwise winning bracket. 14,000 samples of near chalk generated brackets would be effectively useless, first upset kills the entire pool.

As for a buy out you have to take any life changing amount offered. If you would not go to Vegas and lay down your life savings on one bet you should not do the same even if your cost to play was zero. The banked offer becomes your life savings that you are gambling for the billion.

[Logan]

I'd gladly take $10MM after tax, especially if it meant my name not getting out. A billion is so much money it might as well be $100 billion. Knowing I missed out on banking all that money because some prick from Florida missed a free throw while I was chasing an unspendable sum of money would probably drive me insane and into a depression for the rest of my life.

[sterlingice]

How often has a perfect bracket survived the opening weekend? Ever?

Average NCAA Tournament upsets by year | Putting Out the Vibe

Even the "chalkiest" year in 2007 produced 10 upsets in the first weekend.

SI

[Matthean]

Obama picks MSU. He picked Indiana last year so hopefully the Spartans last longer.

[Butter]

Originally Posted by Logan

IKnowing I missed out on banking all that money because some prick from Florida missed a free throw while I was chasing an unspendable sum of money would probably drive me insane and into a depression for the rest of my life.

Pretty much what I was thinking. I might begin to get homicidal over some missed Adrian Payne box-out, and that's not good for anyone. Least of all, Adrian Payne.

[nol]

Originally Posted by sterlingice

How often has a perfect bracket survived the opening weekend? Ever?

Average NCAA Tournament upsets by year | Putting Out the Vibe

Even the "chalkiest" year in 2007 produced 10 upsets in the first weekend.

SI

The most accurate bracket I could find was 58 of 63 in 2007, and they had all the Final Four teams. The the biggest first round upsets were two 11s over 6s (VCU and Winthrop, which were both pretty strong mid-majors getting a decent amount of buzz beforehand), and the only non 1 vs. 2 matchup in the Elite 8 was 1 vs. 3 (the two seed that didn't make it, Wisconsin, was kind of a sexy pick to get upset). If nobody was perfect through two rounds that year, I doubt it's happened.

[cmp]

So everyone is picking Florida, Arizona, Michigan State, Louisville I guess?

[whomario]

just a nice little look back to last years championship game with 2 of the main protagonists, Spike Albrecht and Luke Hancock : The Spike and Luke Show: Redux «

[britrock88]

Yahoo brackets close at 1am Eastern, FYI.

[Logan]

How A UVa Fan Bluffed His Way Into The Huddle At The ACC Title Game

[sterlingice]

My look at NCAA tourney bracket pick psychology this year:

South
* Florida doesn't seem to have a challenge in the top half of the bracket. It's really hard to find someone who has them losing anything short of the Elite 8.
* It seems like everyone has Pittsburgh over Colorado so they can claim a non-chalk line.
* SFA is a popular upset against VCU in a game that should be really fun to watch. Beware, tho, as 538 has SFA as the least likely 5 to win at under 25%. On the other half of that pod, UCLA isn't getting a lot of love, in general.
* I wish there was a better 6/11 than Ohio State/Dayton here. If there was a 6 or even 11 that I had more confidence in, I would be able to slide them past Syracuse into the Sweet 16. However, in Buffalo and the matchup against the zone being the 2nd game on the weekend, and Western Michigan looking like the wrong profile for an upset- I think you almost have to go Syracuse out of that pod, despite all their recent struggles.
* No one picks a 2 to lose to a 15 but Kansas sans-Embiid is certainly capable of losing that second game. The hard part about the bracket is which of Standford and New Mexico to pick for the upset. New Mexico looks a bit stronger on paper
* In short, it's pretty much Florida unless you're banking on Embiid to make it back and dominating, both long shots. The single biggest key to winning a large bracket pool is right here. If Florida stumbles and you can figure out who makes it out of this regional instead (UCLA? Syracuse? Ohio State?), it might be the single biggest advantage you can have.

West
* Arizona 2nd round is the puzzle here in the top half. Lute Olsen lost to a 12, 13, 14, and 15 in the 90s so there's still some perception of Arizona as a choker even though he's been gone for a while. The bigger problem for upset pickers is which of Oklahoma State, underachiever with tons of talent, or Gozaga, perennial giant killer, to pick. That said, if the Wildcats win their second round game, most people have them through to the Final Four.
* There are pretty much zero expectations for the second quad of this bracket. Whoever emerges is picked to lose in the Elite 8 to whoever comes out of the top, be it Arizona, Oklahoma State, or Gonzaga. Inside there's some bracket intrigue as North Dakota State is a popular 12/5 upset winner over Oklahoma. But while Louisville and Michigan State are 4s expected to make some noise, no one seems to have San Diego State going anywhere, even with that stifling defense.
* This is the quietest bottom half of the bracket in the tournament. Bo Ryan has never made a Final Four and the Badgers are a popular upset pick at the hands of Oregon or Creighton or even the Baylor/Nebraska winner. If you think they can do damage, this would be a shrewd play to get to the Final Four or even deeper.
* Speaking of Oregon, pretty much everyone is banking on them to win against BYU because of the Collinsworth injury.
* Creighton feels a little too high to me as there will be a bad shooting day from 3 and that might sink them. Because of their style and not being a big name, they're popular upset fodder in brackets. That said, they're also a popular upset pick to go far. Unfortunately, because of their defense, I think I'm picking them to bow out too early rather than stay too late.
* Picking the Baylor/Nebraska game makes me uncomfortable. Do I go with ceiling (Baylor) which can disappoint or go with floor (Nebraska). Baylor is a tempting upset pick in the 2nd round, especially being in San Antonio. However, relying on Drew to win 2 games or to do much of anything is dangerous.

Midwest
* Poor Wichita State. Their reward for going undefeated is being the #1 least picked to get to the second weekend and 3rd most popular team in their own bracket to reach the Final Four with Michigan close behind them. Not only do they have Kentucky, a popular boom/bust pick, or Kansas State, an in-state "rivalry" game, waiting for them but almost no one has them making the Final Four with Louisville, Michigan, and Duke looming.
* Louisville is a computer darling, 2nd in KenPom, but hated by ye olde RPI, which has them at 19. That's what happens when you play a pretty bad non-con schedule and are in the American conference. They've won 11 of their last 12 but only 4 of those wins were against other tournament teams and so was the loss. However, they were highly touted coming into the season and boast great offensive and defensive efficiency. It's easy to say they're the safety valve for venge-picking against Wichita State but they're also being touted as criminally underseeded and the most likely bracket winner at 15% on fivethirtyeight.com. It's like basketball hipster heaven. As there are picks of them making it everywhere from the Sweet 16 to winning it all, picking where they lose, if at all, could be the key to a lot of pools.
* St Louis is another popular 5 seed to get upset. And no one has them beating Louisville. I mean, seriously, in Yahoo 8/9s Oklahoma State, Gozaga, and Kentucky are all picked more often to win a Sweet 16 game as are all the 7 seeds except Texas, which is 0.2% below them.
* Let's get the "little" games out of the way as everyone (89%) has Duke and Michigan advancing out of their pods. Only Virginia-Michigan State is even close in terms of bracket expectations. Tennessee is getting a lot of love to beat UMass and is a fairly safe upset pick. I wouldn't be surprised to see the betting line favor Tennessee if they get past Iowa. However, if Iowa wins, I think UMass is favored. Either one is probably playing for the pleasure of getting beat by Duke but Tennessee is getting a little dark horse love.
* Texas-Arizona State is one where either someone tries to pick up some hedged upset points with Arizona State or thinks Texas is the better team. Either way, everyone is treating Michigan to the Sweet 16 like an automatic backstop so there aren't many points bled out either way.
* So, about that Michigan-Duke game. If it happens, it will be one heck of a game to watch as the teams are 2-3 in the country in offense and both triple digits on defense. Kenpom gives an edge to Duke on the rankings with Duke at 6 and Michigan at 15. 15 you say? And both with defenses over 100? Shouldn't that give you pause when you're assuming them automatic to the Sweet 16? If one of the other 6 teams in those two pods wins, a lot of bracket quadrants get broken quick. That said, on 538, there's an 80% chance that one of those two teams gets that Elite 8 slot and the slight edge is to Duke. Choose wisely.

East
* This is like the Midwest lite: take the "automatic" of Michigan-Duke and combine it with the #1 people expect to lose and a #4 that's underrated. Most years, Virginia would be the #1 seed everyone expects to lose as they were the last one on that line. However, Wichita State's unique situation makes them the trendier pick to look smart picking against. So, quietly, Virginia-Michigan State is getting picked in nearly 80% of brackets, which is all but automatic for the Sweet 16. Skipping straight to the main course for the top part of the bracket, Michigan State-Virginia is a 50/50 proposition at 538 but Michigan State is being picked at a 2:1 rate.
* People love Izzo in March and most neutral basketball fans even want him to do well as long as he's not playing their team but Michigan State is getting a hugely disproportional amount of love this year so beware!
* For the undercard matches in the top half, we have Memphis-George Washington and Harvard-Cincinnati. In the former, Memphis is favored by 3, which is surprising, considering how close it looks on paper with the teams at 40 and 41 in kenpom. The latter game is getting the 5/12 love and, along with Saint Louis, 538 has Cincinatti at 42% to lose a 5/12 matchup. I think there will be a lot of brackets that have both Harvard and NC State winning because even if Cincinnati and/or St Louis win, they are facing the juggernaut pair of 4s in Louisville and Michigan State. Because when making out your bracket, being able to pick upsets, even the common 5/12 variety, is found money when it's the early rounds and you're almost certain the team is going to lose the next game anyway.
* The bottom half of the East holds one of the more intriguing teams in the tournament for me in Villanova: 7th in kenpom and 5th in RPI. They lost to Creighton twice and Syracuse once during the regular season and were starting to get some run for a #1 seed until a loss to Seton Hall to start the Big East tournament. The this is only the third time Jay Wright has had a team with more than 25 wins entering the tourney and the last two times he made the Final Four (2009) and Elite Eight (2006). I think they're getting slept on as only 10% of people have them winning the region, below Iowa State (11), Virginia (20), and Michigan State (52). Meanwhile, 538 has Michigan State (24), Virginia (23), and Villanova (21) more closely bunched with Iowa State down at 8%. I think Villanova to the Final Four is a really sneaky play.
* Connecticut is about a 2:1 favorite against St Joseph and 538 gives them a 26% chance to beat Nova. It's not great odds but they're the best of any 7 out there. However, some of that is the product of being the most favored 7 so this is probably not the place to go shopping for a 7/11 upset.
* Providence is an instance where I don't see what the computers see as 538 has this about a 2:1 game rather than the 80:20 I see it is. Fan brackets much more favor Carolina but that probably has a lot more to do with name recognition. North Carolina Central is almost 20% on 538, by far the highest of the 14 seeds (highest of any seed below 12), to beat Iowa State, a program which has had trouble in the first round before. NCC is a good smaller school but I'd have felt better about the upset against a slumping Syracuse or if they had somehow been bumped up to the 12 seed to take a crack at some of those 5s but I think Iowa State is a bad matchup. Still, odd things can happen at those late night first round games. If the seeds hold, Iowa State (16o/54d) is 23 in Kenpom and North Carolina (52o/19d) is 26 so that, along with the blue blood pedigree and the potential 1st round upset, probably explain why the Tar Heels are getting a lot of run as a potential 6 over 3 upset.

SI

[Lathum]

Originally Posted by Logan

How A UVa Fan Bluffed His Way Into The Huddle At The ACC Title Game

great, now the terrorists have a roadmap for murdering Tony Bennet. Thanks Dany.

I bet this kid would never pull that stunt at a South Carolina game.

[tarcone]

Go Iowa!

[Lathum]

You can tell from McCafferys interview he is very emotional right now, curious to see how the team comes out.

[Matthean]

UMass would be beatable as well. Duke would be tough.

[tarcone]

Iowa by 2 at half. Let down at the end of the half. Iowa needs to avoid a 2nd half scoring drought. That is one of their downfalls. Long periods of time without scoring.

[cartman]

Originally Posted by britrock88

Yahoo brackets close at 1am Eastern, FYI.

Yep, 45 minutes to go.

[Julio Riddols]

When I win the billion dollar challenge, I will buy this board an island.

[SportsDino]

Well so far I have a pool that uses only 0.0000015 of the solution space, unfortunately that still amounts to 139 billion members and involves some picks that may be considered aggressive. I did come up with a reasonably nifty way for designing these pools and creating equations to relate the number of combos in the pool and the probability of the bracket solution being inside the pool, basically the product of every one of the assumptions being correct.

All of those assumption probabilities start to bring down the statistical accuracy of the pool very quickly, however, to allow for 2-15 upsets you basically need to make the conscious decision to use up a bracket for that upset AND every combination of all other game decisions that include that upset. This means you could easily add trillions of brackets to your pool just so you can claim coverage over a particular upset.

A probabilistic bracket doesn't help with this too much, if it is giving away 3% of the 2-15 matchups to a 15 seed or even a much smaller weighted amount it is reducing the quantity of brackets covering more likely coin-flip variants like 1/2, 2/3, 8/9.

Anyway the design of my system allows for expert knowledge to be applied, whatever you would term expert knowledge (giant stat computer calculating matchups, historical records, panel of experts, crowd vote).

To get down to the 139 billion pool I made some of the following decisions:

All 1/2/4 seeds win first round.

All 3 seeds always win first round, except Iowa State.

Florida is a lock for the first three games.

Arizona, Virginia, Michigan State, Michigan, Louisville lock the first two games.

The following teams always lose when they make it to the round of 32:
SF Austin, NC Central

The following teams always lose when they make it to the round of 16:
Dayton, Providence, NC State, Kentucky, Kansas State

The following teams always lose when they make it to the round of 8:
Stanford, New Mexico, St Joes, BYU

That rule set is all I have used so far to restrict the set. Even if my particular choices are not very expert the basic concepts are simple, the best ways to reduce the size of the pool are to predict early round guaranteed wins and limitations on how far a particularly team can make it. Both of these should generally be reasonable to predict and estimate the probability within basic basketball knowledge. For instance, the likelihood of a late seed reaching the Final Four is extremely low historically (0.7% of all seeds 8-16).

Each rule costs some probability to use, for instance if every rule had a 5% chance of being broken, by the time you have 10 rules there is a 40% chance your entire pool will be busted by a 5% event.

To get down to 15 million I probably would need to do a pretty good job of predicting the maximum placement of every seed greater than 4. Calling more picks to expire at round 32 is essential, which is probably an easier decision and makes no assumptions about the many coin flips in the first round. It is easier to predict that an 11 will not win twice than to predict it will lose to that 6 seed, so my bracket pool allows for that.

Anyway I have found this an amusing problem, I don't think I am smart enough about the teams this season to make the right picks, but I think I could code this into a web gui so anyone could create an experimental pool.

[sterlingice]

And that further assumes you can get people working together on 15M brackets

SI

[Butter]

Originally Posted by sterlingice

My look at NCAA tourney bracket pick psychology this year:

South
* I wish there was a better 6/11 than Ohio State/Dayton here. If there was a 6 or even 11 that I had more confidence in, I would be able to slide them past Syracuse into the Sweet 16. However, in Buffalo and the matchup against the zone being the 2nd game on the weekend, and Western Michigan looking like the wrong profile for an upset- I think you almost have to go Syracuse out of that pod, despite all their recent struggles.

SI

While we're wishing things, I wish there was a better poster than sterlingice. He is a dumb-dumb head.

GO FLYERS!

[mauchow]

I fell asleep when Iowa was up 3 or 4 later in the game and I said to my wife, Iowa will lose this game by more than a handful.

Man, Iowa was just something else this year.

[sterlingice]

Originally Posted by Butter_of_69

While we're wishing things, I wish there was a better poster than sterlingice. He is a dumb-dumb head.

GO FLYERS!

Being called a dumb-dumb head is one of the kinder things I've ever been called

SI

[Lathum]

some fun stats and memories

http://en.wikipedia.org/wiki/NCAA_Men's_Division_I_Basketball_Championship_upsets

[Lathum]

Originally Posted by mauchow

I fell asleep when Iowa was up 3 or 4 later in the game and I said to my wife, Iowa will lose this game by more than a handful.

Man, Iowa was just something else this year.

Watced the game with one of my best friends, huge Iowa alum and fan. It felt like a horse race where your horse was ahead the whole time but you just know he will get run down in the stretch.

It also solidifies my opinion on the first four nonsense. I asked him if it felt like his team played in the tournament or if they were playing for a spot and he said the later. Sucks for fans of those teams.

[mauchow]

Which is how it is in everyone's eyes except for the bracket saying "1st round".
I am so fuckin pumped up right now. Gotta go to court first to get a garnishment placed on a customer then its off to watch some games. Boom!

[sterlingice]

Originally Posted by mauchow

Which is how it is in everyone's eyes except for the bracket saying "1st round".
I am so fuckin pumped up right now. Gotta go to court first to get a garnishment placed on a customer then its off to watch some games. Boom!

Time for the actual first round today in just about an hour.

SI

[Logan]

Originally Posted by Lathum

Watced the game with one of my best friends, huge Iowa alum and fan. It felt like a horse race where your horse was ahead the whole time but you just know he will get run down in the stretch.

It also solidifies my opinion on the first four nonsense. I asked him if it felt like his team played in the tournament or if they were playing for a spot and he said the later. Sucks for fans of those teams.

I think it sucks much more for fans of the teams who earned an autobid but get stuck in the play in game. If you want to play in the games that people actually care about, win more games during the season.

I don't like the expanded field, but if you're going to have it, I'm all for making the last 8 in play in these games.

[PackerFanatic]

Nice write-up, SI - although just to correct a minor thing, the Badgers did make the Final Four in 2000 (they lost to eventual-champ Michigan State)

[sterlingice]

Originally Posted by PackerFanatic

Nice write-up, SI - although just to correct a minor thing, the Badgers did make the Final Four in 2000 (they lost to eventual-champ Michigan State)

Whoops- that was supposed to be Bo Ryan so it's corrected.

SI

[mauchow]

Wisconsin gets over their drought and 9 pt deficit with a 10 pt halftime lead.

[Kodos]

C'mon, Dayton!

[Lathum]

Originally Posted by Kodos

C'mon, Dayton!

They have started to get real tight the last few minutes. I think the OSU pressure is getting to them. Lots of out of control shots and no plan on offense

[BishopMVP]

Aaron Craft can do no wrong in the eyes of announcers.