Breaking The Brackets
Tuesday, March 17, 2009 11:40
Last year, a hot chick I know won 1500.00 in her NCAA Basketball tournament office pool. Guys who lost in earlier rounds were more pissed than a Fenway Park urinal. Their anger didn’t stem from her gender—although that did rankle their ego—but from her methodology, which went something like this:
“Bruins? I love those cute little bears… And I just luvvvvvvvv cats, Wildcats, and Tigers… Ooh, I had a neighbor named Blair Davidson. He had really pretty eyes. Jay Hawks? What’s a Jay Hawk, anyway? My cousin lives in Kansas and she sent me a T-shirt. I’ll pick them.”
Yes, she did pick three out of the Final Four teams.
The reason her kind of approach gets so much attention when it succeeds is because, in reality, that type of method rarely wins—but it sure does make “bracketology” fun.
Let’s try to do a little more than throw up our hands to the hot chick system and see what the numbers tell us…
- Extreme scenarios do not happen. Never will a 64-team bracket play out with every higher seeded team winning every game until the four top seeds reach the Final Four. Nor will two 16 seeds play in the title game.
- Never has a top seed lost to a sixteen seed—although the odds say it should have happened by now. But, in the first round there is over a 50% chance a 2, 3, or 4 will get beat. At least one 2-6 seed WILL be upset in the first round during any given tournament.
- One or two 11 or 12 seeds will advance to the Sweet 16 in a typical 64-team bracket.
- Usually, the teams are seeded as accurately as human management will allow, which means there will be some serious, even laughable mistakes.
By now you are asking, “So Mr. Smarty Pants, how do we apply your axioms to this year’s bracket?”
Let’s take a gander at the First Round, region by region…
FIRST ROUND
MIDWEST REGION
Sleeper Special – Dayton over West Virginia…Other than that little gem, I see the region playing true to form. The only other upset might come from Arizona over the Utes. The Wildcats are limping into the tourney and have been inconsistent all year.
THE WILD WEST
Maryland has the best shot for the Upset Special. The ACC is tougher and more demanding than the Pac-10. Cal leans on the outside game; the Terps force things inside and have the personnel to do it well.
Same call for Utah State. Marquette has been inconsistent and their recent performances during the Madness have been pedestrian.
DOWN SOUTH
Chester Frazier’s hand injury will make the Illini’s task more difficult and the coolly named WKU Hilltoppers showed what they could do as a high seed last year. I am going against the prevailing sentiment and sticking with Illinois.
Look for the surprise when Temple bottles up Arizona State. The Owls had a stellar A-10 tourney; the Sun Devils play defense with anyone. Should be a low scoring nail biter.
IN THE EAST
I just hope Pittsburgh doesn’t play like they did in the Big East tourney. They should approach it like a tough workday and take care of business. The only upset I can see unfolding is the very close to home VCU “hosting” the UCLA Bruins. However, Ben Howland is battle tested and has the troops to overcome any obstacle.
So, Mr. Smarty Pants has spoken with his usual wisdom, confidence, and vast knowledge…
Then again, he knows for a fact the that Blue Devil mascot is totally cute… Oh yeah, and Elvis used to live in Memphis—how can you pick against “The King?”… And Orange is his favorite color…
NEXT WEEK: ROUND TWO
Nate Barlow
says:
March 17th, 2009 at 11:50 am
We could not be on more of the same page in regards to upsets… VCU, Utah State, Temple, Western Kentucky, Dayton (although I don't see Arizona pulling it off).
Would love to see a full-blown statistical analysis of what you could expect round-by-round in terms of upsets.
Gairzo
says:
March 17th, 2009 at 12:43 pm
I plan on doing that next column. If you really look at it though, the field is twice what it really should be. Past the second round there are less true upsets than those among 64 or 32 teams.
I wonder if you could help me decipher the brackets, real quick–anyone–Let's take the West region as an example. Say UCONN wins their first round game and Cornell upsets Mizzou; everything else in that region holds true to form. Does that then mean UCONN will play Cornell in the second round? The way the brackets are laid out it seems the winner of BYU vs Texas A & M would play UCONN.
I feel ignorant asking, but until I began writing for DIS, I filled out my bracket without bothering to study it. I had thought the highest seed always played the remaining lowest seed in a given region.
It seems, given how teams in one region may play at a totally different geographical locale, that isn't the case.
Somebody educate me, please.
Nate Barlow
says:
March 17th, 2009 at 12:56 pm
UConn will play BYU or Texas A&M regardless of what else happens in the first round (save if the Huskies lose themselves, but that's highly unlikely). The NCAA Tournament is unlike the Stanley Cup Playoffs; there is no reseeding after each round.
Each successive round there are less true upsets, unless lower seeds have already been able to pull it off. As long as everything holds to form, each round the teams are more equal.
ernessa
says:
March 17th, 2009 at 5:35 pm
Okay, I didn't understand half of what you said in that column and I kind of hate March Madness, b/c I just really don't like college basketball, but I do love entering MM office pools. I usually pick according to which state and/or city I like best. This is why Indiana will never be found on my form even after a winning season and why the Missouri Tigers always will. Plus, the Missouri Tigers have that really cute paw print. Who can resist voting for that?
Gairzo
says:
March 18th, 2009 at 12:13 pm
Ernessa, one of the great qualities of your writing–besides your unique voice–is your honesty. You "get" it even when you say you don't.
The point of the column was to highlight the "really cute paw print" approach with the "scientific" and statistical approach–both of which are a crapshoot.
I mean some of these MM websites do all these statistical analysis with the attitude that "if you follow our formula, you'll win your office pool" While we can determine statistical patterns or tendencies–like a 2-6 seeded team WILL lose a first round game–the trick is picking which one of those 20 game(s) will result in that upset.
They forget to tell you, even if you do pick one or more first round upsets, the point values are minimal for picking first round winners AND it doesn't mean you'll get a significant percentage of the other 44 games right.
We ought to all submit brackets. I'll forward an e-mail my friend sent. For 20 bucks, you complete a bracket and have a chance to win some cash.
Gary's Brain
says:
March 18th, 2009 at 10:35 pm
Excuse me…excuse me, Gairzo, but I'm betting Math was not your strong suit.
In a 16 team region the 2,3,4,5,6 seeds need five opponents: the 15, 14,13,12,11, respectively. The remaining 6 teams play three games. The above scintillating comment makes sense only if the readers replace the number "44" with "12"…