Wednesday, March 18, 2009

The Perfect Bracket

Do you know how impossible it is to predict every single game correct in an NCAA tournament bracket? I'm not pretending to be a mathematician, but I believe if you were to go take each of the 63 games as a pure 50/50 chance, you'd be looking at something along the lines of 1/2 to the 63rd power, a 50/50 chance iterated 63 times over. To show the ridiculousness of that impossibility, here's that number: 0.000000000000000000108. If my math (or rather my calculator and my spotty remembrance of the rules of decimals) is correct, that's just over a 1 in one quintillion chance. Yet this is my dream. I figure if I'm around for 75 NCAA tournaments, that gives me a 75 in one quintillion chance. Now that doesn't sound so daunting.

Thankfully I have much greater things that I live for, and my hope doesn't rest in a chance somewhere along the same lines as finding a 4-leaf clover orbiting neatly around Jupiter. But I do entertain the fantasy of one day attaining this Jupiterian clover, this white whale. We'll call it a recreational dream - don't matter a lick if it never happens, but man would it be sweet if it did.

In my quest for this one shining moment I've gathered a basic set of rules for filling out an NCAA tournament bracket, based on observations made through the years. Some of these are the same ones you always hear (like always pick a 12 seed to beat a 5), and for good reason - they're correct. And some are more under the radar. But I believe that by observing some general trends played across several tournaments, coupled with the fact that some games are most decidedly not a 50/50 proposition, you can increase your chances greatly, somewhere along the lines of 1 in a trillion. Now we're getting somewhere!

I should note one thing before I give you the list. If you've played Hearts before, you're familiar with the concept of shooting the moon. Picking the perfect bracket is like shooting the moon - just trying to do it will probably end in disaster. So if you're looking simply to win an office pool, you're probably looking in the wrong place. Invariably what happens is you go by these rules, but pick the wrong upsets, so while your 12 seed is losing by 15, some other 12 seed is knocking out your sleeper 5 seed who you picked to the final 4. But at least you've tried, and you've barked up the right tree. If you want to win an office pool, go with the cowardly "all-favorites" bracket (much like President Obama demonstrated on SportsCenter) - you'll get thrown off by a few upsets (who doesn't), but usually the cream does rise to the top and you'll probably be looking pretty good come final 4 weekend. But where's the fun (and right to brag, for that matter) in that. Anyone can pick all one seeds to the final 4. That's the easy way out. I say go for it all, the elusive perfect bracket.

So without further adieu, here are my 12 bracket guidelines that I try to live by each March (15 would have been nice and consistent, theme-wise, but I'd just be making some of them up). I say guidelines because sometimes you just have to go with your gut and bend the rules. But if you stick by these, you'll be in pretty good shape.

First Round Guidelines
1. Always pick a 12 seed to beat a 5 seed.
This is about the worst-kept secret since the parental identity of Luke Skywalker. Every year everyone tells you that at least one 12 seed will beat a 5 seed, and indeed it does happen just about every year. I think only twice has it failed. But why? The reason, I think, lies in the fact that the 12 seed line is the usual cutoff point between the last at-large teams and the very good, but very underrated, mid-major conference winners. These teams are seeded down this far simply because of the conference they play in. Their opponents are weaker, their strength of schedule goes down, their RPI drops, and they fall to a 12 seed, when in actuality their talent level is on par with an 8 or 9 seed. And so they play the supposed giant-killers each year.

My pick this year: Western Kentucky over Illinois. The Hilltoppers did it last year, why not again? And Illinois goes through incredibly long scoring droughts, enough to kill them in an NCAA tournament game.

2. Always pick an 11 seed to beat a 6 seed.
12's beat 5's more often, but there is almost always an 11 seed that takes down a 6 too. Just don't go crazy and pick more than 2 of these - 6 seeds can be among the most frisky in the tournament, and are attractive sleeper picks for the elite 8 (more on that later).

My pick this year: Utah State over Marquette. Marquette is dead in the water after the injury to Dominic James, and Utah State is one of only 2 thirty game winners entering the tourney. They're the perfect example of what I described above, only the got an 11 instead of a 12.

3. Pick at least 6 first round upsets.
I read yesterday that the average number of higher seeded teams beating lower seeded ones in the first round is 7. I usually land somewhere between 7 and 9. I should note that the term upset is used loosely - 9's beating 8's and 10's beating 7's are hardly upsets, but they are higher seeds winning. 

My number this year: 8

4. Don't take any team seeded 14 or higher unless you're REALLY confident in them, and I mean REALLY confident.
16 seeds have never won, 15 seeds win once in a blue moon, and 14 seeds every 5 years or so. But there's a reason the teams seeded 1, 2, and 3 are where they are - they're usually pretty good.

My pick this year: North Dakota State over Kansas. Yes, I'm REALLY confident in NDSU. Kansas lost in the last 2 weeks to Baylor, who would have been a 12 or 13 seed if they'd made it, and Texas Tech, who'd be lucky to get a 14. That's vulnerability, folks, and NDSU is one of the most underrated teams out there. They have a dude who dropped 60 on another NCAA tournament qualifier earlier this year, and they've played giant killers before, though not on this level. Watch out for the Bison!

5. 13 is the new 12.
This is a new one for me, but I've begun to consider taking a 13 seed just about every year. Parity is becoming more a part of college basketball by the year, and with it means very talented and well-coached 13 seeds. With the gap closing, I wouldn't be surprised if 13's started knocking off 4's on a yearly basis.

My pick this year: Cleveland State over Wake Forest. Wake is aggravatingly inconsistent, and Cleveland State is seeded lower than it ought to be. They're basically on par with Butler, who's seeded 9th. They were 1-2 against the Bulldogs, but all 3 games were down to the wire, dead even showings. And Cleveland State beat Syracuse in Syracuse (remember the 70 foot buzzer beater in December?).

Second Round Guidelines
6. Always pick a double-digit seed to advance.
It happens nearly every year, with frightening regularity - a team seeded 10 or higher slips into the sweet 16, leaving the nation crying "shocker" and "Cinderella". But it happens every single year! Be ready. Find that hidden gem, and make the pick. Trust me. It's a very satisfying moment to see the 10 seed you targeted become the national darling for the week.

My pick this year: Cleveland State. See above - the Vikings are this year's Cinderella story.

7. Always pick at least one 2-seed to lose.
While 1 seeds are near locks at this level, 2 seeds are typically vulnerable, and almost every year one or two, sometimes more, go down. Locate that hot 7 or 10 seed and go with it.

My picks this year: California over Memphis, Texas over Duke. I make it a personal rule to pick Duke to lose way before they ought to, and Texas is way too talented to be a 7 seed - talk about underachieving this season! And Memphis hasn't played a major conference team in about 3 months. They've feasted on the dregs that make up the rest of Conference USA, and the step up in competition could catch them off guard. It should be noted that they did play Gonzaga and won convincingly, but they're still not a major conference foe.

8. Pick at least 5 teams seeded 5 or above to reach the Sweet Sixteen.
I keep track of these things (mentally, at least - I have no hard data to back this up), and have noted that at least 5 teams that "shouldn't" reach the Sweet Sixteen - that is, they aren't one of the top 4 seeds in their region - do make it. So don't just drop the top 4 seeds in place, you'll lose.

My picks this year: the aforementioned Cleveland State, Texas and California; West Virginia, Purdue, Arizona State

Later Round Guidelines
9. Pick at least one team seeded 6 or higher to make the elite 8.
Cinderella teams are what make the tourney fun, and usually one threatens, and sometimes even makes, the final 4. Davidson was a 10 seed last year, Gonzaga's done it from that spot a few times, George Mason was an 11 seed when they went to the final 4, and on and on.

My pick this year: West Virginia. They're gelling at the right time and have a favorable bracket. 

10. Add up the total of the seed numbers of your Final 4 teams, and make sure it's between 5 and 10.
Everyone knows that all four 1 seeds never made the Final 4 before last year's anomaly, but there typically aren't many really high seeds either. In fact, it's probably unwise to consider teams seeded 6 or above - it just doesn't happen that often. Usually the Final 4 looks something like this: two 1 seeds, a 2 seed, and a 3 or 4 seed. Total number in that case: 7 or 8. Perfect.

My picks this year: Louisville, Pittsburgh, Oklahoma, and Purdue (total number: 9; two 1's, a 2, and a 5)

11. Pick a 1 or 2 seed to win it all.
It's ok to take a 5 seed like Purdue to the Final 4, but don't get cute with them once they're there, they'll probably lose in the semifinal. Except in some rare instances, the whole tournament is won by the 1's and 2's of the world. So pick your favorite of the bunch and go with them.

My pick this year: Louisville. Crazily, the Cardinals are the only 1 or 2 seed (other than Memphis, who by playing in Conference USA doesn't count) who won their last game. All the others were bounced from their conference tournament early. When in doubt, pick the hot team.

12. Expect your bracket to be in shambles by 10pm Friday.
No matter how well you prepare, it always seems to happen. Have fun with it while you can, then give in to rooting for all upsets and general mayhem, then begin to dream again next year.

And finally, here's the full rundown of my picks. Enjoy the madness!

MIDWEST
Second Round
Louisville over Siena
Cleveland State over Utah
West Virginia over NDSU
Michigan State over USC

Sweet Sixteen
Louisville over Cleveland State
West Virginia over Michigan State

Final: Louisville over WVU

WEST
Second Round
UConn over BYU
Purdue over Washington
Missouri over Utah State
California over Memphis

Sweet Sixteen
Purdue over UConn
Missouri over California

Final: Purdue over Missouri

EAST
Second Round
Pitt over Tennessee
Xavier over Florida State
Villanova over UCLA
Texas over Duke

Sweet Sixteen
Pitt over Xavier
Villanova over Texas

Final: Pitt over Villanova

SOUTH
Second Round
North Carolina over LSU
Gonzaga over Western Kentucky
Arizona State over Syracuse
Oklahoma over Michigan

Sweet Sixteen
North Carolina over Gonzaga
Oklahoma over Arizona State

Final: Oklahoma over North Carolina

FINAL FOUR
Semifinals
Louisville over Purdue
Oklahoma over Pitt

Final: Louisville over Oklahoma

Next week: I said we'd return to football for next week, but this is so much fun we might stick with basketball. Stay tuned...

No comments: