NCAA Bracketology

I am participating in the March Madness NCAA basketball bracket for the first time. It's $5 to buy in and there are 7 of us in the office pool. The exercise of picking teams and the results with no clue as to what is good or bad is rather interesting.

For those who do this seriously, the whole picking the bracket thing is quite an arcane art. There are folks who will 'analyze' your bracket for a fee. Sports betting and stock markets are places where statistics are as a wit once said, "There is lies, damned lies and statistics". As in the stock markets there is no real optimal strategy. So what is promised is mostly some mumbo-jumbo with a generous dash of snake oil. Such knowledge gives someone like me comfort since chance levels the playing field for us all: noobs, rubes, and the armchair jocks.

The theoretically possible brackets are 2^63 ~ a few quintillion. This means that while a perfect bracket is possible, it is out of the realm of probable events. ESPN has offered a $1M prize for someone who manages to do that. There is a better chance of finding a cure for AIDS than getting a perfect bracket. After the first 32 matches -- only 317 people out of about 6 million brackets submitted had perfect picks. It should be mentioned that 5 people had it completely wrong which is a considerable feat as well, though easier to get it all wrong than all correct.

It is obvious that it can't really be a few quintillion possibilities. All match-ups are not even odds. So, if you factor the seeding in question adjusting probabilities (for eg: #1 has a much better chance then a #16) you come down to a more manageable number a probability of 1:150,000,00. Still an improbable event. As of this morning 9am EST, the leaders on the leaderboard do not have a perfect bracket. It's down to only 3 people who made just a single error so far and it's only Day 3. So, ESPN will most likely never give up its 1M prize in the course of NCAA history, or even in the course of human history.

Back to the office. K.T. who has picked straight seeds is actually winning the office the pool. She was winning quite handsomely till the first 32 matches were played. As there were only 3 upsets, she was on top. However, as of this morning, she was still winning, but she picked Pittsburg to win and they lost in an upset to Butler. Implication: She's not going to make too many points in subsequent rounds. Hence the default no-brainer strategy may do well on average, but will never win. You need to pick some upsets. You need to take some risks.

I decided to use a three-prong strategy:

a) the National Bracket
This is the average of everyone's bracket. It is said to be a good strategy, actually the strategy that is guaranteed to be so safe that you will avoid embarrassment - meaning you won't finish last in your office pool. This will overcome my lack of knowledge of the game. It's like picking an stock index fund. Nothing spectacular, but avoids personal manias. It was a little different from straight picks, accounting for sentiment. I had left sentiment out of while doing this, so I picked MSU to win (they lost) and my home school to lose (they won). If I had shown some loyalty I would've been 20 points richer and if Michigan continues to win, it's going to destroy my future earnings. Shameful!

b) Taking a few risks
But, winning is what counts, so to win some, you gotta gamble some. So, just by simple probability there are a few good gambles. For example, looking at past histories of the Elite Eight and Final Four, it's obvious that all #1's do not make it to the Final Four. So, I put in one #2 in the Final Four. Now, if I had taken advantage of the maximum allowed 10 brackets I could have hedged my bets across the different selections, but then that would ruin the fun of the office pool and also increase my exposure to more than $5. I picked Florida(#2) in the Final Four and gambled on UConn(#3) to make it the Elite Eight over SDSU(#2).

c) Game Theory
Since everyone mostly thinks OSU is going to win, it makes sense to just go with OSU to neutralize the points from that win. KT and LT had picked Pitt to win and since Pitt lost last night, the huge risk of picking a different winner did not pay off. So, now everybody else in the office pool has OSU picked that result is not that interesting. It's going to boil down to the Final Four picks that really going to make the difference.

So, as of today how am I doing? I am in third. It must be said that of all the people in the office pool, I have the largest possible points remaining (PPR). So while I am losing by 30pts, it could well go my way. Let's see. In full disclosure, I have not yet watched a single game.

1 comment:

vcim said...

Thank you for a awesome article.You have given me some ideas and a different way to to write articles.

vCommissionPublisher Signup