In Part I of 'If you have no soul,' successful bracket strategy for the SBR Bracket Contest was laid out. With the brackets filled now, it's time to put that strategy to the test.
In the first part, we discussed maximizing your points in the early rounds. If you want to make the money, you need to focus on the bracket’s rules, and may end up making plays that are a bit counter-intuitive.
In SBR’s March Madness contest, there were 540 entries in 2008. I expect there to be over 800 due to SBR’s increasing popularity. There are only prizes for the Top 10 finishers, so it’s a tough game. Regardless of which contests you are playing in, you need to estimate how many players it will have, and look at how far down the prizes go. In SBR’s contest, you need to finish in close to the top 1% to make the money.
Gaming theory can help you maximize your equity in this contest. If you do not finish in the money, there is no difference between finishing 11th and finishing 800th – either way you walk away empty handed. In this contest, you will notice that it is very “top heavy” – first and second prize pay $2000 and $1000 respectively, whereas the next eight prizes pay out a total of $2000.
When you are picking your Final Four and tournament winner, you have two objectives: minimize the odds of someone making the same picks, and maximize the likelihood of hitting a perfect final four. In a contest as top-heavy as this, it would be a mistake to pick all No. 1 seeds to win each region. While this is the most likely result, it is also the most common selection. If you were in a head-to-head bracket against a friend, this would be an excellent strategy. However, in a contest where you need to beat 98%+ of the other players, you are better off “gambling” on a few underdogs in the Final Four.
If you could find a Final Four bracket (including winner) that has a reasonable chance of happening and that only you picked, you would have the best of it in this contest, win or lose. You should still pick favorites for a vast majority of your selections, but tweak your Final Four to make it unique.
There is one other “rule of interest” – the points awarded for the final round are multiplied by the seed number if it is correct. If you selected a No. 4 seed to win the tournament, you would earn 16 x 4, or 64 points. From a gaming theory perspective this should not change your selection process much. If you pick the final team correctly and miss the other finals team, you will earn more points than a player who picked both finals team correctly and missed the winner regardless of the seed. The one impact of the rule is that more people will be picking long-shots to win the tournament. Consequently, you may find the greatest value on the favorites for the last round. Unless you have a compelling argument for a weaker team, I would go with a No. 1 seed to win the finals.
How did I fill out my SBR Bracket? I started with my “Sweet 16,” all No. 1 through No. 4 seeds, except I picked Purdue instead of Washington. That was based purely on the spreads in the first game – Purdue was an 11-point favorite and Washington was only a 5-point favorite. I picked the lower seed for almost every first round matchup unless I had that team in the round of 16. Yes, it is boring, but this is about money, right?
Given the rules for this contest, you might see close to one-third of all entries being very similar to yours. You want a unique final four that is likely to occur, but not likely to be duplicated. Going with three No. 1 seeds and one lower seed will get me there. When you select the teams in the finals, one last thing to look at is which teams likely had the easiest route there. Sometimes, a weaker team is more likely to win the tournament than the stronger due to the path in the bracket. It looks like UNC has the easiest path to the finals, making them a reasonable pick to win it regardless of who they play.
© Copyright 2009. Reprinted with permission of the author.