/* ]]> */
Oct 132009
 

Office football pools have different payout frequencies, including:

Single (at season-end)
Weekly
Mixed (both weekly and season-end)

You should apply different strategies for each, as I explain below.

1.  Single payout:  Pick favorites consistently

A season payout pool is not a game of prediction, it is a game of chicken.  The first participant to flinch and try to predict an upset loses most of the time.  If you start guessing upsets you’ll likely fall behind and might never climb out of your deficit.  Worse yet, should you get lucky and guess a few upsets correctly, you might get overconfident and wrongly conclude that you really can pick upsets.  Don’t let this happen to you.  Let the others make the mistake.  Nobody can predict upsets consistently so don’t try.  You cannot afford a mistake.  In a season payout, until the final few weeks of the season there is no such thing as picking too many favorites!

2.  Weekly payout:   Pick neglected teams

Instead of rushing to predict upsets, pause for a moment and consider the other pool participants’ self interests. Most of them understand the dominance of picking favorites for the Single Payout game. But they don’t stop to consider whether that strategy is a good one for winning the most weekly payouts. (Hint: it could be if so many of them didn’t already use it.)

Everyone obsesses so much about predicting which teams will win, but winning the highest number of weekly payouts depends as much on selecting neglected teams, teams nobody (or few others) in your pool also pick.  Being neglected doesn’t improve a team’s chances of winning but it does improve your chance of coming out on top if they do.

 Weekly Payout Maximization (WPM) is a strategy I designed to exploit the advantage of selecting neglected teams in your pool.  It considers all the teams’ Win Probabilities, the number of participants in your pool, and the likely distribution of picks on each team.  Once I model the distribution of participant picks I run thousands of Monte Carlo simulations to identify how many underbets win the most weekly simulations.  In some scenarios you could win twice as often as the Average Joe.  WPM typically suggests between 1 and 4 upsets each week, depending on the number of participants in your pool and which teams are most neglected.

3.  Combination payout:  Start with the Single Payout strategy, but if you fall significantly behind shift to Weekly Payout

First, do a few calculations on your pool’s payouts and decide which is more significant – weekly payouts or year-end.  If the year-end pot is considerably less than 3x the size of a weekly payout, you should probably just focus on the weekly payouts and follow the Weekly Payout strategy.

Otherwise, start out with the Single Payout strategy.  If by 2/3 through the season  (around Thanksgiving) you conclude you’re hopelessly behind — maybe you missed a week or you didn’t begin Single Payout strategy until well into the season — start playing the Weekly Payout strategy.  (In the future I plan to add tools to identify when you’re behind enough to shift from picking only favorites, and when you’re hopelessly behind).

Your chance of winning the year-end payout might be slightly better than 1/N, where N is the number of players in the pool, but probably no better than 2/N.  Meanwhile, you have 17 chances to win the weekly pool in an NFL season.

 Posted by on October 13, 2009 at 6:05 pm