Reading, Writing, and Reflecting

March Madness is here and that means bracket time.  While I’ve religiously filled out NCAA tournament brackets for many years, the last 3 years I’ve also participated in an NCAA auction.   A different and fun way to enjoy the madness.

Here’s how it works:

1. You get 8 people who physically are together who each have to purchase 8 teams from the 64 team field from an agreed upon fixed budget– basically the entry fee to the auction; in this example, entry fee is $25.   So, all 64 slots in the tournament (the 64^th through the 68th play in teams, the individual teams wouldn’t be purchased, you could have Winner of the match as the slot to be purchased) will be purchased.
2. The teams are purchased in an auction style starting with the overall #1 seed and ending up with #64.  This follows the “S” curve to determine the order among teams with the same seeding.
3. Bidding is done in round robin fashion and each team is up for bid in sequential order. Once you drop out of bidding for a particular team, you cannot re-enter.
4. You must bid higher than the current bid to stay in bidding – except you can match the bid if that’s your maximum bid available.
5. Minimum bid is $0.25 (and all bids are multiples of $0.25).
6. Each person must end up with 8 teams. Therefore, the maximum you can bid on one team is $23.25 (leaving you with $0.25 for each of the remaining 7 teams you will get).

After you purchase your eight teams, they then each get points as they progress during the tournament, similar to the typical brackets based pool.  The point system we used was:

1. 1 point for a victory in the 1st round, 2 more for the 2nd round, etc. (thus, the champion gets you 21 points, not counting bonus points)
2. You get an additional bonus point for each victory that is an upset. An upset is defined as follows:
   • 1st and 2nd rounds only – A seed differential of more than 3 where the lower seeded team beats the higher seeded team (e.g. #13 beating a #4)
   • Subsequent rounds – any lower seeded team beating a higher seeded team (e.g. #2 seed beating a #1)

In preparing for this year’s auction, I analyzed the results of the last four auctions that this group has participated in.    Some of the key questions I wanted to explore were:

1. Does buying the most expensive team, in other words the favorite to win the tournament, generally lead to winning the pool?
2. Do you ultimately have to buy the team that won the tournament to win the pool?
3. Is there an ideal level of your total budget that should be spent on your most expensive team which leads to success in the pool?
4. Is purchasing say two # 3 and # 4 seeds vs buying one more expensive # 1 seed a better strategy?
5. Will buying more teams from the major conferences (ACC, Big East, Big Ten, SEC, Pac 10, Big 12) lead to greater success in the pool?
6. Does having the most purchases at the minimum price (in other words, the majority of budget was spent on first 3-4 teams) lead to failure in the pool?

Below is a summary of the average values in some key metrics over the past 4 years for the #1 through # 8 finishers.

Note-in the 2008 auction, there was a 3 way tie for first-I’ve included each of their values in the Rank 1, so the average for Rank 1 includes 6 players instead of 4.  Likewise, rank 2 and rank 3, do not include 4 data points but only 3.

What this data tells me is

1. Buying the overall number 1 seed (and generally so called favorite to win the tournament) doesn’t necessarily lead to a first place finish.   Only 1 of the 6 participants that won or tied as winner of the auction over the last 4 years had purchased the most expensive team.    Furthermore, only in one of the four years was the most expensive team on either the first or second place finisher, the 2 spots which are in the money.
2. There doesn’t seem to be an ideal target of total budget to allot to your most expensive team.   Rank # 1 was at 56%, which is similar to Rank # 6 at 55% and in between Rank # 7 and Rank # 8 at 62% and 48% respectively.
3. Interestingly, the average winner had purchased 2 teams between a 1 and 4 seed, while Rank # 5 player and Rank #8 player had purchased even more top seeds between 1 and 4.     Generally, the winners of the pool had some seeds above # 4 that were critical to their total point level differential from player #2 and #3.
4. Buying more or less major conference teams didn’t seem to have much bearing on the winners of the pool-pretty even distribution of number of major conference teams across all players.
5. Like most bracket pools, purchasing the team that ultimately wins the tournament does lead to winning the auction.
6. Having 3 or more teams purchased at the minimum price of $0.25 seems to be a bad indicator-rank # 7 and rank # 8 were the only players to average 3 or more for this metric.

Overall, to me, the data suggests (without doing time consuming regression analysis) that each year seems to be different and that there isn’t a clear combination of selecting the most expensive team vs combination of multiple # 2 through # 4 seeds that gives an advantage over other strategies.

However, try to limit the number of teams purchased for the minimum fee which indicates you may have overpaid for your earlier picks, and realize that you will probably need decent performance with one of your picks beyond # 4 to push you over the edge.  Of course, if you pick the team that ultimately wins the tournament, that goes a long way towards winning the pool, but doesn’t guarantee victory.