Collude For The Underdog!

Some college hoops notes:

• The 3/11 idea of heading to Northern Arizona is looking bleak. Pun City needs Portland State to go 0-4 in conference to close out their Big Sky season, in addition to Northern Arizona going 3-0 to finish it off. PSU plays 4-8 Northern Colorado tonight in the first of these 7 games. The Flagstaff idea could hit the deck tonight, gametime is 8:05 CST.
• The 3/11 idea of heading to Los Angeles is looking solid. A friend of the blog is around then and will be able to host. A possible return trip to Pepperdine and/or Cal State Fullerton may be on the horizon, as this blog was unable to get to their bookstores previously. Long Beach State is just a cool campus, so that might get added on also.
• Pun City wishes UWM would be able to make the NCAA tournament, but that scenario would appear unlikely at this juncture. The Panthers almost certainly would need to get a win in Hinkle Fieldhouse against Butler, which seems unlikely at best.
• One thing that Pun City has given some thought to recently is conferences sending a maximum number of teams to the NCAA tournament. Some say the Pac 10 could send 8, some say the Big East could send 8, you get the idea. Someone's bubble always gets burst though, when an unexpected team wins the automatic bid. This raises an interesting question: If Pun City were the conference commissioner for a conference like the Horizon League, would there be incentive to have your likely at-large bid team lose in the conference final? For instance, if UWM (with no at-large hopes this season) were facing Butler (very close to an at-large lock) in the final, would there be good reason to collude for a UWM victory, thus sending both to the NCAA tournament?

According to CNN Money, the NCAA distributes TV revenue to each conference based on the number of games teams from their conference have played in the tournament over the last 6 years. Having 2 teams in the tournament instead of 1 obviously produces more opening round games for a conference. The question to ask would be, would a team like Butler be significantly less likely to win their first round game if their seed suffered due to the loss?

Hypothetically, if a loss in the Horizon League title game would only drop Butler from a 5 to a 7 seed, and the Horizon would pick up UWM's opening game, would their expected number of games go up?

Well, yes. From 2002-2006, 5-seeds have a 10-6 record in first round games. 7 seeds have a 10-6 record also. So in either case, there's a 5/8 chance of 2 games for the at-large team in this hypothetical. Adding 1 game from the automatic bid and you have 2.63 as your expected value of games in the first 2 rounds. This is in opposition to the 1.63 if UWM loses.

The 3rd round is where the stats come back. 5 seeds are 10-4 in the 2nd round against either 4 seeds or 13 seeds (between 2002 and 2006). Conversely, 7 seeds are 6-9. So the expected value of the 3rd round would be 0.71 for 5 seeds that make it there (0.44 combining odds from 1st and 2nd round). Expected 3rd round value for 7 seeds that make it would be 0.4, with a combined odds of 0.25.

4th through 6th rounds of the NCAA tournament provide expected values of 0.125 for 7s and 0.21 for 5s.

Overall expected value for 5 seeds is 1.285 rounds, overall expected value for 7 seeds is 1.0 rounds. Over 6 years, allowing the "at-large" (Butler) team to win the game is going to net an extra 1.71 games for that team. Over one year, allowing the "auto bid" team (UWM) to win is going to net 1 sure extra game for the conference and 0.285 less games for the at-large. A net of +0.715 games every year, plus any bonus games the "auto-bid" team would pick up (13 seeds expect 0.224 extra games, essentially giving the conference 0.937 extra games every year).

So, Pun City isn't one to encourage collusion, but Conference USA, The Horizon League, the Missouri Valley Conference and maybe even the West Coast Conference would probably average about an extra 0.937 games' worth of revenue each year if they allowed a low-rated team to defeat their likely at-large team in the conference championship game each year.

Based on the CNN Money report, one game's worth of revenue translates to $6.25 million. Collusion has its benefits.