2008 Round 1 & 2 Host sites Take 2

I've taken another crack at possible host sites for this year.

This time I didn't use Top Drawer since they are lacking on RPI info compared to all of us educated by UPSF.

Also, the no conf matchups first weekend complicates things and increases travel, particularly for Pac 10 teams. The numbers after each team are RPI based on UPSF up to #100 and ncaa.com for 100+.

#1 seeds

UNC 1 ( C Caro 124, SC 30, ASU 33)

Stanford 2 gets shipped to NJ (Princeton 25, WF 21, Rutgers 22)

UCLA 3 (LB St 26, SD 23, Ill 35)

U of P 4 ( WSU 42, Minn 27, S. Dak St 125)

# 2 Seeds

FSU 5 (Mercer 130, Ala A&M 246, Aub 36)

ND 6 (Wis Mil 19, Evansville 99, MSU 32)

FL 7 (UCF 20, Miami 24, Cal 34)

BC 8 (BU 60, Harvard 40, LIU 81)

#3 Seeds
A&M 9 (LSU 38, Mc Neese St 89, 41 Geo)

Duke 10 (37 Charlotte, UNC GR 48, E Caro 44)

PSU 11 (Navy 139, Sienna 98, ASU 33)

OK St 12 gets shipped to Kan or Missouri (39 kan, 43 Missouri, LMU 46)

#4 Seeds

Vir 13 (JM 28, USC 17, Utah St )

Tex 14 get shipped to VT or WV (VT 29, 31 WV, 45 Purdue)

Col 15 (Denver 51, Weber St 194, BYU 18)

UW 16 gets shipped to: (Toledo 79, Murray St, Cincinatti)

I am not totally happy with this and did not spend a lot of time since it will change with remaining games and conf tourneys. The main thing I wanted to show is that the auto bids for teams with a large RPI is something like 16 teams. Thus, 64-16 means teams with RPI 1-48 make the at large berths meaning both UW and WSU get in. Right now UW gets seeded because of RPI 16 and they get shipped. WSU comes here. If UW loses their seeding I expect they will continue to get shipped and WSU comes here.

That is a fantastic piece of work. I agree that at this point, it's not worth trying to refine it further since a lot will change over the next week and a half. Still, keep it ready for a week from Sunday night or the next Monday morning so we can see if we can guess ahead of time what the Women's Soccer Committee will do.

One thing we should home in on as we get closer to the end: A team with a record below .500 cannot get an at large berth. The one team I see right now in the top 48 that might run into that is Georgia. If a top 48 team runs into that barrier, we'll need to drop down a spot. Also, we'll need to look at teams just below the top 48 to see if head-to-head results or results against common opponents would justify elevating any of them.

Again, great work!

Ooooh- clumping the high altitude schools together -- I like it! for some reason, though, the NCAA always seems to split this group (BYU, Colo, Denver, Weber st.) up.

One thing that has been churning through my brain since I read it was that The NCAA will use route planning software when laying out the brackets. (The NCAA aslo says it won't be the only method used, or even the primary one)

Route planning is a complex thing. UPS, for example, has spent millions on its route designing software that splits the delivery of packages among the drivers. UPS' software takes into account several criteria. It their case, it includes the directive that the truck drivers make no left turns if possible because they are the most dangerous.

So - what are the planning criteria for the NCAA seeding? According to the September committee meeting:

1) the top 16 seeds shouldn't meet until "normal bracketing procedures" say they should "if possible".
ordinarily, 1 seeds shouldn't meet until the semi's. They shouldn't meet 4 seeds until the quarters, 2 and 3 seeds shouldn't meet until the quarters etc.

UPSF showed that last year, at least, the NCAA set up the brackets so that all teams were seeded, even thought the NCAA says they don't do that.

2) the brackets overall should be designed for minimum number of flights. This may take priority over (1) , but it's unclear if that will be politically possible with the 16 seeded teams, at least.

3) After (2) is accomplished, the ground assignments should yield the minimum number of travel miles between venues, since that's how the NCAA reimburses schools.

Any others?


The last one (3) is what kept me thinking about how software might group the teams, because one method of determining the least milage pairing if all the scools travelled on the ground is fairly straightforward. You first pair the closest two schools, then the next two closest, then the next two closest, until all the pairings are done. Game theory shows this can often lead to the lowest total miles (I forget, but there may actually be a proof for this).

I'm not sure what the algorithm might be for determining the least number of flights. I'm sure it's related to the ground algorithm, but I'm not sure it's simply grouping closest teams. I can think of simple scenarios where that might not yield the fewest flights. The key is the directive is fewest flights, not shortest flights, so it may involve maximizing the number of 4 team pods in 400 mile radius clusters. That may indeed be closely related to the milage algorithm.

Additionally, the algorithm may actually be tuned for cheapest flights, and I don't think we can determine what those are unless the travel agent of choice choses to disclose the contracts it has made (doubtful).

One hint of further complication is the directive to "avoid remote or expensive travel regions". Venues outside the Continental US are discouraged, but there may be other areas. For all we know, they may mean Portland. Judging from our previous experience with the NCAA, they wrote that for us.


One random thought I keep having is that the use of software may split the country into four regions even more that the NCAA has done in the past.

Who would have thought I'd be pulling out 40 year old math books again?
Have any of you younger folks done any game theory lately?

Wow... great work, Gnarly! I was fiddling around with the seedings stuff a week or two ago - definitely not an easy task at all.

Kudos, and awesome job!

Whoever has Florida in their bracket has a clear shot to the College Cup, they looked like garbage last night, and have for most of the season, along with Texas A&M the two most overrated teams in the nation. Although Florida beat us, we out shot them 16 to 5, and of their 2 shots on goal 1 went in, bad luck that day.

FSUfan - You are a very welcome guest on our forum, and I am glad you joined. Gives us a little taste of what's happening in the East and in the South, which is truly refreshing.

I know Florida is your rival, like Santa Clara is ours. And you both have to live in the same state, which makes the rivalry even more personal and bitter. (Like the Oregon-Oregon State "Cival War" rivalry in our state - man, can it get nasty!!!)

I know you believe FSU is better than the Gators and you obviously feel that they stink. I have not seen the Gators play. (I missed the Pilots' match with them back in September.) However, they do sport a 16-2-1 overall record and an 11 - 0 record in the SEC going into the conference tournament.

Going 11-0 in any conference is quite an accomplishment. They must have something going for them. Even though FSU might have really outplayed them, Florida did win the match, and that is what counts in the standings and the RPI calculations.

Good luck with the rest of your season, and the NCAA tournament. Perhaps we will meet in the Final Four!