RebKell's Junkie Boards

[beknighted]

Judy Southard repeated something last night that she’s said before, which is that the WBCA had urged the committee to focus on geography in setting up the brackets, and that was why they’d taken the approach they did. I’m almost afraid to say this, but I think I believe her. I also suspect that she didn’t understand what was being said, or at least what the coaches really meant. (My guess is that most WBCA members want geographic seeding for their own teams because they know that they’re usually one and done. That’s different than how teams with championship aspirations see the bracket, and I doubt that most WBCA members want the two through six seeds assigned that way.) Anyway, I’ve been pondering the ideal way to set the brackets (or at least a better way), and thought I’d see what other people think.

This all proceeds from the assumption that the committee has picked all the teams and then sets the S-curve. Their procedures say that they nominate 8 teams for 1 seeds, pick the 1s, then nominate 8 teams for the 2 seeds, pick those, etc. They’re also allowed to move teams up or down a line on the bracket if they think they have to do so to fix some other problem. With that in mind, here’s what I would have them do:

1. After the 1 seeds are picked and ranked, assign them to regions in rank order, with the #1 #1 getting the best geographic match, etc. That is what they say they do today.

2. After the 2 seeds are picked, assign them initially strictly based on the S-curve - #5 gets the last #1 seed, #6 gets the #3 #1 seed and so on. If this does not result in any conference matchups, set the bracket this way. If it does put a #1 and #2 from the same conference against each other, make the minimum changes necessary to fix the problem, with minimum defined as whatever leaves you with a total ranking closest to 9 in each region (e.g., #1 #1 plus #4 #2 is 9, and #2 #1 plus #2 #2 is 8).

3. Repeat the process for the #2 seeds with the 3 and 4 seeds, except that (a) the initial placements are based on achieving balance (that is, the totals of the positions on the S-curve for the 1 to 4 seeds in each region come as close as possible to 34 {1 plus 8 plus 9 plus 16}) and (b) it’s permissible to have a 2 and 4 from the same conference in the same region and to have a 1 and 3 from the same conference in the same region, so long as they are not both among the top 3 teams from a conference. The balance condition may sound wacky, but it’s in the rules already and actually not so bad if you start out by trying to make the 1-2 lines balanced. A team could be moved to a different region and/or bumped up or down a line to address conference pairing issues issues.

4. Repeat the process for the remaining seeds through #9, except that the only constraints are that no two teams from same conference can meet until the regional final and that teams from any given conference will be allocated as evenly as possible among the brackets. Again, a team could be bumped up or down a line to solve any problems.

5. For the #10 through #16 seeds, place the teams the same way you place them for the #1 seeds, with the same constraints as for 5 through 8 seeds. However, use the subregional locations for this purpose, not the regional locations.

I pondered whether to apply the last rule to teams with better seeds, but I suspect that coaches with those higher seeds would rather get the “right” spot in the bracket than the most convenient location. Anyway, this approach seemed considerably more likely to meet the two constraints that seem relevant – rewarding and penalizing teams according to their merits and getting the teams that are likely to be going home early a chance to be seen by the home town folks – than the current system. Also, and not a minor issue, it makes doing the top of the bracket a lot easier because it presumes that you start with a balanced bracket, rather than starting with assignments that have nothing to do with balance.

So, what have I missed?

[LadyDevilFan]

[kage]

1. You lost me on your math in the parentheses in Rule 2. I think I get how the first #1 plus the 4th #2 = 9. But how does the 2nd #1 plus the 2nd #2 = 8...? What am I missing?

First #1 = 1. Fourth #2 = 8. 1 + 8 = 9.

Second #1 = 2. Second #2 = 6. 2 + 6 = 8.

_________________
http://img211.imageshack.us/img211/6329/lindseyhardingsig3oh2.jpg

[LadyDevilFan]

[braveniler58]

The top four seeds are placed on the S-curve, 1 2 3 4. The next four teams, all number 2 seeds, are "ranked" 5, 6, 7, and 8. Thus, the second #2 seed is the 6th team (the first #2 team would be fifth on the S-curve).

EDIT: It's not multiplication, but a list of some sort.

the S-curve is slanted as this,

1 2 3 4 (#1 seeds)
8 7 6 5 (#2 seeds)
9 10 11 12 (#3 seeds)
16 15 14 13 (#4 seeds)

So, theoretically, the number 1 overall seed in the NCAA tournament should have the worst #2 and #4 seed along with the best #3 seed in their bracket.

_________________
Phoenix Mercury fan for life!

[beknighted]

[LadyDevilFan]

Thank you braveniler and bek... so it IS some secret math I wasn't taught as a wcbb newbie. Thanks for enlightening me!

[myt]

beknighted, I see we are going to have to sneek you into the selection process room next year!

_________________

SuziQ wrote:

ima say this only once, and I'm never gonna say it again. Parker's damn good.

[braveniler58]

You're welcome. =)

_________________
Phoenix Mercury fan for life!