Hi, Joe:

Love your work, but I have a question about how you (or the NCAA) apply the S-curve ratings to the actual bracket. Basically, my question comes down to this: Does the NCAA actually rank the teams 1-65 and then seed strictly according to the S-curve? If this were the case, the No. 1 overall seed should be in the same region as the worst No. 2 seed (eight on the S-curve) and the best number No. 2 (nine on the S-curve).

However, in the [Monday, Feb. 22] bracket, Kansas is grouped with Villanova (seven on the S-curve) and Vanderbilt (12). Shouldn't that region have Kansas as the No. 2 with West Virginia (eight) as the No. 2 and Ohio State (nine) as the No. 3 and so on?

Additionally, does the NCAA base regional [pairings] for the No. 1 seeds on S-curve placement (for example, should Kansas line up to play the worst No. 1 seed in the Final Four?). I seem to remember based on previous years that they do not, but wouldn't it make sense?

Thanks in advance. I seem to have this question every year around this time and would appreciate knowing the actual procedure once and for all.

Brian Herrmann

Durham, N.C.

Brian raised all good and common questions. Let me answer two of them right off the bat and then take you through the process of seeding the top four lines in the current (Feb. 22) bracket.

First, the committee does indeed rank the entire field 1-65 before actual bracketing begins. This so-called S-curve is followed to the extent that the various bracketing principles are not violated. In addition, teams are given geographic preferences so long as no unreasonable imbalance is created among the four regions.

Second, the four regions are indeed paired by the S-curve ranking of the No. 1 seeds. So instead of a set rotation of region pairings (East vs. West, South vs. Midwest, etc.), the regions are assigned so that the best No. 1 seed would play the "worst" No. 1 seed, etc., if all reach the Final Four.

Now let's go through the Feb. 22 bracket for teams 1-16 on Monday's S-curve:

(1) Kansas: As the No. 1 overall seed, KU gets the most favorable region (Midwest, St. Louis) and sub-region (Oklahoma City) placement.

(2) Kentucky: UK's natural assignment would seem to be South (Houston) and New Orleans. However, the campus is much closer to the East Regional (Syracuse) and Milwaukee sub-regional.

(3) Syracuse: The Orange cannot play where they host, so the next-closest available region is the South (Houston). This is mitigated somewhat by the convenient Buffalo sub-regional.

(4) Purdue: Our final No. 1 seed goes West (Salt Lake City), but receives the closest available sub-regional (Milwaukee).

(5) Duke: Remember, geography takes precedence over S-curve ranking so long as the overall regions are relatively balanced. The Blue Devils are closest to the East Region (Syracuse) and Jacksonville sub-regional. A count begins for the S-curve totals in each region, with the East sitting at seven (Kentucky-2 plus Duke-5).

(6) Kansas State: K-State cannot be in the Midwest with Kansas, so the Wildcats get the South Region (Houston) and the second pod in the Oklahoma City sub-regional. The S-curve count for this region is nine (Syracuse-3 plus Kansas St-6).

(7) Villanova: With only the Midwest and West available, these Wildcats find themselves in the closer Midwest (St. Louis) and convenient Providence sub-regional. The S-curve count for the Midwest is eight (Kansas-1 plus Nova-7).

(8) West Virginia: All that's left for WVU is the West, which is handy in that the first three teams from any conference have to be in separate regions as a bracketing principle. The Mountaineers get Salt Lake City and the second Buffalo pod. We are watching the S-curve count in the West, however, which is a little out of balance at 12 (Purdue-4 plus West Virginia-8).

(9) Ohio State: The Buckeyes could go anywhere except Purdue's region as the top No. 3 seed, but regional balance is becoming an issue. In a close call, we'll put OSU in the South (Houston) to give that region an S-curve count of 18. For the first weekend, with Milwaukee and Buffalo unavailable, Ohio State earns a ticket to Providence.

(10) Georgetown: The Hoyas are clearly closest to the East (Syracuse) region. Bracketing principles also prohibit teams from one conference being seeded 1-4 in the same region unless a fifth team from that conference is seeded that high. So there is no choice on Georgetown, as it turns out. Meanwhile, the S-curve count for the East is now 17 and the closest first weekend destination for G'town is Jacksonville.

(11) New Mexico: It's an easy call to place the Lobos in the West (Salt Lake City) and San Jose. The West's S-curve count is at 23 and bears continued watching.

(12) Vanderbilt: The last open slot for a No. 3 seed is the Midwest (St. Louis). The Commodores go there and grab the first New Orleans pod for the first weekend. The Midwest S-curve count is 20.

(13) Pittsburgh: As the fifth Big East team on the top four lines, Pitt can go anywhere. However, we need the strongest possible team to help bring the West Region into balance. So the Panthers go West (Salt Lake) and take the only non-West sub-regional still on the board (New Orleans). Our S-curve count in the West is at 36 (with 34 being "perfect" balance).

(14) BYU: The Cougars cannot go West because Mountain West rival New Mexico got there first, plus Pitt is also in the way. BYU's only other option is the East (Syracuse) because of the school's "no play Sunday" rule. The Cougars also need a Thursday/Saturday sub-regional, and there is one available in San Jose. The S-curve count for the East is 31, not ideal but within the committee's "acceptable" range.

(15) Michigan State: The Midwest and South are open, but Ohio State has blocked Sparty in the South. So Michigan State gets the Midwest (St. Louis) by default and also Spokane for the first weekend, as there are no other open pods remaining. The Midwest S-curve count is 35.

(16) Gonzaga: The last available slots are South (Houston) and Spokane, so you can see how likely it is that the Zags will get to play close to home. Our final S-curve count in the South is a perfect 34.

At the end of the day, we have adhered to all bracketing principles and given favorable geographic assignments wherever possible. We'd like a little more balance in the East and West regions, but nothing is out of the acceptable range.

The committee at this point would "lock" the top four lines of the bracket and move on with its business.

Joe Lunardi is the resident Bracketologist for ESPN, ESPN.com and ESPN Radio. He also teaches Fundamentals of Bracketology online at Saint Joseph's University. Comments may be sent to bracketology@comcast.net.