PDA

View Full Version : S-Curve.. Still in use??



gofurman
03-06-2017, 09:56 PM
Yes, this can probably be morphed in some other thread but I just wondered.. Do they still use the S-Curve at all? Is it used once you are beyond putting the one-seeds in their own regions or what? thx for quick answer

brevity
03-06-2017, 10:08 PM
Yes, this can probably be morphed in some other thread but I just wondered.. Do they still use the S-Curve at all? Is it used once you are beyond putting the one-seeds in their own regions or what? thx for quick answer

_No.

duketaylor
03-06-2017, 10:14 PM
the "No" answer I find surprising. Should be an S-curve with the exception that same-conference teams can't meet until 3rd round, or has that changed, too? I plan (trying) to formulate my own bracket before 6:30 Sunday and see how close I can get.

Nugget
03-06-2017, 10:20 PM
the "No" answer I find surprising. Should be an S-curve with the exception that same-conference teams can't meet until 3rd round, or has that changed, too? I plan (trying) to formulate my own bracket before 6:30 Sunday and see how close I can get.

Nope -- now teams from the same conference can meet as soon as the 2nd round if, due to an unbalanced schedule, they only played once during the season. For example, Lunardi has us today as a 4 seed with Virginia the 5 seed in our region.

Bluedog
03-06-2017, 11:31 PM
the "No" answer I find surprising. Should be an S-curve with the exception that same-conference teams can't meet until 3rd round, or has that changed, too? I plan (trying) to formulate my own bracket before 6:30 Sunday and see how close I can get.

S-curve has been gone for several years now. Teams get seeded 1-68 and the top 16 overall get placed according to geographic preference. So, if the top 1 and top 2 seeds are both closest to the East site, that's where they go. There is some rule that the regions can't be wildly out of whack by having the top 1, 2, 3, AND 4 seeds in the same region (they add the seed ranks and it must fall in some range), but that'd be extremely unlikely to happen by chance.

The rule about meeting a team from your same conference has been changed more recently, however, as nugget said.

gofurman
03-06-2017, 11:58 PM
Nope -- now teams from the same conference can meet as soon as the 2nd round if, due to an unbalanced schedule, they only played once during the season. For example, Lunardi has us today as a 4 seed with Virginia the 5 seed in our region.

The UVA comment. I don't see it. I just looked and I see Lunardi has us vs SMU in Greenville. If we beat Princeton. He has UVA in Milwaukee I think? Am I seeing something wrong?

http://www.espn.com/mens-college-basketball/bracketology

Nugget
03-07-2017, 03:02 AM
[/U]

The UVA comment. I don't see it. I just looked and I see Lunardi has us vs SMU in Greenville. If we beat Princeton. He has UVA in Milwaukee I think? Am I seeing something wrong?

http://www.espn.com/mens-college-basketball/bracketology

My error. It's Jerry Palm at CBS who has that in his current bracket

SCMatt33
03-07-2017, 07:10 AM
The S-curve NEVER existed in that there was never a world in which the number 1 overall seed would be locked into a region with #8 overall (save for procedural conflicts like those two being from the same league). Geographic concerns have always trumped matching up exact seeds. This continues to be true today. For example, unless they both fall onto the same seed line, there's about a 99% chance that Gonzaga and the top team from the Pac 12 are both in the eat region, even if they were to somehow end up as #4 and #8 overall. There is however, a true seed list of 1-68 for the field and there has been for a long time (though it's only been made public since 2012). They will use these seeds o try and add some general competitive balance to the regions (there are limits on the sum of the true seeds of the top for teams in each region) to prevent say the top team on each line or the worst team on each line from populating a single region. In the scenario I describe where Gonzaga and the top PAC 12 team are 4 and 8, they would likely place stronger 3 and 4 seeds to make up for it (though they wouldn't necessarily have to be the top 3 and 4 seeds, just stronger ones).

As for conference teams meeting in the second round. While the committee does have the leeway to do this, it is something they generally try hard to avoid and when they are forced into it, they are even less likely to make it two teams that are favored in the first round. I would wager a lot of money that you won't see Duke and Virginia in the second round. If it were to happen, it would be much more likely to be a 3-11, 2-10, or 4-12 game where one of the teams needed to upset someone in the first round to actually make the matchup happen.

Lunchab1es
03-07-2017, 08:09 AM
S-curve has been gone for several years now. Teams get seeded 1-68 and the top 16 overall get placed according to geographic preference. So, if the top 1 and top 2 seeds are both closest to the East site, that's where they go. There is some rule that the regions can't be wildly out of whack by having the top 1, 2, 3, AND 4 seeds in the same region (they add the seed ranks and it must fall in some range), but that'd be extremely unlikely to happen by chance.

The rule about meeting a team from your same conference has been changed more recently, however, as nugget said.

This is correct, with the added factor that the top 16 teams can not be at a home court "disadvantage". For example, if Duke were a 4 seed in Greenville, and Furman somehow made it into the tournament as a 13 seed, Furman would be sent elsewhere. Much like conference matching, this protection only applies to the first round, I believe.

duketaylor
03-07-2017, 09:20 AM
Thanks for the corrections!!

TexHawk
03-07-2017, 10:51 AM
S-curve has been gone for several years now. Teams get seeded 1-68 and the top 16 overall get placed according to geographic preference. So, if the top 1 and top 2 seeds are both closest to the East site, that's where they go. There is some rule that the regions can't be wildly out of whack by having the top 1, 2, 3, AND 4 seeds in the same region (they add the seed ranks and it must fall in some range), but that'd be extremely unlikely to happen by chance.

The rule about meeting a team from your same conference has been changed more recently, however, as nugget said.

This is true until it isn't. Just last year, Villanova was shipped to the South region to avoid giving #1 UNC a disadvantage in the East region. Xavier was shipped east. Kansas's geographic preference was Chicago but was shipped to Louisville, so Michigan State could have a "reward" for winning the Big10 but not getting a top seed.

To be clear, I appreciate that the committee works this way. I absolutely think they should take a step back and look at the big picture before locking everything in. Some years it is messier than others.

English
03-07-2017, 01:29 PM
This is true until it isn't. Just last year, Villanova was shipped to the South region to avoid giving #1 UNC a disadvantage in the East region. Xavier was shipped east. Kansas's geographic preference was Chicago but was shipped to Louisville, so Michigan State could have a "reward" for winning the Big10 but not getting a top seed.

To be clear, I appreciate that the committee works this way. I absolutely think they should take a step back and look at the big picture before locking everything in. Some years it is messier than others.

To clarify, did the Committee Chair explicitly say that these things happened this way, or is that basically what everyone just knew? I cannot remember hearing the Chair say them, and it doesn't seem like something the Committee would actually say. Again, it is certainly something that was commonly accepted as what happened.

Bluedog
03-07-2017, 02:01 PM
This is true until it isn't. Just last year, Villanova was shipped to the South region to avoid giving #1 UNC a disadvantage in the East region. Xavier was shipped east. Kansas's geographic preference was Chicago but was shipped to Louisville, so Michigan State could have a "reward" for winning the Big10 but not getting a top seed.

To be clear, I appreciate that the committee works this way. I absolutely think they should take a step back and look at the big picture before locking everything in. Some years it is messier than others.

True, sometimes the committee does something a bit different based on a set of circumstances and gives a GREATER preference to higher seeds. Since UNC was a #1 seed, the committee "felt it was unfair to put ACC champion North Carolina [...] into a regional that might require it to play for a spot in the Final Four against a team [Villanova] with a huge potential home-court advantage" (philly.com) considering Nova played 3 home games there that season (the max allowed by the NCAA to NOT be considered a home court officially). The committee did not say the above explicitly as far as I can tell -- people just assumed.

This year, the #1 overall seed will get to choose their geographic preference. I'd imagine the committee saw Louisville vs. Chicago as not THAT big of a difference from a geographic footprint perspective for Kansas, but it was for MSU -- where Chicago is clearly B1G country, but Louisville certainly is not. So, yeah, they did some balancing too admittedly there.

Nugget
03-07-2017, 02:47 PM
True, sometimes the committee does something a bit different based on a set of circumstances and gives a GREATER preference to higher seeds. Since UNC was a #1 seed, the committee "felt it was unfair to put ACC champion North Carolina [...] into a regional that might require it to play for a spot in the Final Four against a team [Villanova] with a huge potential home-court advantage" (philly.com) considering Nova played 3 home games there that season (the max allowed by the NCAA to NOT be considered a home court officially). The committee did not say the above explicitly as far as I can tell -- people just assumed.

This year, the #1 overall seed will get to choose their geographic preference. I'd imagine the committee saw Louisville vs. Chicago as not THAT big of a difference from a geographic footprint perspective for Kansas, but it was for MSU -- where Chicago is clearly B1G country, but Louisville certainly is not. So, yeah, they did some balancing too admittedly there.

I think for KU last year it might also have been that the Committee's computer program literally just compares distances and, technically, Louisville is slightly closer (545 miles vs. 548) to Lawrence than Chicago, so that's where KU was assigned. If that was their rationale, it was kind of silly since KU (like MSU) obviously would have preferred to be in Chicago than Louisville.

TexHawk
03-07-2017, 03:11 PM
I think for KU last year it might also have been that the Committee's computer program literally just compares distances and, technically, Louisville is slightly closer (545 miles vs. 548) to Lawrence than Chicago, so that's where KU was assigned. If that was their rationale, it was kind of silly since KU (like MSU) obviously would have preferred to be in Chicago than Louisville.

It can't be entirely automated. The committee clearly (and publicly!) pushed UK to Atlanta in 2012, when they had a better geographic option in St Louis. Lexington is closer to St Louis, but because of UK's stated preference as a "southern school", they got Atlanta.

Fwiw, I'm fine with all of that. The committee needs to start from a set of templates with loose rules/expectations, or the whole system would be questioned. But at the same time, small shenanigans or tweaks should be accepted as long as fairness is preserved as much as possible, and top seeds are rewarded geographically.

Hindsight is always easy here... the alternative of KU playing in Chicago against a hot Izzo that had just won the Big10 was not exactly appetizing. While Nova was sitting in Louisville coming off a bad loss in their conference tournament, and their starting big man was injured. Of course, Sparty got Middle Tennessee-d in Round 1, and Nova won a title. Such is sports sometimes.