It's about time to start this thread going. Based on experience over the last five years, it's possible to use the current ARPI rankings to identify the reasonably likely field of possible candidates for #1 and #2 seeds. The outside edge of each reasonably likely pool is the team with the poorest ranking over the last five years that achieved the particular seed -- i.e., a #1 seed or a #2 seed. The reasonably likely fields for these two seeds actually are getting pretty constrained at this point in the season (in order of their ARPI rankings: field of 9 for #1 seeds; field of 29 for #2 seeds): Reasonably Likely #1 Seeds Stanford UCLA FloridaState PennState BYU NorthCarolinaU Duke SanDiegoState FloridaU Reasonably Likely #2 Seeds Stanford UCLA FloridaState PennState BYU NorthCarolinaU Duke SanDiegoState FloridaU SantaClara WakeForest MissouriU TexasA&M TennesseeU Baylor VirginiaU PortlandU VirginiaTech BostonCollege UCF Pepperdine WashingtonU NotreDame MichiganU IllinoisU MarylandU WestVirginiaU Georgetown Denver
As we approach the end of the season, it seems like a good time to start looking at the candidates for seeds in the NCAA Tournament, the teams that are relatively highly ranked but that still are reasonable possibilities not to get at large selections, and the teams in the next tier down but that still have reasonable possibilities of getting at large selections. The following lists are based on comparing the Adjusted RPI rankings of teams at this exact stage of the season (through the 10/21 games) to the eventual tournament seeds and at large selections over the 2007 through 2011 seasons. The outer perimeters of each list match the most extreme outer perimeters over the five year period, so they should be pretty good and conservative as outer limits although there always could be a new outer limit set this year. Over the next two weeks, these lists will compress into narrower ranges. #1 seed candidates: 1 through 8 in the ARPI rankings: Stanford FloridaState UCLA PennState BYU SanDiegoState NorthCarolinaU SantaClara #2 seed candidates: the #1 seed candidates plus 9 through 20 in the ARPI rankings. Add: FloridaU WakeForest TennesseeU Baylor VirginiaU Duke TexasA&M Pepperdine OhioState UCF MichiganU BostonCollege #3 seed candidates: the #1 and #2 seed candidates plus 21 through 30 in the ARPI rankings. Add: WestVirginiaU NotreDame CaliforniaU PortlandU KentuckyU MissouriU Marquette WashingtonU WisconsinU Dartmouth #4 seed candidates: the #1 through #3 seed candidates plus 31 through 43 in the ARPI rankings. Add: Georgetown CentralMichigan TexasTech MarylandU VirginiaTech Denver LongBeachState Rutgers StephenFAustin OregonState ArizonaState UtahU Louisville Teams with relatively high rankings but that are reasonable possibilities not to get at large selections: teams ranked #26 or poorer (which means that teams ranked #25 or better are highly likely to get at large selections). Teams with poorer rankings but that are reasonable possibilities to get at large selections: teams ranked #58 or better (which means that teams ranked #59 or poorer are very unlikely to get at large selections). Current at large bubble area: teams ranked #26 through #58: MissouriU Marquette WashingtonU WisconsinU Dartmouth Georgetown CentralMichigan TexasTech MarylandU VirginiaTech Denver LongBeachState Rutgers StephenFAustin OregonState ArizonaState UtahU Louisville ColoradoCollege MinnesotaU Princeton LaSalle WashingtonState Auburn LoyolaMarymount ArkansasU Dayton MiamiOH MiamiFL UCIrvine SouthFlorida IllinoisU MississippiU
In order to qualify for an at large selection for the Tournament, a team must have at least an 0.500 record. Here are the teams that are current potential "bubble" teams that look to be at risk of not meeting the requirement: #48 Auburn 10-9-0. Remaining games Alabama and SEC Tournament. If they lose to Alabama and in the first round of the SEC Tournament, they are out of the running. #51 UC Irvine 9-8-1. Remaining games CS Northridge, Hawaii, and Big West Tournament. If they win one of the two remaining regular season games (most likely Hawaii), they'll meet the 0.500 requirement. #52 Arkansas 9-8-1. Remaining games LSU and SEC Tournament. If they lose to LSU and in the first round of the SEC Tournament, they are out of the running. #57 Illinois 8-7-2. Remaining games Michigan and Big 10 Tournament. If they lose to Michigan and in the first round of the Big 10 Tournament, they are out of the running. #61 LSU 8-7-2. Remaining games Arkansas and possibly SEC Tournament. They actually are outside the current bubble group but close, so I've included them. If they lose to Arkansas, nevertheless get into the SEC Tournament, and then lose in the first round, they are out of the running. If LSU could beat Arkansas and run deep into the SEC Tournament, they might move up enough in the rankings to be a bubble team. Both Arkansas and LSU have lots riding on their upcoming game.
I guess we should be getting worried about how the Committee is going to seed (or not seed) San Diego State and BYU. It'll be interesting to see how much they get downgraded by the criteria for seeding that is independent of the RPI (wins against teams selected for the tournament, wins against the teams initially selected for seeding, last 8 games, etc). On the surface, this isn't an RPI issue - as they place highly in the RPI, just as they apparently would in an Elo rating system (judging by the first column Massey ratings). At a glance, and coming from an Elo rating perspective, I believe these teams deserve seeds and pretty high ones, probably 2nd seeds at least (traditional seeding #8 or higher). But by playing in conferences that may not contain as many playoff teams such as the ACC, Pac12 and, potentially, the SEC - their record may not stack up as well in the Committee's criteria (and based on how the Committee interprets or applies those criteria).
I just ran through a process of applying the NCAA's decision criteria (which are advisory but not mandatory for seeding) to the top 6 teams in the Adjusted RPI, using data from games through yesterday (Wednesday 10/31). I use a pretty rigorous process, with the only real judgment involving ranking teams' results over teams already selected for the bracket including automatic qualifiers with ARPI rankings of 75 or better. Since I don't know who the automatic qualifiers will be, I included all teams ranked 75 or better -- but I don't think that made the slightest difference. For the top 6, I ended up with the teams in the following order (without reference to how the bottom teams would fit with teams ranked below them in the ARPI for purposes of seeding beyond the four #1s): 1. Stanford 2. Florida State 3. BYU 4. Penn State 5. UCLA 6. San Diego State So, my take is that the criteria would award the four #1 seeds to the top 4 on this list.
I'll take a stab at how the seeds might look like: 1 Stanford 1 Penn St 1 Virginia 1 SDSU 2 UCLA 2 FSU 2 Florida 2 BYU 3 Baylor 3 tOSU 3 Wake Forest 3 UNC 4 Tennessee 4 Duke 4 Marquette 4 Santa Clara I expect a group of death quadrant of 1-2-3-4 Virginia, FSU, Wake, & Duke. From a Stanford point of view, seems like Idaho St will be the 1st-rd opponent, then a potential 3rd-rd matchup between SantaClara-Cal, should these teams win out to that point.
I'm calling the four one seeds using URPI and results against other top teams. Stanford BYU Penn State FSU I don't think Virginia gets many points for splitting with FSU and splitting with #10 UNC or 20+ RPI Maryland, no matter how good they looked today ( and they did look good). Each of these wins is wiped out by losses to the same team. So their best unambiguous win is 15 Duke. And they lost to Penn state, who lost to BYU and Stanford. BYU's only loss was Stanford 1-0 FSU's best unambiguous wins are 15 Duke, 14 Wake( which Virginia lost to) 10 UNC, and 8 Florida. Even though they crashed at the end, my gut says FSU is the fourth 1 seed. I do wish they would stop scheduling 270+ teams and make the decision easier. SDSU just didn't play hard enough teams besides losing to Stanford 1-0. Their best win is 18 Portland 2-1 The only other candidate, UCLA, pooched they shot losing to 101 USC. Their best win was 11 Tennessee.
I'd like to think that they won't cluster the ACC teams. Last year they separated them to the maximum degree possible (eight of the teams in separate eighths of the bracket with only one of the worst ACC teams, Virginia Tech, having to play another before the quarterfinals). I would hate to see it, but I could see Duke dropping out of the seeds. I would think that Wake would have to be at least equal or ahead of UNC and Duke, going 2-0-1 against them + a road win over Virginia. Don't think that UNC and Duke have anything better than that on their resume.
Oops, yeah (#62). But they still beat Penn state(3) which is still a better win than the other candidates. They also beat Portland, if you are comparing to your SDSU team. And it was their ONLY loss. Their other blemish is a tie with 11 Santa Clara. And I really don't think that the committee could bring itself to seed 4 West Coast teams when it is this close.
For those who want to see a well-informed projection of the at large selections and seeds, I recommend AllWhiteKit's Chris Henderson and his prognostications, here: http://www.allwhitekit.com/. I have a process I go through to see what I think the at large selections and seeds should be, using the NCAA's criteria that are mandatory for at large selections and advisory for seeds. Sadly, it is a very pedantic and time consuming process and I probably won't complete it before the bracket announcement tomorrow. I'll do it anyway, but I'm guessing what I come up with will be very close to the AWK projections, especially for the at large selections.
Oops, you're right. I was trying to seed just 3 Western teams in the top8 but the Sam Adams kept me from thinking straight. FWIW, this is how I see the committee sort things out: 1 Stanford 1 Penn St 1 Virginia 1 Florida St 2 BYU 2 Florida 2 Wake Forest 2 UCLA 3 SDSU 3 tOSU 3 UNC 3 Baylor 4 Tennessee 4 Duke 4 Marquette 4 Santa Clara
They do absolutely consider "bad losses." I don't know in what proportion to a team's good wins. Common sense would lead me to assume that a win over PEnn St outweighs a loss to Utah, in this instance. I wasn't saying SDSU should be ahead of BYU, just pointing out that BYU's loss to Utah has more effect than if they lost to Stanford.
I hope Chris is wrong about Dayton & Colorado College. I'd certainly rank them above LBSU and Miami, but I don't know if that really makes sense the way the NCAA does it.
Dayton basically has zero chance. Colorado College is essentially #65. I'd give them roughly a 40% chance of making it over #64, Oregon State.
You can make a dominant case for Dayton. They should make it with their decent RPI (49) and dominant wins. You see 8-1-2 results against a good conference, the Atlantic10, with the tie of VCU in the tournament not counting as a mark against them, and their wins over teams are pretty dominant. Oregon State should not advance on the merits of one early win and one good tie. They have too many blemishes.
What weighs more, bad losses or good wins? Seems like this varies from year to year depending on the committee.
Dayton has no case. They were ranked #15 of 15 bubble teams I ranked to make a final projection. They were the only team in that list without a win or a tie against a team projected to be in the field, a fellow bubble team, or a RPI Top 75 team that won an auto bid. If they get in, it'd be utterly unexplainable given the NCAA's selection criteria, and trust me, I like Dayton, so that hurts to say.
This year's bracket: http://www.ncaa.com/sites/default/files/external/gametool/brackets/soccer-women_d1_2012.pdf
Rapid reaction: UCLA a 3 seed? Santa Clara not seeded and gets Stanford in second round? I thought UNC would get a lower seed than 2. But the ACC got 6 seeds and 9 teams total, which was predicted. Maryland got a 3 seed! Called it The Big East got only one seeded team: #4 Marquette in BYU's bracket
Sick and tired of the same teams year after year coming to Gainesville, lol. I mean, if you can send Auburn to Washington, why not send some other teams to Florida? seems like we see FGCU UCF and/or Miami every year. Harsh 2nd round for Stanford. UNC got a head scratching favorable draw. I mean, seriously.