View Full Version : A seeding guess

02-27-2007, 08:05 PM
Ok, we're in the last week of the regular season, big games and all that for seeding implications and who is in and out of the tourney. So of course, I started looking at numbers and playing around with things and well, it might be contrary to what some others say, but here's how I think the seedings should be made up(I won't make any claim that this is how they stand now, or will end up, just that if the season ended right now, I think this is how they should go):
Seeds(* represents presumed conference winner):
1: UCLA*, Ohio St*, UNC*, Florida*,
2: Memphis*, Wisconsin, Pitt*, Texas A&M*
3: Duke, Southern Illinois*,Kansas, Georgetown
4: Maryland, UNLV*, Arizona, Nevada*
5: Kentucky, Air Force*, Michigan St, Virginia Tech
6: Tennesee, Louisville, Oregon, Boston College
7: Villanova, Marquette, BYU, Washington St
8: Indiana, Notre Dame, Old Dominion*, Butler*
9: Creighton, Illinois, Xavier*, Texas
10: Vanderbilt, Missouri St, Virginia, USC
11: Stanford, Drexel, Appalachian St*, Syracuse
12: Purdue, Florida St, Oklahoma St, Gonzaga*
13: Georgia Tech, Winthrop*, Holy Cross*, Vermont*
14: Akron*, Penn*, Sam Houston St*, Long Beach St*
15: Oral Roberts*, East Tennesee St*, South Alabama*, Marist*
16: Austin Peay*, Weber St*, Central Connecticut St*, Delaware St*/Missippi Valley St*

Given that setup, I wouldn't expect Ga Tech to make it in without winning at least two of their next three(either BC or UNC + an ACC tourney is the minimum to be considered, win any 3 and they are probably in). Teams like Old Dominion, Butler, Xavier, App St, and Winthrop are likely ahead of them should they not win conference tourneys. The 3 to the 6 seeds are really hard to differentiate IMO. This year it really does seem almost like a whim to place one team ahead of another. If I were correct, the teams teams that I'd think are hardest/easiest to deal with at each seed would be(hardest first, then easiest)
1: UNC/Ohio St
2: Wisconsin/Memphis
3: Kansas/Southern Illinois
4: Maryland/UNLV
5: Michigan St/Air Force
6: Louisville/Oregon
7: Washington St/BYU
8: Indiana/Old Dominion
9: Texas/Creighton
10: Vanderbilt/USC
11: Syracuse/Drexel
12 and beyond... I'm not going to pretend to have a clue. Anyway, am curious what others think(and there are some teams I'm pretty sure will get seeded lower/higher than I have, but I can't figure out why they are projected so looking at the numbers and all).

02-27-2007, 08:30 PM
I last updated my bracket prior to yesterday's games, but I think you're pretty close. A few things I think you're off on:

I think you have Arizona and Kentucky too high - their records and play over the past month don't justify their higher RPI. I had them both as 8 seeds.

I think 3 of your 10 seeds - Vandy, UVa, and USC, are all too low. They all have enough really good wins and a sufficiently strong conference record to overcome lower RPIs. I had UVa as a 4, Vandy a 5, and SC a 6.

I don't think Oklahoma State is anywhere close to a bid. Their record in February is abysmal.

I had Ga Tech as one of my last 2 in, and FSU as one of the last two out. The big difference for me came down to Douglas being out and their record since he went out. Unless he gets back and they play well with him, I think that's really going to hurt FSU's chances.

02-27-2007, 09:23 PM
You are probably right about the overseeding of Arizona and Kentucky. I have a bias towards teams that play a tough schedule, and do decently well with it(playing 12 top 50 teams and going 1-11 vs them isn't any credit to a team, but going 5-7 vs that group is worthy of respect IMO). Virginia and Vandy are two teams that I did struggle a bit with. Vandy's only good road win was Kentucky, no other top 75 teams on the toad, coupled with some bad early losses push them down to me(but admittedly, their last 2 wins vs Florida and Kentucky are a wakeup call, but again both at home, sandwiched around an away loss to Missippi St... I think they need to do well in the SEC tourney to justify a high seed). Virginia's resume is fairly vanilla to me, one early bad loss to Utah in a tourney(the App St loss in the same tourney doesn't look so bad as things worked out), and a nice win at Maryland, and not much else of really big note. I really don't get the love for USC. They beat Arizona home and away, and aside from that, have nothing particularly positive to note on their record. Beat patsies to open the season, losing their only 3 major conference nonconference games(South Carolina, Kansas, and Kansas St). If they win their last two on the road(especially Washington St) they can slide up, but otherwise, I'll be mystified if they are above a 7.

Anyway, will be interesting to see how it works out.

02-27-2007, 09:42 PM
*kansas is closer to a 1 than a 3.
*no way is unlv a 4. nevada better not be a 4 given that they have played nobody and beaten nobody.
*kentucky, mich st and air force all are worse than 5s.
*villanova is more like a 10 or 11 and wash st is more like a 3 or 4.
*uva and usc are more like 4s or 5s.
*if ga tech gets in (that's a big if), then they will be seeded higher than 13 due to their sos and rpi.

02-27-2007, 10:10 PM
As someone who runs his own unofficial Duke-centered bracketology blog, I'll share my thoughts:

1s: I think either KU or aTm will make a run for a 1 seed. Florida is playing their way out of one as we speak.

3/4: Duke, Maryland, and VT have very similar profiles as of today. VT has tons of Top 50 wins to make up for their bad losses. I would give Duke the slighest edge today, but I would rank Duke and VT consecutively on the S-curve. UNLV will probably be a 4 seed if they keep winning and should be seeded higher than Nevada.

5: Air Force is playing its way down a little bit, and Michigan State should be ranked more similarly to IU and UI. (They gave Amaker a much needed W tonight.)

7: BYU will be seeded to play in a Thursday-Saturday bracket the entire time. They have already played UCLA, so I see them playing in the east across from UNC.

8/9: Butler deserves a seed at this level but will be seeded higher because they beat a lot of good teams in November. Illinois is probably a little high given their overall profile. Texas has won a bunch of games in a row, so I would rank them as a 4-6 right now.

13: Winthrop deserves to be higher. I think they could be a surprise at a 8-9 if they keep winning like they have.

As far as "pods" go, I like your version because it should put us in Winston for the 1st 2 rounds, and I have tickets. ;) I think we end up in Winston if we are higher seeded than VT, UVA, or MD but not if we are equally seeded. (Hello Buffalo or Spokane.) If we are a 4, we still run a serious risk of ending up in Spokane if we are paired with a west coast team like USC, Nevada, or UNLV as a 5, and I'd rather not see us make that trip.

02-27-2007, 10:11 PM
When you look at Vandy head-to-head with Kentucky, Vandy has basically every relevant measure over Kentucky, including a head-to-head sweep.

USC vs Arizona is closer, but again USC swept Arizona, and also swept Oregon, and leads Zona by 2 games in the standings. Arizona has Memphis at home, and @ Oregon, but also has 5 home losses.

UVa is leading arguably the best conference in the country, and is 3-3 against the other top 5 teams in conference (4 of which were on the road).

I would say that USC and UVa probably don't have a lot of latitude on their higher seeds at this point - if they finish poorly they could drop to the 8-9 range or so.

I agree with dukie8 that the MWC teams are overseeded as well. In fact, with Air Force's loss today, I'm not even sure they're in any more.

02-28-2007, 12:02 AM
Air Force is still easily in. RPI 18 going into the night, and losing to the RPI 23 team won't change their profile all that much(down perhaps to a 7 or so).

As far as Vandy and Kentucky, Kentucky has the obvious edge in RPI and strength of schedule. They don't have a single bad loss this year(worst one, on the road to RPI 53 Georgia). Vanderbilt has lost to Wake and Auburn on the road, and Furman at home. Vandy has some better better wins, but the Kentucky win is the only really significant one they have away from home, and Vandy's homecourt is always a big edge.

Arizona is the team that has played a killer schedule(#1) with an RPI of 16, has 0 bad losses(every single loss they have is to a top 50 RPI team) and some nice wins. USC is RPI 50 - i.e. worse than Clemson and usually where you are wondering if the team will make the tourney at all. They have swept Arizona and Oregon, but those 4 games are the entirety of their good wins. If they can win their next 2 on the road they can get a boost, but if they split or lose two, they have work yet to do to ensure they are in the field.

UVa is 0-3 vs the teams which will be ACC tournament seeds. They have only 2 away wins of any note, Maryland and Clemson. I can see moving them up into the 8-9 range(with things being as close as they are), but they aren't close to sniffing a top 6 seed(rpi 38) unless they make a run at least to the ACC final, if not winning it.

And finally the Big 12 is kind of odd. The conference as a whole this year was lousy. Which meant that both Texas A&M and Kansas only had 2 games vs RPI top 25 teams, and both went 1-1. I don't see any justification for giving them top seeds(in fact, their profiles aren't that dissimilar from Nevada's, nice records vs not great competition, in their case a few more losses thanks to playing a few better teams, but not by a great amount). As far as Texas goes, it all comes down to their last 2 games, with A&M and Kansas. Split those and hop up to 7ish, win both and a 4 is a possibility, lose both, and 10/11 is where they are at.

duke2x, what's your site? I'm always curious to check out other opinions on such matters. And I'll probably crunch some of the numbers on Friday or so in anticipation of the end of the regular season.

02-28-2007, 12:22 AM
I have VCU over ODU. Certainly over Drexel.

02-28-2007, 03:55 AM
ODU took out Georgetown on the road, Drexel took out Creighton, Villanova, and Syracuse on the road, while VCU's big wins were ODU and Drexel. VCU's wins just aren't as impressive to me(and I've never though head to head should determine these sorts of things, you have an overall profile of 30 games, look at all 30, not just one matchup).

02-28-2007, 07:38 AM
1. Man, I know this isn't a popular sentiment on this board, but I really think a No. 3 might not be in the cards for us, unless we win out from here. Maybe I'm misreading things, but it seems like there are enough teams who are locks to be better seeds than us to push us down to no better than s No. 4. I'd love to be wrong, because I'd much rather be a 3 than a 4, but it just seems like a stretch right now.

2. Like others said, UK, Nevada, UNLV and Arizona are a bit high. Otherwise, I'll endorse your predictions.

02-28-2007, 07:56 AM
Deslok, I think you're generally over-relying on the RPI in seeding - it is usually not as important as things like big wins, road/neutral record, or how a team is playing coming into the tournament (I think of a team like Cal last year which had an RPI of 52 and was a 7 seed, or Washington with an RPI of 35 and a 5 seed).

Air Force should be in trouble - they've lost 3 straight, 5 of 10, their big wins are UNLV (6ish seed), @ Stanford (10ish seed), and neutral vs Texas Tech (10ish seed). I think they're probably still in, but 10 seed territory, and had better not lose in the 1st round of the MWC again.

The CAA is going to be a PITA again this year. I think ODU is clearly ahead of VCU for the top spot - if you compare RPI in conference games only, ODU is ahead of VCU even though VCU has a better record, due to the imbalanced schedule in the CAA. Plus the win @ Georgetown will be huge for ODU. Drexel is a tough one to place - they have all the road wins, but 4th place in the conference. That conference tournament is going to be interesting.

02-28-2007, 07:57 AM
ODU took out Georgetown on the road, Drexel took out Creighton, Villanova, and Syracuse on the road, while VCU's big wins were ODU and Drexel. VCU's wins just aren't as impressive to me(and I've never though head to head should determine these sorts of things, you have an overall profile of 30 games, look at all 30, not just one matchup).

ODU and Drexel admittedly have better OOC wins, and VCU's loss to Bradley hurts. But VCU also dominated the conference and stands in first place. ODU MIGHT get an at large over VCU. I see no way that Drexel does.

02-28-2007, 08:21 AM
i have no idea how you can compare kansas and texas a&m to nevada. nevada has played nobody and is a glaring 0-1 against the top 50. their best win was against gonzaga, which was not an at large team even before shroomgate. kansas has beaten the following tournament teams: florida, bc and usc. texas a&m has beaten the following tournament teams: winthrop, kansas (road) and texas. with florida, wisc and unc tripping up, kansas is looking like a 1 seed right now.

02-28-2007, 08:44 AM
I don't mean quite so literally that Kansas and Texas A&M haven't played anyone of note(as you note with Nevada). But rather, for teams eyeing top seats, their schedules are very weak. I mean come on, your best nonconference win for Texas A&M was Winthrop? That pretty much rests my case for them. For Kansas, the have a neutral site win vs Florida, and the only other RPI top 40 win they have is home vs BC. Its not that I think Kansas is horrible or anything, but that is not the profile of a top seed unless their record is like 31-1. Its the same reason Memphis has no shot at a top seed, even with a better record than Kansas. Memphis beat Kentucky on a neutral court, but their next best win is Mississippi.

Another admitted bias I do have for teams is I've never been a fan of relying heavily on the last 10 games. As a simple example, Air Force and Louisville's profiles are not dissimilar, with more of the overall numbers being in Air Force's favor, but because of the last few games, people are perceiving Louisville at a 4/5 seed, and Air Force as a 10 which to me is absurd. That is way overrating the last few games vs the season for the teams. Yeah it matters some, but last few games should be no more than a seed or two in each direction.