Thursday, July 20, 2017

New selection guidelines. Break it down!

http://www.ncaa.org/about/resources/media-center/news/di-men-s-basketball-committee-redefines-quality-win

A change in the nitty-gritty sheet. In the past, the sheet divides games into 4 categories: games against the RPI Top 50, 100, 200, and 201+.

The big change is the change in cutoff for each of the 4 categories. The first tier is now Top 30 home games, Top 50 neutral site games, and Top 75 road games. And so forth.  We'll do a full breakdown below.

This is a significant on the following levels:
1) They're clearly listening to everyone else. Everyone knows road games are tougher. Now, instead of relying on the committee to view the sheet and make adjustments in their mind, the adjustments will be made on the sheet itself.
2) This shows the emphasis they were putting into these tiers in the first place....which is something most people have underestimated over the years.
3) This is the first step towards getting rid of the RPI as a whole and/or coming up with a better ranking system to analyze teams.  This is a positive step towards recognizing a single number cannot be representative of the value of beating a team.  More information is always needed.

There's still a problem with the arbitrary way they've chosen the cutoffs. There's nothing special about 30, 50, or 75, or any other numbers. Still, this is better than the alternative.

Now, let's look at the actual hard numbers they're using.

Tier 1:
home games against RPI Top 30
neutral games against RPI Top 50
road games against RPI Top 75
- Changes in the tier:  subtracting home games against 31-50 and adding road games against 51-75.

I like the selection of Top 30 for home games.  With very few exceptions, every team in the top 30 is a tournament team.  So every game there represents at least some kind of value.

However, by losing 31-50, something interesting happens.  Some teams in the past, like Kansas or Duke, in order to build a strong SoS, schedule prospective conference champions from mid-major conferences.  Many of these teams wind up in the tail end of the RPI Top 50, causing the SoS of the giant to be inflated higher than it should be, and causing their W-L vs. the Top 50 to be inflated as well.  Using last year as an example, home games against UNC-Wilmington, Middle Tennessee, and Illinois St lose value.  Now with this change, home games against mid-major conference champions lose value.  Will opportunities dry up?  And how does this impact mid-major at-large bid chances, when one precious signature win chance (home game at conference leader) comes off the board?

On the flip side, with road games against 51-75 now entering this year....there are actually more mid-majors in this range.  Now the giants can play the occasional road game and add a quality win that might not have been viewed that way in the past.  Using last year as an example, road games against New Mexico St, ETSU, Charleston, and Winthrop now represent signature wins.  So you could actually see more giants playing true road games instead of home games against these types of schools.  Maybe not more than 1 in a given year, but spread across a few dozen teams, the impact could be large.

Tier 2:
home games against RPI 31-75
neutral games against RPI 51-100
road games against RPI 76-135
- Changes in the tier:  subtracting home games against 76-100 and adding road games against 101-135.

One impact of swelling to 351 D1 teams is that you have more teams with winning records occupying RPI 101-150.  Now some of these road games represent solid wins in this structure.  This is an improvement.  If you look, this range has a fairly balanced mix of middling majors and quality mid-majors.  I don't see an advantage gained for either faction here.  The major teams will add a couple of road games to the quality win ledger; the mid-major should also add a few such road games, but now will lose home games against RPI 76-100 in this category.

One conference that might get hurt specifically is the A-10.  They've been very proficient at having a few teams in the RPI 75-125 range traditionally, usually leading to solid W-L records against the Top 100 that get them higher seeds (and more bids).  With only Top 75 home wins now counting, they might get hurt a bit.

You'll also see much more attention paid to unbalanced scheduling.  Most conferences will have 4 teams, give or take a team, ranked in between RPI 75-125.  In most conferences with 14 teams, it's probably a given team would only play one of those four teams twice.  In those 5 total games....if 4 are road games, that's 4 Tier 2 games and 1 Tier 3 game.  If 4 are home games...that's 1 Tier 2 game and 4 Tier 3 games, and a big resume hit (in terms of chances at quality wins, and strength of schedule).  In the latter case, the team might be criticized for a soft SoS when it was simply the luck of the draw.

One big challenge for each conference is finding a strategy for unbalanced schedules.  And more importantly, making sure every team has a solid mix of road games and home games.  An imbalance of road/home games and good/bad opponents will cause a team to skew more than normal in the bubble discussion.

Tier 3:
home games against RPI 76-160
neutral games against RPI 101-200
road games against RPI 136-240
- Changes in the tier:  subtracting home games against 161-200 and adding road games against 201-240.

If a bubble team is a tournament team, they really should handle games against RPI 161-200.  No issue moving those wins down a level.

Usually a bubble team will have a loss or two or maybe even three in this category in the past.  Stuff happens.  But losses in Tier 4 are usually devastating.  Some of those Tier 4 road losses now move up a level, to this tier.  I like it.  I almost want to put every road game in this tier, but I do suppose you should leave some for Tier 4.  This will help a team like Monmouth cope with conference play dragging them down just a little bit.

Tier 4:
everything else

Losses to sub-240 teams, even on the road, are pretty unforgivable...maybe a team can still afford one.  The big thing is home losses to sub-160 teams now slip into this category.  Note that for power conferences, only the truly bad teams slip to sub-160, so I don't think the big boys will be hurt by this.  In fact, I think they're helped in that road conference losses in the top conference will almost never slip into this category.

Sunday, April 2, 2017

End of the season



End 'o the season.  I'm out of things to say, so this is the final post of the '16-'17 season.  We'll try and get some offseason posts up, some big-picture thoughts, but that won't be for a few months probably.  So go home, we're all done here.

Saturday, April 1, 2017

FINAL S-CURVE

The 1 line:  Villanova (31-3), Kansas (28-4), Gonzaga (32-1), Duke (27-8)
The 2 line:  Arizona (30-4), North Carolina (26-7), Kentucky (29-5), Baylor (24-7)
The 3 line:  Oregon (28-5), Louisville (24-8), UCLA (29-4), Butler (23-8)
The 4 line:  Florida St (25-8), Florida (24-8), West Virginia (26-8), Notre Dame (25-9)
The 5 line:  Iowa St (23-10), Purdue (25-7), Virginia (22-10), SMU (29-4)
The 6 line:  Cincinnati (29-5), Minnesota (24-8), Creighton (25-8), Michigan (24-11)
The 7 line:  Wisconsin (25-9), Maryland (23-8), Miami (21-11), Northwestern (23-11)
The 8 line:  St Mary's (28-4), Virginia Tech (22-10), Wichita St (29-4), Seton Hall (21-11)
The 9 line:  Arkansas (25-9), VCU (26-8), Dayton (23-7), Rhode Island (24-9)
The 10 line:  Marquette (19-12), Middle Tennessee (28-4), Providence (20-12), Oklahoma St (19-12)
The 11 line:  South Carolina (22-10), Michigan St (19-14), Xavier (21-13), Kansas St (20-13), Vanderbilt (19-15)
The 12 line:  Wake Forest (19-13), USC (24-9), Nevada (28-6), UNC-Wilmington (27-5), Princeton (22-6)
The 13 line:  Vermont (28-5), New Mexico St (25-5), East Tennessee St (25-7), Bucknell (26-8)
The 14 line:  Winthrop (24-6), Iona (22-12), FGCU (23-7), Northern Kentucky (22-10)
The 15 line:  Texas Southern (23-11), Kent St (21-13), Troy (20-14), North Dakota (19-9)
The 16 line:  South Dakota St (16-16), Jacksonville St (18-14), New Orleans (17-11), Mount St Mary's (19-15), UC-Davis (20-12), North Carolina Central (22-8)

Bubble:

Last 3 in:
Vanderbilt
Wake Forest
USC

Last 3 out:
Syracuse
Illinois St
Illinois



*administrative note:  Marking the date of this post as April 1 in order to pseudo-pin this entry to the top of the blog.  If you're paranoid, this S-Curve was also emailed in to the Matrix.

Tuesday, March 28, 2017

The disaster that is South Carolina

I think this may need an explanation.

I think everyone can agree, give or take a Wichita State, that South Carolina got the most flagrant seed from the selection committee.  I had them on the 11 line.  Bracket Matrix on the 9 line.  Very few gave them a 7 seed.  They were the most mis-seeded team in the tournament, period.

So don't get trapped by their run!  Their Final Four run does NOT mean that their 7 seed was justified.  Don't fall for that trap.

The real disaster of the situation is, though, that the committee has been rewarded for their bad seed.  They gave a middling power conference team a much higher seed than they deserved, and they got validated.  This just means that in the future, the committee is much more likely to give more middling power conference teams better seeds, instead of mid-majors.

Notice what happened to Middle Tennessee.  Selection committee said they probably would've been an at-large team if they needed it...but got seeded as a 12, below all at-large teams.  Clearly there's a disconnect between selection and seeding that needs to be bridged.  But if all these teams seeded 7th, 8th, and 9th keep winning multiple games or coming close (see S Carolina > Duke, Wisky > Nova, and the Arkansas/UNC close call, for example), there's going to be no motivation for the selection committee to move mid-majors up into those seeds.

So this is what I'm worried about.  The selection committee is going to use South Carolina as validation to keep over-seeding major conference teams and to bury mid-majors.

Friday, March 24, 2017

Conference analysis

This is late, but I always like to look conference-by-conference, and see if there are any outliers or anything unusual to point out.

America East:
1 NCAA
1 CBI/2 CIT
Not a bad year for the conference given Vermont trucked everyone.  I'm guessing New Hampshire declined a postseason bid, everyone else eligible went.

ACC:
9 NCAA
3 NIT
All the talk was 10 or 11 bids going in, and the ACC kind of hit the floor of 9 instead.  It just proved that there wasn't quite enough wins to go around.  It's tough to push 6-12 conference teams into the NCAAs.  In the end, Pitt, NC State, and Clemson took just a couple too many wins away from the pool.  Everyone over .500 at least got into the NIT.

AAC:
2 NCAA
2 NIT
The long-term trend isn't good for the conference.  The top 2 were obviously good, but the middle of this conference hurt the cause.  Houston was 12-6 in 3rd place, and couldn't mount a serious charge to the bubble.  The key?  Memphis and UConn, at 9-9, were also sub-100 in the RPI.  Those two, Temple, and Tulsa were all in between 100-150 when their goal should be 75-100.  This is something they'll need to work on.  Those 4 teams need to be on most years.  I am willing to chalk some of it up to variance, though.

A-10:
3 NCAA
1 NIT
2 CBI
Probably an average year for the A-10, given their champ was only on the 7 line and URI just squeaked in.  5 teams below them ranked between 73 and 122 in the RPI, which is the secret sauce to their success.  Interesting that some teams eligible for EIEIO tournaments sat, but GW and Mason played.

A-Sun:
1 NCAA
2 CIT
FGCU with the double crown here...interesting that Lipscomb apparently turned down the postseason...and two teams (USC Upstate, Jax) got in the CIT with pretty damn awful resumes.  Either the CIT has a A-Sun fetish or they're struggling.

Big 10:
7 NCAA
3 NIT
The distribution of bids is about on par for the conference.  Their seeds weren't.  I'm surprised though.  It's tough to get 7 bids out of a conference when there isn't a high-end team distributing quality wins to the other teams.  I don't know how to describe it.  Everyone did just enough in the non-con to survive.

Big 12:
6 NCAA
1 NIT
I'd say this is on par with the conference's expectations.  TTU isn't making the NIT, let alone the NCAA, with their scheduling practices.

Big East:
7 NCAA
They overachieved a bit, squeaking out 7 bids.  They kind of had a perfect distribution of wins between teams 2-7, with conference records of 12-6, 4 10-8s and a 9-9.  That's pretty lucky.

Big Sky:
1 NCAA
1 CBI/2 CIT
The 2nd-4th place teams took postseason bids.  Another example conference that likes to be fed by the EIEIOs.

Big South:
1 NCAA
3 CIT
Pretty standard for the Big South.  The other 2 teams that broke away in the conference standings took their CIT bids...as well as a 7-11 Campbell, for some unholy reason.

Big West:
1 NCAA
1 NIT
1 CIT
A disaster year for the conference.  3rd place team barely got CIT eligible.

Colonial:
1 NCAA
1 NIT
A couple strong RPI years for the conference now.  Sadly, it didn't pay off in a better seed for UNCW.  All I can say is to keep plugging along.  Charleston actually got an at-large bid for the NIT, which is a good sign for the future of this conference.  This is trending upwards.  Also interesting:  everyone else turned down the CBI/CIT.  If those tourneys can't pull upper-majors like CAA teams, something's wrong with them.

CUSA
1 NCAA
1 CBI
Considering how far off this conference is, it's kind of amazing Middle Tennessee almost built an at-large profile.  Here's a conference that's shunning the CBI/CIT types, with Rice being the one team that decided to play.  There's a couple teams that deserved some kind of postseason, that obviously passed.

Horizon:
1 NCAA
2 NIT
2 CBI
Pretty interesting...at-large NIT bid for Valpo.  Profile was just good enough, but didn't think the committee would pull the trigger.  Good for them.  Wright St probably turned down the postseason, which is the only way to explain the fraudulent CBI bid that UIC got.

Ivy:
1 NCAA
The conference was secretly really blah behind Princeton.

MAAC:
1 NCAA
1 NIT
3 CIT
Down a bit this year.  Poor, poor Monmouth.  Assuming Siena declined the postseason, a pretty fair distribution of postseason teams for the conference.

MAC:
1 NCAA
1 NIT
1 CBI/1 CIT
They have the resources to do better.  I wonder how long this will go before Akron starts to get fed up, because they're always the one on fringe at-large candidacy and can't get close enough.

MEAC:
1 NCAA
1 CBI/1 CIT
And frankly they got more postseason teams than they deserved.

MVC:
1 NCAA
1 NIT
Just a catastrophic year.  Everyone 3rd on back sat out postseason, presumably as penance for their crimes against WSU and ISU.  My advice in the future:  don't suck.

Mountain West:
1 NCAA
2 NIT
1 CBI
The thing is, I don't think the conference is THAT far away from where it needs to be.  After Nevada, they had 5 teams in between RPI 69 and 98.  This tells me incremental steps forward from those programs will rectify their single-bid status quickly.  They schedule mostly okay, they just need a few swing games.  Obviously they need to be better but I don't see a big philosophical shift needed to fix that.  Interesting that Wyoming in 7th took a CBI bid while 4th-6th in the standings decided not to play (assuming).

NEC:
1 NCAA
1 CIT
Seems fair.

OVC
1 NCAA
1 NIT
1 CIT
The conference was a bit down, and frankly I'm not sure how many more postseason teams they deserve.

Pac-12:
4 NCAA
3 NIT
Kind of a disaster the conference just snuck a 4th team in, when their top 3 was this good.  As it turns out...everyone else was terrible.  No other way to paint it.  It's a pretty clear problem with a pretty clear solution for them.  There's nothing in their scheduling practices I think they need to change.  Just play better.

Patriot:
1 NCAA
1 CBI
I'm guessing some teams turned down the CBI and CIT because the 6th place team here got the CBI bid.  Secret thing about these recaps:  you can see which conferences are shunning these tournaments.

SEC:
5 NCAA
3 NIT
A big year for the SEC, bigger than you think.  Remember they are in the middle of making great efforts to improve scheduling.  This year, those schedule benefits were reaped.  Vandy obviously got in on the strength of theirs, and the conference schedule didn't sink their overall SoS like it would have in years past.  Further, marginal resumes for Arkansas and S Carolina were enhanced by the SoS numbers.  Look at the RPIs of teams that missed:  UGa 53, A&M 93, Auburn 106, Ole Miss 68, Tenn 79.  In years future, those 5 bids will turn into 6 and 7 with those numbers.  Keep it up!  Good job!  I honestly mean that.

SoCon:
1 NCAA
1 NIT
2 CIT
An okay year for them.  All 3 tri-champs got a postseason bid, so all is just.  I'm guessing Chattanooga declined a bid...and why did 8-10 Samford get a CIT invite?

Southland:
1 NCAA
4 CIT
Here's a case where it looks like just about all eligible teams, and basically the top 5 teams in the league, took their CIT bid and went.  Always interesting to compare the conferences who walk away en masse and the ones to take them en masse.

Summit:
1 NCAA
1 NIT
1 CIT
Down a bit from their highs of the past couple years.  Not much else to say here.

Sun Belt:
1 NCAA
1 NIT
2 CBI/2 CIT
A good season overall, but that merely earned them the distinction of "best of the rest".  Best conference not to produce a serious NIT at-large candidate.  As you can see, several teams (interestingly, not Arky State though) cashed in for token postseason appearances...which they all deserved, more or less.  Good for them.

SWAC:
1 NCAA
lol

WAC:
1 NCAA
1 NIT
2 CBI
A big surge up the RPI for this conference thanks to the top 2 teams, but they didn't pay it off into anything tangible.  Those 2 CBI bids are pretty damn fraudulent.  Grand Canyon interestingly didn't play this year.

WCC:
2 NCAA
1 NIT
1 CBI
This is a pretty standard distribution for this conference.  The middle-to-bottom of this conference just isn't strong enough to support better.  All postseason bids and non-bids were just, IMO.

Saturday, March 18, 2017

Let's build a bracket.

Every year, I like to do this exercise.  I build a bracket using the committee's 1-68 seed list.  And see what's different about mine.  Full breakdown below, but I'll put the highlights at the top.  Needless to say, there's a couple questions I have.

- Cincinnati has a legitimate gripe.  They're ranked higher than Maryland; why did Cincy get shipped West instead of them?
- Vandy as the highest 8 seed got sent to Salt Lake City...not sure why.
- I would have swapped VCU and Marquette.  Although Marquette is ranked higher, I think VCU's travel considerations trump Marquette's given what regional sites were available.
- The committee made a mess of the play-in games.  I offer a solution that doesn't involve the Sacramento site below.
- I also would have tweaked the 16 line, but it's no big deal.

Below is the breakdown, and at the end is my bracket.

The 1 line:
1 Villanova - gets Buffalo and the East regional
2 Kansas - gets Tulsa and the Midwest
3 North Carolina - gets Greenville and the South
4 Gonzaga - gets Salt Lake City and West.  You could make an argument Sacramento would be a better site for them, but as it turns out this assignment works better for the Pac-12 teams behind them
Through 1 line, regions remaining:  BUF, GRE, ORLx2, INDx2, MILx2, TUL, SLC, SACx2

The 2 line:
5 Kentucky - their best location is the South regional, and Indianapolis is open
6 Arizona - goes to Salt Lake City.  The West is the obvious regional for them
7 Duke - obviously goes to Greenville.  As far as regionals go, East is better than Midwest
8 Louisville - gets the other spot in Indianapolis, and the Midwest is the open regional
Through 2 lines, seed score:  East 8, West 10, South 8, Midwest 10
Through 2 lines, regions remaining:  BUF, ORLx2, MILx2, TUL, SACx2

The 3 line:
9 Oregon - Sacramento to start.  As for the region, with the West blocked, the Midwest becomes the next best option
10 Florida St - here's where being the 4th best team in a conference is fatal.  Duke, UNC, and UL are ahead of FSU.  With those 3 occupying 3 of the regions, FSU is forced to take the last available region - the West.  At least Orlando is open for them for the 1st 2 rounds
11 UCLA - they're going to have to travel no matter what...but of the two available regions, the South is better than the East.  They take the second spot in Sacramento
12 Baylor - they get the second spot in Tulsa.  As for regionals, the East is the last one available.  The South would be better, but Baylor as the last 3 seed gets no power in this decision.  Let's put a pin in this and revisit later
Through 3 lines, seed score:  East 20, West 20, South 19, Midwest 19
Through 3 lines, regions remaining:  BUF, ORL, MILx2

The 4 line:
Before we begin, we note the conflicts:  Butler can't go in the East, Florida can't go in the South, and WVU can't go in the East or Midwest.  Actually this is pretty lucky, usually at this stage you'd have more conflicts
13 Butler - Milwaukee is their ideal sub-regional, and it's still available.  As for region, South and Midwest are both close to even, but South is slightly better.  We'll put them in the South, knowing we have some flexibility if we need it
14 Florida - Orlando still has a spot available, so they luck out on that front.  Since UF can't go in the South, they can go either in the Midwest or East, with about the same travel.  The East is slightly better, so let's put them there, knowing we can revisit
15 West Virginia - Buffalo is their preferred site out of the 8 total...and somehow is still open.  Amazing that it's working out this well.  14 of 15 teams so far have gotten their first choice of site.  As for region, the Midwest and West are left and Kansas is in the Midwest.  WVU has to go west for now
16 Purdue - Milwaukee is the last site available...and works out pretty well for Purdue.  Midwest is the last region available...and works out well for Purdue again
Through 3 lines, seed score:  East 34, West 35, South 32, Midwest 35

Balance for the top 4 lines looks okay.  Let's revisit a couple things.
- If we flop Baylor and UCLA, we can help Baylor's travel and make things even more balanced.  However, it would make UCLA travel all the way across the country instead of halfway.  Let's leave it as is.
- In a perfect world, it's Purdue as the lowest seed that has to go to the West.  However, the region would them add up to 36, which is teetering on the edge of being unbalanced.  Knowing that, it's better for WVU to go West than Butler or Florida.
- If we put Butler in the Midwest instead, we still can't do anything with Florida...I don't think there's a permutation that helps anyone too much here.

The top 4 lines are set, with very little conflict!  The only issues for individual teams are probably Florida St (but we can't do anything about that) and West Virginia (in a perfect world, Purdue travels).  Regions are pretty balanced.  Let's move on.

The 5 line:
subregional sites for the 5 line:  BUF, ORL, MILx2
17 Virginia - Buffalo is their best choice, but UVa played WVU in the regular season.  This sends UVa careening towards Orlando instead.  This does keep UVa in the East though
18 Minnesota - two Milwaukee spots are open for Minny, but since Purdue is in one of them, Minny takes the spot in the South regional instead
19 Notre Dame - Milwaukee is better for them than Buffalo, but they played Purdue in the regular season.  So they go to Buffalo; although this is the West regional
20 Iowa St - Milwuakee is their preferred site anyway, so this works out well.  Having Kansas in the sweet 16 from this draw isn't ideal, but rematch conflicts sort of tie our hands here.  Plus this puts ISU in the Midwest regional
Through 5 lines, we have not deviated from the selection committee yet.

The 6 line:
subregional sites for the 6 line:  ORL, TUL, SACx2
21 SMU - Tulsa is by far the preferred regional here.  It's the East region, which isn't great, but first weekend takes precedence
22 Cincinnati - Orlando would be better than either Sacramento spot.  This does bury Cincy in the West region, though
23 Maryland - only Sacramento is available.  We'll stick them in the South spot to leave the Midwest spot open for Creighton below
24 Creighton - see above; they get the other Sacramento spot
We have hit our first deviation.  In real life, the committee sent Cincinnati out west and kept Maryland in Orlando.  I'm not really sure why they did this.

The 7 line:
subregional sites for the 7 line:  GRE, INDx2, SLC
25 St Mary's - Salt Lake City in the West region is open.  Perfect
26 South Carolina - Greenville is available for them, which is basically the committee's wet dream.  Perfect, again.  East regional too
27 Michigan - the two Indianapois sites are available for Michigan and Dayton.  Perfect again.  Honestly, it's kind of amazing how neat this all works out.  The South and Midwest regionals are the two available.  Mileage is really close for both teams and regions...Dayton going South is the best advantage for either team, so let's do that
28 Dayton - see above.  By the letter of the law, Michigan should get the South spot, but the advantage is so small that consideration for Dayton should trump it
We do not deviate from the committee on the 7 line.

The 8 line:
subregional sites for the 8 line:  BUF, GRE, TUL, SLC
29 Wisconsin - actually, Buffalo is the best site for them, over Tulsa.  But it's close enough that I wonder if consideration of the regional should mater more.  Midwest over East?  Let's put Wisky in Buffalo for now, which is East
30 Miami - they can't go to Greenville (UNC), so to Tulsa and the Midwest they go instead.  Note that this was the alternate Wisconsin path, but it's better to use it for Miami to help them
31 Arkansas - Greenville and the South is miles better than SLC and the West, obviously
32 Northwestern - the perils of being the last team on a line, they have to take the hike to SLC and the West
We do not deviate from the committee on the 8 line.

The 9 line:
subregional sites for the 9 line:  BUF, GRE, TUL, SLC
Here's where things get fun with conflicts.
Vandy can't go to Greenville, SHU can't go to Buffalo, MSU can't go to Buffalo or SLC, VT can't go to Greenville or Tulsa
33 Vanderbilt - Tulsa and the Midwest region is best available
34 Seton Hall - after conflicts, Greenville and the South become best available
35 Michigan St - Buffalo and SLC are the remaining sites.  See note above.  Oops.  Back to SHU we go!
34 Seton Hall - if we kick SHU out of Greenville to make room for MSU, we have to send Seton Hall to SLC and the West region
35 Michigan St - Greenville and the South region for Michigan St
36 Virginia Tech - Buffalo and the East region is available for them
We have a deviation!  In real life, the committee sent Vandy out to Salt Lake City, and Seton Hall to Greenville, and Michigan St to Tulsa.  In my world, the committee sends Vandy to Tulsa, SHU to SLC, and Michigan St to Greenville.  Vandy as the highest team on the 8 line avoids the travel.  Not sure why the committee went the other way on this.

The 10 line:
subregional sites for the 10 line:  GRE, INDx2, SLC
37 Oklahoma St - they can take either Indianapolis spot; they'll take the Midwest one over the South one
38 Wichita St - Indianapolis works for them too, and they'll take the South one that OSU passes up
39 Marquette - Greenville and the East is obvioulsy better than the western stuff
40 VCU - here's another example of unfairness in being the last team on a line.  Greenville and the East is PERFECT for VCU, while for Marquette it's merely a little better than SLC.  However, SLC and the West for VCU is horrible.  By NCAA rules, VCU has to get screwed so that Marquette is helped slightly.  This is a spot where I would pull an executive decision.
So we deviate from the committee on VCU and Marquette.  Common sense rules.

The 11 line:
subregional sites for the 11 line:  ORL, TUL, SACx2
We've got play-in games, and this gets really messy really quick.  Orlando is a Thu/Sat, Tulsa and Sacramento are Fri/Sun.  Automatically we know Orlando is getting someone to play Cincinnati.  USC has to go to a Fri/Sun site because of travel reasons.
Here's what we know for the PIG guys:  Providence can't go to SAC, Wake can't go to ORL, USC can't go to ORL, SAC, or TUL, and KSU can't go to Tulsa.
So um, you see a minor issue.  USC can't go to ANY of the sites.  They're blocked out by travel (Orlando), regular season rematches (Tulsa), and conference conflicts (Sacramento).  Of the 3 choices, regular season rematches are the one that the committee will violate in this case.  So Tulsa is hosting a PIG winner, and USC is one of the ones going there.  Since Wake can't go to Orlando, let's put them with USC going to Tulsa.  Providence/Kansas St to Orlando.  Let's proceed:
41 Xavier - for non-PIG participants, just Sacramento is left.  X gets the South spot over the Midwest spot
42 Providence - see above
43 Wake Forest - see above
44 Rhode Island - Sacramento in the Midwest is last available
45 USC - see above
46 Kansas St - see above
In real life the committee lost its mind.  They set both PIG participants to Fri/Sun sites, including one in Sacramento.  They did USC/Providence going to Tulsa and Wake/K-State going to Sacramento.  I'm not really sure why they took this route.

The 12 line:
subregional sites for the 12 line:  BUF, ORL, MILx2
Things get a little easier now with one-bid conferences.
47 Nevada - Milwaukee, in the Midwest over the South
48 Middle Tennessee - it's actually pretty close, but Milwaukee is better than Orlando or Buffalo for them.  As it turns out, sending them to Milwaukee helps the two teams below them as well
49 UNC-Wilmington - Orlando, easily
50 Princeton - Buffalo, easily
We do not deviate from the committee on the 12 line.

The 13 line:
subregional sites for the 13 line:  BUF, ORL, MILx2
51 Bucknell - Buffalo is their best spot
52 East Tennessee St - Orlando is their best spot, although it's close
53 Vermont - Milwaukee is left, they get the Midwest because of the team below them who strongly prefer the South
54 Winthrop - the other Milwaukee spot
We do not deviate from the committee on the 13 line.

The 14 line:
subregional sites for the 14 line:  ORL, TUL, SACx2
55 New Mexico St - they'll take Tulsa obviously
56 FGCU - FGCU in Orlando. Perfect
57 Kent St - stuck in Sacramento, we'll give them the South over the Midwest
58 Iona - stuck in Sacramento
We do not deviate from the committee on the 14 line.

The 15 line:
subregional sites for the 15 line:  GRE, INDx2, SLC
59 Northern Kentucky - obviously going to Indianapolis.  South regional is better for them than Midwest, slightly
60 Troy - Greenville is their best fit
61 Jacksonville St - Indianapolis is their choice
62 North Dakota - goes to Salt Lake City
We do not deviate from the committee on the 15 line.

The 16 line:
subregional sites for the 9 line:  BUF, GRE, TUL, SLC
Buffalo and SLC are Thu/Sat; Greenville and Tulsa are Fri/Sun.  Buffalo is earmarked for one PIG winner, we could go for either Fri/Sun site.  Let's give them to the #2 overall seed (Kansas) in Tulsa.  Davis needs the Fri/Sun site for sure; New Orleans would be 2nd on the geographic consideration list, plus Tulsa would be a great regional site for them.  MSM and NCCU can play Thu/Sat then, in Buffalo, which is a decent location for either school.
63 Texas Southern - after the above, SLC and Greenville are left.  TSU actually slightly prefers Greenville...
64 South Dakota St - and SDSU prefers SLC, obviously.  Let's send TSU east to help SDSU out
65 UC-Davis - see above
66 North Carolina Central - see above
67 New Orleans - see above
68 Mt St Mary's - see above
We actually differ from the selection committee on what we do with the PIG.  Not sure why they had Davis play NCCU.

EAST 34
@Buffalo
1) Villanova vs. 16) North Carolina Central/Mount St Mary's
8) Wisconsin vs. 9) Virginia Tech
@Orlando
4) Florida vs. 13) East Tennessee St
5) Virginia vs. 12) UNC-Wilmington
@Tulsa
3) Baylor vs. 14) New Mexico St
6) SMU vs. 11) USC/Wake Forest
@Greenville
2) Duke vs. 15) Troy
7) South Carolina vs. 10) VCU

WEST 35
@Salt Lake City
1) Gonzaga vs. 16) South Dakota St
8) Northwestern vs. 9) Seton Hall
@Buffalo
4) West Virginia vs. 13) Bucknell
5) Notre Dame vs. 12) Princeton
@Orlando
3) Florida St vs. 14) FGCU
6) Cincinnati vs. 11) Providence/Kansas St
@Salt Lake City
2) Arizona vs. 15) North Dakota
7) St Mary's vs. 10) Marquette

SOUTH 32
@Greenville
1) North Carolina vs. 16) Texas Southern
8) Arkansas vs. 9) Michigan St
@Milwaukee
4) Butler vs. 13) Winthrop
5) Minnesota vs. 12) Middle Tennessee
@Sacramento
3) UCLA vs. 14) Kent St
6) Maryland vs. 11) Xavier
@Indianapolis
2) Kentucky vs. 15) Northern Kentucky
7) Dayton vs. 10) Wichita St

MIDWEST 35
@Tulsa
1) Kansas vs. 16) UC-Davis/New Orleans
8) Miami vs. 9) Vanderbilt
@Milwaukee
4) Purdue vs. 13) Vermont
5) Iowa St vs. 12) Nevada
@Sacramento
3) Oregon vs. 14) Iona
6) Creighton vs. 11) Rhode Island
@Indianapolis
2) Louisville vs. 15) Jacksonville St
7) Michigan vs. 10) Oklahoma St

Wednesday, March 15, 2017

1-68 seed list analysis

NCAA rank listed first, then my ranking in parentheses.

1 Villanova (Villanova)
2 Kansas (Kansas)
3 North Carolina (Gonzaga)
4 Gonzaga (Duke)
5 Kentucky (Arizona)
6 Arizona (North Carolina)
7 Duke (Kentucky)
8 Louisville (Baylor)
9 Oregon (Oregon)
10 Florida St (Louisville)
11 UCLA (UCLA)
12 Baylor (Butler)
13 Butler (Florida St)
14 Florida (Florida)
15 West Virginia (West Virginia)
16 Purdue (Notre Dame)
17 Virginia (Iowa St)
18 Minnesota (Purdue)
19 Notre Dame (Virginia)
20 Iowa St (SMU)
21 SMU (Cincinnati)
22 Cincinnati (Minnesota)
23 Maryland (Creighton)
24 Creighton (Michigan)
25 St Mary's (Wisconsin)
26 South Carolina (Maryland)
27 Michigan (Miami)
28 Dayton (Northwestern)
29 Wisconsin (St Mary's)
30 Miami (Virginia Tech)
31 Arkansas (Wichita St)
32 Northwestern (Seton Hall)
33 Vanderbilt (Arkansas)
34 Seton Hall (VCU)
35 Michigan St (Dayton)
36 Virginia Tech (Rhode Island)
37 Oklahoma St (Marquette)
38 Wichita St (Middle Tennessee)
39 Marquette (Providence)
40 VCU (Oklahoma St)
41 Xavier (South Carolina)
42 Providence (Michigan St)
43 Wake Forest (Xavier)
44 Rhode Island (Kansas St)
45 USC (Vanderbilt)
46 Kansas St (Wake Forest)
47 Nevada (USC)
48 Middle Tennessee (Nevada)
49 UNC-Wilmington (UNC-Wilmington)
50 Princeton (Princeton)
51 Bucknell (Vermont)
52 East Tennessee St (New Mexico St)
53 Vermont (East Tennessee St)
54 Winthrop (Bucknell)
55 New Mexico St (Winthrop)
56 FGCU (Iona)
57 Kent St (FGCU)
58 Iona (Northern Kentucky)
59 Northern Kentucky (Texas Southern)
60 Troy (Kent St)
61 Jacksonville St (Troy)
62 North Dakota (North Dakota)
63 Texas Southern (South Dakota St)
64 South Dakota St (Jacksonville St)
65 UC-Davis (New Orleans)
66 North Carolina Central (Mt St Mary's)
67 New Orleans (UC-Davis)
68 Mt St Mary's (North Carolina Central)

Biggest differences:
South Carolina 15 (NCAA 26, me 41) - this ranking was nonsensical
Vanderbilt 13 (NCAA 33, me 45) - you know, losses are supposed to count too
Middle Tennessee 10 (NCAA 48, me 38) - this is part of the general disrespect for mid-majors
Rhode Island 8 (NCAA 44, me 36)
Dayton 7 (NCAA 28, me 35)
Michigan St 7 (NCAA 35, me 42)
Wichita St 7 (NCAA 38, me 31)
Virginia Tech 6 (NCAA 36, me 30)
VCU 6 (NCAA 40, me 34)

Takeaways:
- The SEC's efforts to schedule up did pay off more than I thought it would.
- I don't know what to do with the A-10
- Losses don't matter as much to the NCAA given what Vandy and MSU got

I got 68/68 teams...I missed a team's overall rank by an average of 2.838 spots per team.

I wonder how 2.838 compares to other bracketologists.  If you want, give me your 1-68s and I can calculate.