The 2017 Tournament’s Most Overseeded Teams Comment Count

Alex Cook March 14th, 2017 at 6:44 PM

32132954746_a9f4dcfa7f_zMaryland’s Mark Turgeon [Marc-Gregor Campredon – MGoBlog]

MGoBlog’s Tournament Sponsor is HomeSure Lending. NMLS 1161358.

March Madness is wonderful. Millions of people across the country will fill out a bracket, predicting winners for each game, hoping that they’ll be the one to pick the right upsets, and trust the right teams to make it to the Final 4. Of course, no amount of research is any guarantee of success, and part of the fun is the insane randomness that determines the course of these next few weeks. Still, becoming acquainted with the participants is key to making informed bracket choices, and I’ll be writing up a few posts with that in mind.

In order to sort out which teams are better or worse than where they’re seeded, I took the list of teams sorted by the committee 1-66, and I compared it to a composite computer ranking of tournament teams based on the average of Ken Pomeroy, Jeff Sagarin, and Bart Torvik’s metrics. I then calculated the difference between where a team should theoretically be ranked given their strength according to that composite ranking and where they actually wound up.

This is the resulting bracket:

seed difference

The seven most overseeded teams (which are clustered in just two regions) are

  • West #6 Maryland
  • West #2 Arizona
  • South #5 Minnesota
  • South #4 Butler
  • South #7 Dayton
  • South #10 Seton Hall
  • West #3 Florida St.

They’ll be discussed after the jump.

[JUMP]

Maryland

Seed Rank: 23 (6-seed). Computer Rank: 40 (10-seed). Round of 64 vs. Xavier (11); Round of 32 vs. Florida St. (3) / Florida Gulf Coast (14).

After years of computer rankings habitually overrating Bo Ryan’s Wisconsin team, they’ve started to habitually underrate Maryland: early-season success in close games – particularly in three one-point victories (two of which came against NCAA Tournament teams) – led to a wide gap between human and computer polls. The Terrapins started off the season with a 20-2 record and the advanced metrics were never particularly impressed; a very young team lost six of its last ten games, including a first round exit in a pseudo-home environment in the Big Ten Tournament against Northwestern.

Maryland the good fortune to wind up as the 6-seed in a region with a weak 2 (Arizona) and 3 (Florida State), and they have a veteran point guard – Melo Trimble –  leading the show; they also have three freshmen playing significant roles – Kevin Huerter, Anthony Cowan, Justin Jackson – and none of those freshmen are NBA players. Their first round opponent, Xavier, lost their star point guard Edmond Sumner to a season-ending injury, so an upset there might not happen, but the computer rankings are very down on the potential for Maryland to make a long run.

Arizona

Seed Rank: 6 (2-seed). Computer Rank: 21 (6-seed). Round of 64 vs. North Dakota (15); Round of 32 vs. St. Mary’s (7) / VCU (10).

I was surprised to see how little the various advanced stats metrics liked Arizona; the Wildcats went 30-4, tied for the Pac-12 regular season title, won the conference tournament, and only lost to teams that eventually earned protected seeds. U of A only finished with the third-best efficiency in league games, but they finished a seed line ahead of Oregon and UCLA. Most of the Pac-12 is thoroughly mediocre (even though there are potential Final Four teams at the top), so that probably adversely impacted Arizona’s statistical profile. The committee judged their resume fairly though, giving them a 2-seed out West, and manageable potential matchups against Florida St. and Gonzaga in the Sweet 16 and Elite 8.

As is custom for him in Tucson, Sean Miller has a young and talented team. Freshmen Lauri Markkanen, Rawle Atkins, and Kobi Simmons have played significant roles this season and sophomore Allonzo Trier has made an impact after a PED suspension kept him ineligible for most of the season. Markkanen is a likely one-and-done lottery pick from Finland and one of the most efficient players in the country – the 7-footer shoots 54% on twos, 43% on threes, and 82% from the line. Arizona’s solid, but not exceptional, on both sides of the floor, and boast one of the tallest teams in the country.

Minnesota

Seed Rank: 18 (5-seed). Computer Rank: 33 (9-seed). Round of 64 vs. Middle Tennessee St. (12); Round of 32 vs. Butler (4) / Winthrop (13).

It’s been quite the turnaround for Minnesota, which has improved its win total by 16 and finished with a protected seed in the Big Ten Tournament. Despite having a meh non-conference schedule, the Golden Gophers successfully gamed the RPI and find themselves a few seed lines higher than they should be. With Akeem Springs’s Big Ten Tournament injury, Richard Pitino was forced to go to a six-man rotation in the semifinals against Michigan; the Gophers have talent, but not depth, and the tactical acumen of Pitino is very much an open question. Minnesota finished with thee second-best seed in the Big Ten. They weren’t the second-best team in the Big Ten.

Before the injury to Springs late in the BTT win over Michigan St., Minnesota had been playing quite well, winning 9 of their last 10, including that game after a 5-game skid in January. Middle Tennessee St. will be a popular 5 / 12 upset pick; the Blue Raiders dominated their conference after returning most of their key players from a team that pulled a historic upset over Michigan St. in the tournament last season. MTSU is one of the best under-the-radar mid-major teams in the country and should be a tough out for the Gophers. Whoever wins that game, will likely play…

Butler

Seed Rank: 13 (4-seed). Computer Rank: 26 (7-seed). Round of 64 vs. Winthrop (13); Round of 32 vs. Minnesota (5) / Middle Tennessee St. (12).

…Butler, another team that’s overseeded. The Bulldogs had an early exit from the Big East Tournament, but they racked up some good wins over the course of the season: they swept #1 overall seed Villanova, defeated Arizona at a neutral site, beat Marquette on the road, and took care of Vanderbilt and Northwestern in the non-conference portion of the season. Butler did record a few losses against non-tournament teams, most notably a one-point loss to lowly Indiana State, but a mostly great non-conference showing and a second-place finish in the Big East was good enough to get them a 4-seed.

Junior wing Kelan Martin is the go-to guy for the Bulldogs, an inside-outside threat leading the team in shot attempts by a wide margin. He’s joined by Andrew Chrabascz on the wing, a versatile player who does most of his damage by drawing fouls. Sixth man freshman Kamar Baldwin provides energy and shot-making off the bench, and veteran point guard and North Carolina State transfer Tyler Lewis is a low-usage set-up man with one of the best individual offensive ratings in the country.

Butler’s opening round matchup (Winthrop) is much easier than Minnesota’s (MTSU), but if you’re looking for a spot to put a double-digit seed in the Sweet 16, the MTSU Blue Raiders might be a good choice – they’d face the equivalent of a 9- and 7-seed to get there.

Dayton

Seed Rank: 28 (7-seed). Computer Rank: 39 (10-seed). Round of 64 vs. Wichita St. (10); Round of 32 vs. Kentucky (2) / Northern Kentucky (15).

The Flyers have seen a considerable amount of NCAA Tournament success under head coach Archie Miller; they made it to the Elite 8 in 2014, won a First Four game and made it to the Round of 32 in 2015, and have made it to the tournament four seasons in a row. Dayton only lost one home game all season – to fellow mid-major stalwart St. Mary’s – and won the Atlantic 10 regular season title running away, but a season-ending loss to George Washington and a loss to Davidson in the first round of the conference tournament signal that the Flyers may be hitting a rough patch after a nine-game winning streak that extended from January to March.

Unfortunately for Dayton, they drew perhaps the worst first-round matchup in the entire field, another mid-major that’s experienced recent March success: Wichita St. The Shockers avoided the bubble by winning the Missouri Valley Tournament, but they wound up as a 10-seed after finishing tenth in the aggregate computer rankings. Wichita St. would play Kentucky in the Round of 32 and could be an intriguing Cinderella pick. At the very least, Dayton is deeply unlucky to see them in the 7 / 10 game.

Seton Hall

Seed Rank: 34 (9-seed). Computer Rank: 44 (11-seed). Round of 64 vs. Arkansas (8); Round of 32 vs. North Carolina (1) / Texas Southern (16).

Seton Hall is one of a handful of teams that played its way into the tournament late in the season; after a 22-point home loss to Villanova on February 18th, they were just 16-10 (6-8 Big East), but after a five-game winning streak and heartbreaking loss to the Wildcats in the Big East Tournament, the Pirates were safely in the field of 68. At one point, they lost five of six games, but they recovered over the second half of conference play and weren’t really at risk of missing the field by the end of the season. Still, the computer ranking composite considers the Pirates to be one of the worst at-large teams to have received a bid.

In any case, Seton Hall is unlikely to advance very far in the tournament. Their opponent in the 8 / 9 game (Arkansas) isn’t anything special, but North Carolina looms as the Round of 32 opponent and the Tar Heels’ size inside could pose problems for a team that only has one player over 6’6 – excellent junior big man Angel Delgado – among their five players with the most playing time. Even if they do beat Arkansas, getting to the Sweet 16 would be quite difficult since getting there means that the Pirates (who are a team with the quality of an 11-seed) would have to beat one of the national title favorites.

Florida St.

Seed Rank: 10 (3-seed). Computer Rank: 19 (5-seed). Round of 64 vs. Florida Gulf Coast (14); Round of 32 vs. Maryland (6) / Xavier (11).

After a four-year NCAA Tournament drought under longtime head coach Leonard Hamilton, the Seminoles exceeded expectations this season, winning 25 games and finishing tied for second in an ultra-competitive ACC alongside Louisville and Notre Dame. Florida St. has a highly touted outside scorer (sophomore Dwayne Bacon), a versatile defensive stopper (Xavier Rathan-Mayes, who also has an excellent assist-to-turnover ratio), a tantalizing one-and-done freshman (Johnathan Isaac) whose skill at 6’10 will get him drafted in the lottery, and a host of other spectacular athletes, including two rotation centers over 7-feet tall.

While FSU has cooled off as the season has gone on – they started off 18-2 with wins over Virginia, Duke and Louisville before finishing 7-6 down the stretch with few great wins – they landed in a nice spot: as the 3-seed in the West, they’d have to play Maryland and Arizona to make the Elite 8 if chalk holds – the two highest teams on this list. FSU only ranks 19th in the computer ranking, but Arizona is 21st. If you’re looking for a sleeper in that part of the bracket, 7-seed St. Mary’s could be a worthy choice.

Comments

SD Larry

March 14th, 2017 at 6:57 PM ^

not in your most overseeded list.  IMO, with just 19 wins, they should have been considered a bubble team, particularly when Illinois State was left out with 27 wins.  

 

Jangalang

March 15th, 2017 at 9:22 AM ^

do you play a tougher non conference scehdule and lose to everyone (Sparty) or do you play a cupcake non con and clean up?  Is it better to lose to the top teams or pad your wins with crappy RPI blowouts?

I think Sparty got it wrong by scheduling so many tough opponents.  They should have diversified a little more with their non conference schedule to get the win total up more.

Their ratio should have been like 3-9...top teams-cupcakes...

MH20

March 14th, 2017 at 7:06 PM ^

Man, I'd kill for Michigan to be in that West Region.  VCU sucks, Arizona is very beatable and Florida State is the weakest 3-seed.

MH20

March 15th, 2017 at 9:20 AM ^

Especially when paired with a 16-2 regular season record and a tournament championship in the Pac-12. I guess I'm just not overly impressed with the Pac-12 in general.  And KenPom agrees with me (top three teams are 16, 18, 21 -- Michigan is 20).

UMinSF

March 14th, 2017 at 7:48 PM ^

Unlike the other listed teams, I don't think Arizona is overrated. They lost to Gonzaga with both their guards out, and Trier missed almost the whole season (not to make excuses for him, he got busted for PED's). 

With Trier back, they've beaten all their top foes in the PAC, and I think they match up very favorably with Gonzaga.

I agree, how did MSU not make this list?

mi93

March 14th, 2017 at 9:30 PM ^

they looked good against Oregon and UCLA.  The computers clearly thinks the PAC is trash.  All three of them are overseeded per this perspective.  The eye test says otherwise, to me, in this case.

Scout96

March 14th, 2017 at 8:11 PM ^

People are asking why MSU is not on this list. It mentions it's only ranking the teams that made the tourney, a 10 seed is I think the last seed for a power 5 team to make the tourney, so he evaluated MSU as -2, which probably would have pushed them to a 10 seed, the bottom position for power 5 teams. It is just not one of the larger overseeding differences. pushing them out would been -8 at worst. FSU is last on his list at -9.

funkywolve

March 14th, 2017 at 11:11 PM ^

is a 10 seed and listed as one of the most over seeded teams.  The computers that Alex looked at and averaged all had MSU in the 9/10 seed range so at least in terms of what Alex looked at, MSU wasn't overseeded.

Caille33

March 15th, 2017 at 9:33 AM ^

Glad to see I'm not the only that picked those 2 upsets!  I wouldn't be surprised if MTSU beats Minnesota by double digits.  Don't know much about Winthrop but I have seen Butler a couple times and they don't impress me at all.

charblue.

March 14th, 2017 at 9:35 PM ^

while Illinois is playing Valparaiso tonight in the NIT having finished the regular year with an 18-14 record, which is exactly what MSU would have had had they not won their first Big Ten Tournament game. Not saying that the Illini and MSU are equal based on SOS or quality wins, but a one-win difference has netted Sparty a far greater return than many other tournament teams. And also conisder that they get to play Friday night instead of as the opening tip contest of the tournament, the timeslot which Michigan drew with five more wins and two wins apiece agaist the conference's top teams, Purdue and Wisconsin.

J.

March 14th, 2017 at 10:58 PM ^

I have as much disdain for Staae as anyone, but the entire reason that we have tens of different computer ratings is that comparing teams by record is insufficient.

You say "[You're] not saying that [they're] equal...," but then you go on to claim that MSU got into the NCAAs because of a one-game difference in its record.

MSU got in because of a much stronger SOS.  Your Illinois record includes a win over Division II McKendree University.  The best Illini win is the Sir Maverick Morgan game, and they lost to both Winthrop and Rutgers and got swept by PSU.  MSU beat Wichita State (which apparently only mattered when seelcting the teams, not when seeding them), beat Michigan -- but also beat Wisconsin and swept Minnesota.

The bubble was very weak this year, but not so weak that KenPom #70 Illinois should have gotten a bid.

charblue.

March 15th, 2017 at 10:15 AM ^

You still have to win in order to get in. For crying out loud,How many other at-large teams got in with fewer than 20 wins?  Illinois fired their coach. And Michigan was on the bubble at 19 wins.

The point is, Sparty got in with a 19-14 record and a 9th seed based on reputatipm and their AD serving as tournament director. But in addition to getting a 9th seed against an ACC team which actually beat some of the reams that Sparty got credit for schoduling but losing to. And they nearly lost at home to Florida Gulf Coast in the preseason. The point is Sparty got every advantage including a late Friday tournament tip time compared to the opening tip time that Michigan got in the same Midwest bracket.

I'm not comparing technical data. I'm just making an observation about how some schools get treated with more respect than others. 

funkywolve

March 14th, 2017 at 10:08 PM ^

When the bracket came out I thought UNC had an easy draw and this kind of confirms it. Lots of people are talking about UK, UCLA and UNC being in the same region but with that draw it would be a big upset if UNC isn't in the Elite 8.

oriental andrew

March 14th, 2017 at 10:23 PM ^

We've heard that they rely quite a bit on RPI, so did a few calcs on most teams (quit once I got into the small conference auto-bids). 

I've highlighted a few below. Negatives indicate underseeded (NCAA seed lower than RPI), positive indicates overseeded (NCAA seed higher than RPI). 

RANK SCHOOL W-L RPI NCAA Seed Over/Under
1 Villanova 31-3 0.665 1 0
2 Arizona 30-4 0.655 6 -4
3 Kansas 28-4 0.653 2 1
4 Kentucky 29-5 0.652 5 -1
5 North Carolina 26-7 0.647 3 2
6 Duke 27-8 0.645 7 -1
7 Louisville 24-8 0.645 8 -1
8 Gonzaga 32-1 0.644 4 4
9 Oregon 28-5 0.643 9 0
10 Florida 24-8 0.639 14 -4
11 Baylor 24-7 0.637 12 -1
12 Cincinnati 29-5 0.631 22 -10
13 SMU 30-4 0.631 21 -8
14 Florida State 25-8 0.631 10 4
15 Butler 23-8 0.625 13 2
16 UCLA 29-4 0.624 11 5
17 St. Mary's 28-4 0.621 25 -8
18 Virginia 22-10 0.617 17 1
19 Purdue 25-7 0.613 16 3
20 Minnesota 24-9 0.611 18 2
21 Iowa State 23-10 0.61 20 1
22 VCU 26-8 0.61 40 -18
23 Notre Dame 25-9 0.609 19 4
24 West Virginia 26-8 0.605 15 9
25  Michigan 24-11 0.604 27 -2
26 Creighton 24-9 0.603 24 2
27 UNC-Wilmington 27-5 0.602 49 -22
28 Arkansas 25-9 0.602 31 -3
29 Nevada 28-6 0.601 47 -18
30 Dayton 23-7 0.601 28 2
31 Rhode Island 24-9 0.599 44 -13
32 Wichita State 29-4 0.599 38 -6
34 Maryland 23-8 0.596 23 11
35 Middle Tennessee 29-4 0.595 48 -13
36 Wisconsin 25-9 0.593 29 7
37 Xavier 21-13 0.593 41 -4
38 Vanderbilt 19-15 0.591 33 5
39 Wake Forest 19-13 0.589 43 -4
40 Oklahoma State 19-12 0.588 37 3
41 Southern California 24-9 0.588 45 -4
42 Miami (Fla.) 21-11 0.587 30 12
43 South Carolina 21-10 0.587 26 17
44 Seton Hall 21-11 0.586 34 10
46 Vermont 28-5 0.584 53 -7
47 Princeton 22-6 0.584 50 -3
48 Virginia Tech 22-10 0.582 36 12
50 Michigan State 19-14 0.579 35 15
51 Northwestern 23-11 0.578 32 19
55 E. Tennessee State 25-7 0.571 52 3
56 Providence 20-12 0.571 42 14
57 Kansas State 20-13 0.57 46 11
59 New Mexico State 25-5 0.568 55 4
61 Marquette 19-12 0.566 39 22
63 Bucknell 26-8 0.564 51 12

Teams most underseeded include:

  • UNC Wilmington (-22) - Going by RPI of 27 puts them at a 7 seed vs. the 49 the committee gave them
  • VCU (-18) - RPI of 22 equates to a 6 seed vs. the 40 from the committee
  • Nevada (-18) - RPI of 29 equates to an 8 seed vs. 47 from committee
  • Rhode Island (-13) - RPI of 31 equates to 8 seed vs. 44 from committee
  • MTSU (-13) - RPI of 35 equates to 9 seed vs. 48 from committee 

Teams most overseeded include: 

  • Marquette (+22) - RPI of 61 equates to a 16 (!) seed vs. 39 from the committee
  • Northwestern (+19) - RPI of 51 equates to a 13 seed vs. 32 from committee
  • S Carolina (+17) - RPI of 43 equates to 11 seed vs. 26 from the committee
  • michigan state (+15) - RPI of 50 equates to 13 seed vs. 35 from committee
  • Providence (+14) - RPI of 56 equates to 14 seed vs. 42 from committee

Michigan's RPI is 25 vs. NCAA seed rank of 27, so slightly under but not enough to impact our 7 seed. 

Okie State RPI is 40 vs. NCAA seed rank of 37, so slightly over and would have been an 11 seed. 

Going strictly by RPI, the potential 10 seeds Michigan would've played are Wisco, Xavier, Vandy, and Wake. 

J.

March 14th, 2017 at 10:28 PM ^

The RPI is terrible.  Any indication that they're not seeding via the RPI is a good thing.

Not only does the RPI ignore margin of victory and overrate home losses, it weights your opponents' opponents records exactly as your team's own record.  It was put together using the technology that was available in the early 80s and hasn't had a meaningful update since.  Alex's approach of averaging three computer rankings passes my personal sniff test better than the RPI does.

oriental andrew

March 14th, 2017 at 11:27 PM ^

Which is that the perception is that the committee strongly values the RPI, and the delta between RPI and the actual tournament seedings. I never made any value judgement about the quality of the RPI. 

At any rate, it looks like Alex's method and RPI both suggest that teams like Seton Hall and Maryland are very overseeded, while Wichita State and WVU are underseeded, although there is more agreement with the RPI comparison than there is with the computer composite comparison.

J.

March 14th, 2017 at 11:32 PM ^

No offense, but I think you're missing your own point.  The seedings are obviously the definitive word as to what the committee values; so, to the extent that teams are overseeded or underseeded by the RPI, that's a repudiation of your theory that the committee values it highly.

Furthermore, if the RPI isn't meaningful, then noting which teams are overseeded or underseeded according to it is inherently meaningless as well.  For example, both MIchigan and MSU are highly underseeed by the metric "distance from Ann Arbor, MI," but that's obviously not useul for determining who is likely to win an NCAA tournament game. :-)

oriental andrew

March 15th, 2017 at 6:50 AM ^

you just restated my point - that it's basically a test to see how closely the committee aligns to RPI (not MY theory, as its been talked about quite a bit here and elsewhere). in some cases, very much so. in others, not so much. The final seed list doesn't tell us WHY they're in that order. The theory is that the RPI is a big driver.

this is as "meaningless" an exercise as the computer composite comparison. I don't know why you're fighting it so hard.

(And for the record, I'm not the guy who's been down voting you)

mGrowOld

March 15th, 2017 at 11:58 AM ^

Forget RPI....forget BMI......forget Kenpom and for sure forget all  those new-fangled "fancy stats".  We know what we see and we use the eye-test for seeding.  It's worked the past 60 years for us and it'll work this year too!

Image result for group of very old men