Skip to main content
Menu
It’s Crowded At The Top Of The NCAA Tournament

FiveThirtyEight’s March Madness predictions are up and ready for your perusal. But while we’re really excited about the tournament, we have to say this is not a year in which we’re going to be able to help you all that much, at least in the Elite Eight and beyond.

That’s because there’s as much parity in the tournament as there’s ever been — not necessarily from the very top of the field to the very bottom, but certainly among a group of No. 1 and No. 2 seeds that aren’t all that easy to distinguish from one another. Meanwhile, the blue bloods have to navigate a minefield of underseeded teams such as Wichita State and SMU, with some having more perilous paths than others.

The one team that potentially stood out from the pack — defending national champion and No. 1 overall seed Villanova — has been undermined by a difficult draw. While the Wildcats are still the nominal favorite to win the tournament, they have only a 15 percent chance of doing so, which is tied for the lowest probability for a frontrunner in our seven years of making tournament predictions.1

Our methodology for making these projections is exactly the same as it was last year. They’re based on a blend of six computer rankings: FiveThirtyEight’s Elo ratings, Ken Pomeroy’s ratings, Jeff Sagarin’s “predictor” ratings, ESPN’s BPI, Joel Sokol’s LRMC ratings and Sonny Moore’s computer power ratings. We also use two human-generated rating systems: the selection committee’s 68-team “S-Curve” and a composite of preseason ratings from coaches and media polls. The preseason ratings have some predictive power when used carefully, serving as a hedge against teams that may have overachieved or underachieved relative to their talent level and are due to revert to the mean.


VIDEO: A No. 16 seed will win, but don’t bet on it


Our ratings also account for injuries and travel distance — playing in a familiar gym an hour from campus is a big advantage compared to traveling across the country. And once the tournament begins, the ratings will be updated to reflect the results of previous tournament games. If a No. 2 seed needs overtime to defeat a No. 15 seed, for instance, it can be an inauspicious sign for their prospects down the road.

Let’s take a quick tour of the four regions — starting with Villanova and the East.

RANK AMONG 68 TOURNAMENT TEAMS
SEED TEAM ELO POM SAG BPI LRMC MOOR NCAA PRE
1 Villanova 1 2 2 1 2 1 1 4
2 Duke 4 12 8 6 8 6 7 1
3 Baylor 19 13 16 17 17 20 12 28
4 Florida 16 9 10 8 9 14 14 29
5 Virginia 20 7 7 4 10 10 17 7
6 SMU 9 11 19 19 14 15 21 22
7 South Carolina 50 31 30 38 34 45 26 39
8 Wisconsin 26 23 17 18 22 9 29 9
9 Virginia Tech 37 41 42 45 43 37 36 42
10 Marquette 31 28 29 28 30 30 39 46
11 Providence 40 47 47 48 48 42 42 37
11 USC 48 50 46 50 52 46 45 34
12 UNC-Wilmington 42 49 49 51 44 52 49 47
13 E. Tenn. St. 52 52 52 52 50 53 52 57
14 NM State 56 54 54 55 54 63 55 50
15 Troy 55 58 60 57 59 58 60 65
16 New Orleans 65 64 67 68 68 67 67 68
How the teams compare: East region

ELO = FiveThirtyEight’s Elo rating; POM = Ken Pomeroy; SAG = Jeff Sagarin “predictor” rating; BPI = ESPN’s basketball power index; MOOR = Sonny Moore power ratings; NCAA = Selection committee’s 68-team “S-Curve”; PRE = Composite of preseason rankings

Villanova, the selection committee’s top seed, also ranks well according to the six computer systems, all of which place it first or second overall. But it got absolutely no help from the committee, who stuffed the East region with tough teams. Duke — which began as the preseason No. 1, slumped in the middle of the season, and then recovered to win the ACC tournament this weekend — engenders some disagreement among the various ratings systems but is no pushover as No. 2 seed, to say the least. That sets up a possible grudge match between Villanova and Duke, the past two tournament champions, in the East regional final at Madison Square Garden.

But Villanova has their work cut out to get there. Wisconsin, the No. 8 seed and ‘Nova’s potential opponent in the Round of 32, should have been seeded several slots higher according to the computers and has a history of tough postseason play. No. 4 Florida and No. 5 Virginia also get a lot more respect from the computers than they did from the committee, with blemished win-loss records masking difficult schedules and strong defenses.

On Duke’s side of the bracket, No. 6 SMU could be a tough out and is about even-money in its potential Round of 32 matchup against No. 3 Baylor. But if you’re looking for a first-round upset, there aren’t any slam-dunk candidates. No. 7 South Carolina is probably no better a team than No. 10 Marquette, but they have the advantage of playing a de facto home game in Greenville, S.C.

RANK AMONG 68 TOURNAMENT TEAMS
SEED TEAM ELO POM SAG BPI LRMC MOOR NCAA PRE
1 Kansas 2 10 9 9 7 8 2 2
2 Louisville 10 6 6 5 11 3 8 12
3 Oregon 8 16 13 13 16 24 9 5
4 Purdue 18 15 12 11 12 13 16 14
5 Iowa State 15 17 15 23 15 11 20 21
6 Creighton 25 27 27 27 27 29 24 18
7 Michigan 12 21 22 20 21 22 27 33
8 Miami 27 32 28 28 40 36 30 23
9 Michigan St. 44 40 35 38 39 38 35 10
10 Okla. State 29 24 23 26 26 25 37 44
11 Rhode Island 30 36 40 31 41 28 44 20
12 Nevada 34 46 45 42 42 47 47 52
13 Vermont 45 51 51 45 49 50 53 53
14 Iona 61 57 57 58 55 54 58 48
15 Jax. State 63 62 64 62 64 68 61 66
16 UC-Davis 64 68 65 67 67 66 65 62
16 NC Central 66 61 62 61 61 65 66 67
How the teams compare: Midwest region

ELO = FiveThirtyEight’s Elo rating; POM = Ken Pomeroy; SAG = Jeff Sagarin “predictor” rating; BPI = ESPN’s basketball power index; MOOR = Sonny Moore power ratings; NCAA = Selection committee’s 68-team “S-Curve”; PRE = Composite of preseason rankings

As a Michigan native, I’m a firm believer that Midwest is best, but that doesn’t extend to this year’s tournament, where the region is a bit lackluster. Kansas — with a 28-4 record from a very good Big 12 conference — might seem like a no-brainer No. 1 seed. But other than Elo, the computer ratings are a bit down on the Jayhawks, noting that they won many games by narrow margins that could reflect luck as much as clutch performance. Still, they’re the favorite here. Louisville is a perfectly adequate No. 2 seed, but they have to survive a difficult Round of 32 game — both No. 7 Michigan and No. 10 Oklahoma State are underseeded. And the regional finals will be played in Kansas City, advantaging KU.

A lot of people are also pointing toward Kansas’s Round of 32 matchup — against either No. 8 Miami or No. 9 Michigan State — as a difficult spot. But as much as it pains me to say this as an East Lansingite, I’m not sure this is Michigan State’s year. They’ve won only five road or neutral-site games all year, and while a lot of credit should be given to the Spartans for how many tough teams they scheduled, they also didn’t come away with a lot of wins. Then again, Tom Izzo has made a fool of everyone else’s best-laid plans and projections in the past.

No. 3 seed Oregon is another case of a team that might seem to be underseeded but probably isn’t. They lost senior forward Chris Boucher to an ACL tear in the Pac-12 tournament, substantially weakening their front-court depth.

RANK AMONG 68 TOURNAMENT TEAMS
SEED TEAM ELO POM SAG BPI LRMC MOOR NCAA PRE
1 UNC 6 3 3 3 4 4 3 6
2 Kentucky 3 4 5 10 6 5 5 3
3 UCLA 14 18 14 14 13 16 11 15
4 Butler 23 26 26 25 24 23 13 27
5 Minnesota 33 33 34 36 35 35 18 55
6 Cincinnati 21 22 20 15 18 18 22 25
7 Dayton 41 35 36 34 46 44 28 38
8 Arkansas 32 37 37 40 37 27 31 41
9 Seton Hall 28 45 44 47 47 33 34 30
10 Wichita St. 11 8 11 15 5 12 38 24
11 Wake Forest 43 30 33 30 29 32 43 49
11 Kansas State 51 29 31 32 28 34 46 36
12 Mid. Tenn. 38 43 48 40 32 48 48 51
13 Winthrop 54 56 56 55 57 56 54 58
14 Kent St. 57 59 58 58 58 57 57 59
15 N. Kentucky 60 60 59 63 60 60 59 63
16 Texas Southern 67 66 66 63 62 64 63 60
How the teams compare: South region

ELO = FiveThirtyEight’s Elo rating; POM = Ken Pomeroy; SAG = Jeff Sagarin “predictor” rating; BPI = ESPN’s basketball power index; MOOR = Sonny Moore power ratings; NCAA = Selection committee’s 68-team “S-Curve”; PRE = Composite of preseason rankings

UNC and Kentucky are the top two seeds in the South — and are a tossup to advance from the region according to our forecast (each one has a 30 percent chance). Working in the Tar Heels’ favor: they played one of the toughest schedules in memory, rank slightly ahead of Kentucky in most of the computer rankings (although not in Elo) and are on the opposite side of the bracket from Wichita State, which was ridiculously underseeded. (The Shockers ranked in the top 10 overall according to two computer rankings, LRMC and Pomeroy, and yet they are just a No. 10 seed in their region according to the committee.) Helping Kentucky: the Wildcats are a bit healthier than UNC, come in hotter (having won 11 games in a row and the SEC tournament), might have a bit more top-level talent (as reflected in their preseason ranking) and the location of the regional final, in Memphis, could be slightly favorable to them.

If it’s not UNC or Kentucky, the obvious alternative to emerge from the South is No. 3 UCLA, but the various computer rankings are not as bullish on the Pac-12 as the conventional wisdom seems to be. The South also features perhaps the best bet for a 5-vs-12 upset: No. 12 Middle Tennessee, which knocked off Michigan State last year, has a roughly 50-50 shot of beating No. 5 Minnesota, according to our forecast.

RANK AMONG 68 TOURNAMENT TEAMS
SEED TEAM ELO POM SAG BPI LRMC MOOR NCAA PRE
1 Gonzaga 7 1 1 2 1 2 4 12
2 Arizona 5 19 21 24 20 21 6 11
3 Florida State 22 19 18 21 19 17 10 35
4 West Virginia 13 5 4 7 3 7 15 17
5 Notre Dame 17 25 24 22 23 19 19 31
6 Maryland 49 42 41 42 45 41 23 19
7 St. Mary’s 24 14 25 12 25 26 25 16
8 Northwestern 47 38 39 42 33 43 32 40
9 Vanderbilt 39 34 38 36 31 39 33 26
10 VCU 36 44 43 35 38 40 40 32
11 Xavier 46 39 32 33 36 31 41 7
12 Princeton 35 48 50 48 53 49 50 43
13 Bucknell 53 53 53 54 51 51 51 54
14 FGCU 59 55 55 53 56 55 56 56
15 North Dakota 62 63 61 60 66 59 62 61
16 South Dakota St. 58 65 63 65 65 62 64 45
How the teams compare: West region

ELO = FiveThirtyEight’s Elo rating; POM = Ken Pomeroy; SAG = Jeff Sagarin “predictor” rating; BPI = ESPN’s basketball power index; MOOR = Sonny Moore power ratings; NCAA = Selection committee’s 68-team “S-Curve”; PRE = Composite of preseason rankings

In contrast to recent seasons, when the committee was often forced to ship in teams from the eastern half of the country to claim the top two seeds in the West, this year’s bracket features two geographically appropriate representatives in No. 1 Gonzaga and No. 2 Arizona. That’s important to keep in mind, because travel can be a big factor in the West region, with teams potentially flying a long way to play games in an unfamiliar time zone.

Otherwise, the West is a bit … weird. For one thing, Gonzaga — with a 32-1 record largely assembled outside a top conference — isn’t an easy team to get a handle on. A few of the computer ratings have the ‘Zags as the No. 1 overall team, while the committee’s rankings — and Elo — are more skeptical of them because they beat up on middling opposition. No. 2 Arizona, meanwhile, isn’t well liked by the computers, with its strength of schedule not as good as you’d expect from a 30-4 Pac-12 champion. Michigan State, whom the Wildcats beat to start their season, was Arizona’s best nonconference win.

West Virginia, the No. 4 seed, is another highly divisive team. They lost eight games but those losses came by an average of just 4.6 points, so the computer formulas think they’re badly underseeded. It’s a tough break for the Mountaineers to not only have drawn Gonzaga in the Round of 16 but also to have to fly across the country to play the game.

The West also features a number of potential spoilers from strong academic schools — Northwestern, Princeton, Vanderbilt — that will probably create a lot of buzz if they win. (All those Northwestern journalism grads won’t hurt.) None of those teams are especially underseeded, however. A better dark-horse pick is No. 7 St. Mary’s, which could give Arizona fits in the Round of 32.

Check out our March Madness predictions.

Footnotes

  1. Fourth-seeded Louisville also had a 15 percent chance in 2014.

Nate Silver is the founder and editor in chief of FiveThirtyEight.

Comments