Today's rankings, read with March eyes

The season started this morning. Here is the full preseason Top 25, translated into the only language March speaks: seed lines, trap games, and the unranked names that become Cinderellas.

One game is already in the books

Notre Dame beat Villanova 74–72 in Rome on the morning of Sunday, November 1, 2026 (Eternal City Tip-Off, Palazzo dello Sport). Both teams were unranked; Villanova had 6 votes in the poll.

Monday, November 2: more than 100 games, headlined by No. 1 Florida vs No. 14 Miami in Tampa and No. 5 Arizona vs No. 22 UCLA in Las Vegas.

Nobody's March fate moves this week. But the picture that March will rearrange starts here.

The full preseason Top 25

Others receiving votes: Missouri 167, Nebraska 70, Purdue 53, Ohio State 39, North Carolina 20, TCU 17, Saint Louis 13, VCU 12, Iowa 12, Virginia Tech 10, Indiana 8, Texas A&M 6, Villanova 6, Auburn 4, Marquette 3, Nevada 2, Grand Canyon 2, Utah State 2, Santa Clara 1.

Seven of the 25 are SEC teams, led by Florida at No. 1.

Michigan, which won the April 2026 national title, opens down at No. 16.

What these bands become in March

November ranks are not March seeds.

The committee seeds the field on the whole season, and four months of games will shuffle this list.

But the bands rhyme with the seed lines, and the seed lines carry 40 years of counted history (the full analysis).

Read today's list in bands:

• Ranks 1 to 4 (Florida, Duke, Illinois, UConn): the 1-seed line. One seeds have won 26 of 40 titles and 158 of their 160 first games. The safest seat in sports, with two famous exceptions.
• Ranks 5 to 8 (Arizona, Michigan State, Texas, Tennessee): the 2-seed line. Strong, not safe: 15 seeds have beaten 2 seeds 11 times, about once every four years.
• Ranks 9 to 12 (Virginia, Houston, Louisville, Arkansas): the 3-seed line. Three seeds lose to 14 seeds about 14 times in 100.
• Ranks 13 to 16 (St. John's, Miami, Gonzaga, Michigan): the 4-seed line. The first bullseye. Four seeds lose to 13 seeds 33 times in 160, about one time in five. The reigning champion, Michigan, sits in this band today.
• Ranks 17 to 20 (Kansas, Iowa State, Texas Tech, Alabama): the 5-seed line. The most famous trap in the tournament. Five seeds lose to 12 seeds 57 times in 160, better than one time in three — and no 5 seed has ever won the title.
• Ranks 21 to 25 (USC, UCLA, Kentucky, Vanderbilt, BYU): the 6 and 7 lines. Favorites in the two most upset-heavy games after the coin flip: 6 seeds lose to 11 seeds about 39 times in 100, and 7 seeds lose to 10 seeds about 39 times in 100.

Vulnerable favorites: where the voters and the model disagree

Two rankings came out this fall.

The voters published theirs, the AP poll above.

The best-known free model published its own: Bart Torvik's T-Rank, an efficiency rating the NCAA itself puts on the selection committee's team sheets.

Its preseason baseline, read on November 1, 2026, ranks all 365 teams before almost any of them has played a game.

Where the model rates a ranked team far below the voters, one of the two lists is wrong.

If the model is right, that team is a favorite standing on a trap line with less under it than its rank suggests.

These are the six biggest disagreements in the Top 25:

Read this the honest way: this is one model's preseason baseline, a projection, never a certainty.

It will be wrong about some of these teams by January.

That is why this page re-scores the list every week instead of carving it in stone.

The disagreement runs both ways:

• Kansas: voters No. 17, model No. 4. If the model is right, a top-line team is hiding in the 5-seed band today.
• Nebraska: unranked, model No. 15. The model's 15th-best team in the country did not make the voters' Top 25 at all, on 70 votes. Somebody moves toward somebody by January.
• Gonzaga: voters No. 15, model No. 8. The model likes the Zags more than the poll does, even on the 4-seed band where 13 seeds win one time in five.

Cinderella Radar: scoring the unranked

March's 11, 12, and 13 seeds are almost never ranked in November.

They are conference champions and fighters from leagues the polls barely watch.

Count the whole first round and about eight games a year go to the lower seed. About two double-digit seeds are still standing in the Sweet 16.

So this page scores the unranked names on three signals. All three are public, and you can check every one:

• Model strength. Torvik's model rates the team in the national top 70 even though the voters left it out. Teams that play above their seed line are where upsets come from.
• Poll notice. The team received AP votes. At least a few voters thought it belonged, which separates it from the 300 or so teams nobody mentioned.
• A live path. Its league's road to March lands teams on the 10 to 14 seed lines, where the counted upset rates run from about 14 to 39 in 100.

The radar is a watch list, not a promise: most of these teams will not make the field, and the page will say so when their signals go dark.

And the famous name both lists agree on: North Carolina. Twenty votes, no ranking, and the model puts the Tar Heels 47th too. A giant starts the season doubted by everyone.

What this page cannot see

Plain words about the limits, because a forecast that hides them is a sales pitch:

• One model, named every time. Torvik's T-Rank is free, public, and on the committee's own team sheets. Other respected ratings sit behind paywalls, or publish nothing in November that can be checked. One model you can verify beats three we cannot.
• Preseason means priors. In November, the model is rating last season's program, returning players, and history, not games this roster has played. Its disagreements with the voters are hypotheses, and this page treats them that way.
• No injuries, no roster news. This page does not track who is hurt or who transferred. By February, that will matter more than anything printed here, and the weekly refresh will say what changed.
• A rank is not a seed. The bands rhyme with the seed lines. The committee holds the real pen, in March.

Why a bracket site keeps a rankings page

A bracket is a set of choices, and the choices start forming now.

Every Monday, this page refreshes with the new poll, restamps its date, and re-reads the bands.

By March, the teams you have been watching all winter will land on seed lines whose history you already know.

That is the whole idea here: your team, your upsets, your calls — made with the counted record open next to them.