— ?Power ratings are not rankings. They measure how strong each team is right now relative to an average D1 team per 100 possessions, so any two teams can be compared, turned into a line, and converted to a win probability. A team can be rated above a team with a better record. Rankings judge résumés (who you’ve beaten) and will come later in the season in the Rankings tab. ?What counts as an upset. A team the Meech line had losing by 15 or more points, before tip, that won the game outright. At 15 points the Meech line gives the underdog about a 10% chance (11% in women’s), so these are genuine shocks without the list going empty for weeks. Swing = how many points better than the line the winner played. Only games with a logged Meech line count, so the list starts Nov 4, 2026.?How home court is measured. For each of the last four full seasons (the COVID year is skipped), every team gets a full-season rating. Then we check how each team did in its home games compared with what its rating said, and the same for its road games. A team’s raw home edge is its home result minus its road result. That raw number is noisy, so it’s pulled toward the league average: a team’s own history counts as n / (n + 200) of its edge, where n is the home games measured. A team with 60 home games keeps about 23% of its own edge; the rest comes from the league average. Altitude teams rise to the top on their own: the model is never told anything about altitude.
The Board ?The table below shows every D1 game today with the Meech line. The Vegas line is also shown from that team's perspective (a “+” number means Vegas has them as the underdog). Lines are locked before tip and never edited.
| ?Meech line minus the Vegas line. A negative number means the Meech model likes the favorite more than Vegas does; a positive number means it likes them less. |
|---|
No games match this filter.
| ?Points per 100 possessions better than an average D1 team on a neutral floor. Offense minus defense. This is what determines the overall ranking. | ?Points scored per 100 possessions above an average D1 offense, adjusted for who the team played. | ?Points allowed per 100 possessions compared to an average D1 defense, adjusted for who the team played. Lower is better. | ?Possessions per 40 minutes the team plays at against an average opponent. | ?Net rating before any games played. This was built from last season’s results, the roster, the conference, and recruiting rankings. | ?How far the Net rating has moved from its preseason number. | ?Share of last season’s minutes that are back on this year’s roster. | ?Points advantage this team gets when playing at home: the league average plus its own home edge from the last four seasons. |
|---|
No teams match this filter.
Conference Strength ?Average team rating in each conference. Hover a bar for the best team and team count.
In my opinion, teams should not be ranked until after the Thanksgiving tournaments. In contrast to power ratings, rankings judge résumés, and in early November nobody has a true one. The first Meech rankings will drop December 1, once teams have played real opponents. Until then, the power ratings show who is playing strongest.
Pick two teams and hit Predict.
This part of the Meech Model is purely experimental. All season we track two things for every team: the pace it likes to play at, and how good it is at forcing games into that pace (pace control). The question we’re testing is whether owning the pace actually helps you win. Do teams that drag opponents to their speed beat expectations, or do the teams that adapt to whatever speed the other team wants have the edge? None of this goes into the Meech lines yet. Five past seasons say the effect is small at best, and it didn’t point the same way for men’s and women’s, so it is research-focused only for now.
Style Map ?Every team by speed (left = slow, right = fast) and pace control (up = drags games to its own speed, down = adapts to the opponent). Control is this season only, measured from games against teams that want a different speed, so it’s jumpy early in the season. Click a dot for the team page.
Fastest Teams
| Rk ?Pace rank: 1 = fastest team in D1. | Team | Conf | Pace ?Possessions per 40 minutes against an average opponent. This is what the Meech lines use. | Actual ?Average possessions per 40 minutes in their games so far, against the opponents they actually played. | Swing ?How much their pace moves from game to game (standard deviation, in possessions). Lower = more consistent. | Control ?When the two teams want different speeds, how far the game ends up toward this team’s speed. 50% = meets in the middle; higher = drags games to its own speed. This season only; faded until a team has 5+ games against a team wanting a different speed. | GP ?Games played vs. D1 teams. |
|---|
No teams match this filter.
Slowest Teams
| Rk ?Pace rank: 1 = fastest team in D1. | Team | Conf | Pace ?Possessions per 40 minutes against an average opponent. This is what the Meech lines use. | Actual ?Average possessions per 40 minutes in their games so far, against the opponents they actually played. | Swing ?How much their pace moves from game to game (standard deviation, in possessions). Lower = more consistent. | Control ?When the two teams want different speeds, how far the game ends up toward this team’s speed. 50% = meets in the middle; higher = drags games to its own speed. This season only; faded until a team has 5+ games against a team wanting a different speed. | GP ?Games played vs. D1 teams. |
|---|
No teams match this filter.
Five-Year Dictators ?Teams that dragged games toward their own speed in at least 4 of the last 5 seasons (2021-22 to 2025-26).
| Team | Seasons ?Seasons above 50% control, out of the seasons the team played. | Avg control ?Average pace control across those seasons. |
|---|
Five-Year Chameleons ?Teams that adapted to the opponent’s speed in at least 4 of the last 5 seasons.
| Team | Seasons ?Seasons below 50% control, out of the seasons the team played. | Avg control ?Average pace control across those seasons. |
|---|
| ?Points this team gets at home in a Meech line. Blue = at least a point above the league average. | ?Approximate seating, matched from Wikipedia’s list of D1 arenas. An estimate: listed capacities aren’t always current, and a few schools match a different arena they also use. | ?Elevation of the arena’s city, in feet. | ?This season’s record in true home games vs. D1 teams. | ?This season’s record in true road games vs. D1 teams. |
|---|
No teams match this filter.
Arena capacities: Wikipedia, “List of NCAA Division I basketball arenas” (CC BY-SA 4.0). Elevations: Open-Meteo geocoding.
| Date | Winner | Loser | Meech ?How many points the Meech line had the winner losing by, before tip. | Win % ?The winner’s chance to win, per the Meech line. | Vegas ?The Vegas spread for the winner (closing line when we have it). + = Vegas also had them as the underdog. | Swing ?Points better than the Meech line said: the underdog spread plus the winning margin. A 20-point underdog that wins by 5 = 25. |
|---|
No graded games yet. Results post the morning after each game.
Miss by Month ?Average points between the line and the final margin. Blue = Meech on every graded game. Light blue = Meech on only the games that had a Vegas line, and gray = Vegas on those same games, so those two are a fair head-to-head. Dashed orange = the 2025-26 test result.
Calibration ?When the model gave the favorite this chance to win, how often did they actually win?
| Model said | Games | Avg said | Won |
|---|
Fills in as games are graded.
Game Log ?Every graded line, newest first, with each team’s final score next to its name. Blue team = the Meech favorite. Closer = whose line landed nearer the final.
| Date | Away | Home | Meech | Vegas | Win % | Miss | Vegas miss | Winner | Closer |
|---|
Fills in as games are graded.
—
Schedule ?Every game this season. Played games show the line logged before tip. Upcoming games show what the Meech line would be if the game were played today. C = conference game. N = neutral site.
| Date | # | Opponent | Meech ?The Meech line from this team’s side: minus = favored, plus = underdog. Blue = this team is the Meech favorite. | Win % ?This team’s chance to win. Played games: as logged before tip. Upcoming games: as rated today. Never shown as 0% or 100%. | Result |
|---|
Rating Trend ?This team’s power rating at the end of each day this season. Dashed line = where it started (preseason). Hover a point for the date, rating and rank.
Ratings
Leaders ?Per-game averages this season for players who’ve played in at least half of the team’s games, sorted by points.
| Player | GP | PTS | REB | AST | MIN |
|---|
No stats yet.
Standings ?Ordered by conference record once conference games start (games above .500, then wins). Until then, and to break ties, ordered by Net rating.
| Rk | Team | Conf | Overall | Net ?Points per 100 possessions better than an average D1 team on a neutral floor. | Offense | Defense | D1 rank ?Where the team ranks among all D1 teams in the Meech power ratings. |
|---|
Top Players ?The conference’s top 10 scorers this season (per game), among players who’ve played in at least half of their team’s games.
| Player | Team | PTS | REB | AST |
|---|
No stats yet.
Vs. Other Conferences ?This season’s record against every other D1 conference, counting home, road and neutral-site games. Most of these games happen in November and December.
Conference Strength ?Average team rating in each conference. This conference is highlighted. Click any bar to open that conference.
The Meech Model rates every Division I men's and women's basketball team and puts out a line for every game before tip-off. Every line is locked before tip, never edited, and graded against the final score and against Vegas’s closing line. The point is twofold:
- To see if a model built from scratch can hang with the numbers and ranking systems everybody else uses.
- Create a model that uses the same engine to rate both NCAA men's and women's teams in a similar manner in order to produce lines.
Rating every team
Game data must be normalized per 100 possessions. A team that scores 85 in a fast-paced game and a team that scores 65 in a slow-paced grinder can be equally good. Each team gets three numbers, all adjusted for who they played:
Points scored per 100 possessions above an average D1 offense.
Points allowed per 100 above an average D1 defense. Lower is better.
Possessions per 40 minutes the team plays at against an average opponent.
Net rating = offense − defense. That's the power rating.
Where a team starts
Before anyone plays, each team gets a starting rating built from last year's team rating, its strength of conference, last season's production from returning players and transfers, and recruiting rankings for incoming freshmen. That starting point counts for about six games' worth of evidence for men's teams and four games' worth for women's, then this season's results take over.
Conferences move together
Early in the season, most teams have played only a couple of tough opponents. When a whole conference keeps beating expectations, or keeps getting blown out, every team in it moves slightly. Once conference play starts, that effect will lock in place.
Making the line
- Matchup: each offense against the other defense, at the pace both teams play.
- Home court advantage: every team gets its own specific home edge from four seasons of home-vs-road results. Most are 3 to 4 points, but some high altitude teams like Utah, BYU and Colorado are over 5. Neutral sites get none.
- Buy games: in November and December, a Power-conference team hosting a smaller school gets a small boost, because historic data says those game lines often underestimate the Power 5 team.
- Win %: how often a favorite of that size actually wins, based on how far real games land from the line. A game with a higher spread, meaning one team is a heavy favorite, will have a higher win % for the favorite.
How every piece earned its spot
There are many inputs that can be placed into models of this kind. In order for an idea to be incorporated into The Meech Model, it was tested using the 2021-22 through 2024-25 seasons as test data. Then, it was graded on the 2025-26 season to remove any training bias. If an idea didn't make the lines more accurate, the idea was not used as an input in the model. One exception: after that check, the buy-game boost was refit using all five seasons, including 2025-26, so this season’s live track record is the true test.
On the 2025-26 grading season the model missed the final margin by about 9.1 points per men's game and 9.7 per women's game, picking the winner 72% and 77% of the time.
Power ratings vs. rankings
Power ratings and rankings answer different questions. Power ratings ask who is strongest right now. These ratings are what the lines come from, and a team can be rated above a team with a better record. Rankings ask who's earned the most, based on who you've played and beaten this season.
In my opinion, teams should not be ranked until after the Thanksgiving tournaments. In early November, nobody has a true résumé, so the first Meech rankings will drop December 1. How they're built will be explained here when they launch.
What's different here than other rating models
- Transparent, team-specific home court instead of one number for everybody.
- Conference-aware ratings that adjust fast in November.
- Women's basketball, rated and graded using the same engine as men's.
- A public, never-edited track record against Vegas.
- Pace control: which teams make opponents play their speed and which ones adapt. This project is in progress, and may be a future input into the Meech Model.
For entertainment and informational purposes only. This is not betting advice. Lines are model estimates and carry real uncertainty, and past results don't guarantee future accuracy. Not affiliated with the NCAA, any conference, school, or sportsbook. Game data from ESPN via sportsdataverse; recruiting rankings from CollegeBasketballData.com (247Sports composite), ESPN, On3 and 247Sports; Vegas lines as shown on ESPN. If you or someone you know has a gambling problem, call 1-800-GAMBLER.