Methodology

What the ranking answers

The D2CFB Power Rankings answer "who is best right now?" first and "who has earned it?" second. Each team's D2CFB Rating is a fixed 70/30 blend: 70% a power rating built from scoring margin, 30% strength of record built from wins and losses. Both are computed only from finals stored on this site, and the rules below are published in full so anyone can check a number.

Power rating (70%)

Every final says "home team's rating, plus a home-field edge, minus the away team's rating, equals the margin". We find the ratings that fit all finals at once (least squares), so a week 1 result counts as much as a week 9 one and every game informs every rating.

Margins are capped at 24 points, so running up the score earns nothing, and a win always counts as at least 7 points. The home-field edge is fitted from this season's games and held between 0 and 4 points. Ratings are centred so the average Division II team is 0: a power rating of +10 means ten points better than an average Division II team on a neutral field.

The early-season prior

With three or four games played, results alone produce odd rankings. Each Division II team therefore starts from the AFCA preseason poll: No. 1 at +21 points down to No. 25 at +7, everyone else at 0. That starting point is worth 3 games in week 1, 2.4 in week 2, and falls by 0.6 games a week until it reaches zero in week 6. From then on the ranking uses this season's results only.

Strength of record (30%)

Strength of record asks how often a team as good as this week's No. 25 would finish with a better record than yours against exactly your schedule, home and away included, with matching it counting half. The chance comes from the power ratings, the home edge and a fixed 16-point spread of results around them (close to what this season's finals show), and it is calculated exactly rather than simulated. A lower figure is a more impressive record. It counts wins and losses only: margin does not enter it.

The blend

Power and strength of record are each turned into standard scores among rated Division II teams, mixed 70/30 and restandardised. The D2CFB Rating puts that on a 0–100 scale where 50 is an average Division II team and each 10 points is one standard deviation.

Schedule strength and opponents outside Division II

SOS is the average power rating of the opponents a team has played; remaining SOS averages the rated opponents still to come. Neither is part of the blend.

Games against FCS, NAIA, Division III and other opponents count. Those opponents are rated alongside Division II teams, starting from their division (FCS +7, NAIA −10, Division III −14, others −14). Because only their games against Division II teams are stored here, that starting point stays worth 2 games all season. They are never ranked.

NPI est.

From 2026 the NCAA selects Division II football's at-large playoff teams with the NCAA Power Index (NPI), counting only games between Division II teams. The NCAA publishes football's settings: winning percentage and strength of schedule weighted 25/75, road wins and home losses weighted 1.05 against 0.95, a quality-win bonus of 0.5 times the amount an opponent's NPI exceeds 52.0, overtime counted like regulation, and at least 5.5 wins retained. It does not publish how those pieces combine. Our NPI est. makes these assumptions:

1. Each result is valued at 25% of its outcome (100 for a win, 0 for a loss) plus 75% of the opponent's NPI, plus the quality-win bonus for a win. A tie is half a win and half a loss.
2. Road wins and home losses weigh 1.05; home wins and road losses weigh 0.95. A team's NPI is the weighted average of its results.
3. The weakest wins are dropped while they lower that average and at least 5.5 wins remain. Losses are never dropped.
4. Every team starts at 50 and the averages are recalculated until they settle.

NPI est. has not been checked against the NCAA's official figures, which alone govern selection. It is shown beside the ranking and plays no part in the D2CFB Rating.

When it updates

A week's ranking is published the Monday after its Sunday. Until the next week's first game it is recomputed each morning so late results land in it; then it is frozen, and past weeks are never quietly rewritten. A frozen week is recomputed only to correct a data error, and a change to any setting below gets a new method version.

Limitations

• No neutral-site flag is published, so every game is treated as played at the listed home team's field.
• Early-season ratings rest on few games and on the preseason poll; they settle as the season goes on.
• A game this site has not stored as final does not count. Records and ratings follow the finals shown on the scoreboard.
• NPI est. is a reconstruction, not the NCAA's figure.

Settings for Week 9