2026 Division III football · projection after Week 9

Can my team get an at-large bid?

If your team doesn't win its conference, its playoff hopes come down to one number and one question: does it finish in the top 11 of the NPI ranking among teams that didn't win their league? Here is what that means, in plain terms.

The short answer

For a team that does not win its conference, the final record tells most of the story.

1 loss

Projected in

For most conferences. Twelve conferences can no longer produce one, six more only through a tiebreaker, and a few more are on the edge.

2 losses

Mostly out

We project about three get in this way in 2026, almost all from the strongest conferences.

3 losses

Out

It didn’t happen once in 400 simulated seasons.

Two doors into the playoffs

The bracket has 40 spots. 29 go to conference champions automatically: win your league and you're in. The other 11, called at-large bids, go to the teams that didn't win their conference but rank highest in a formula called NPI. Nobody votes and no committee picks; the formula decides.

Each square is one of the 40 playoff spots.

What the ranking rewards

Think of NPI as a grade for each game, averaged over the season. Every game's grade has two parts:

Did you win? A win earns a big chunk of credit; a loss earns none.
How good was the other team? Playing a strong team earns credit even in a loss, and beating one earns extra.

The formula also protects teams: a loss can never raise your grade, and beating a weak opponent can't drag it down once you have enough wins.

Why your conference matters as much as your record

Most of your games are against your own conference, so your grade leans heavily on how good the formula thinks your conference is. It can judge that only one way: by how your conference did against other conferences, mostly in September. Those games are nearly all played, so each conference's reputation for 2026 is largely set.

That makes it a shared fate. Every September win by a conference-mate lifted your team's grade a little, and every loss pulled it down.

Games against teams from other conferences, through Week 9, not counting games against teams awaiting NCAA classification (Azusa Pacific, St. Francis). The NESCAC plays no outside games, so the formula treats it as exactly average.

Check your team in three steps

Can your team still win its conference? If yes, that is the easy road in, and nothing below matters. If it might fall short, keep going.
How many losses will it likely finish with? Count the losses it has, then add the conference games you expect it to lose. NESCAC teams play nine games, so for them "one loss" means 8-1.
Find your conference in the table below and read across to that number of losses. The last column says whether your conference can even produce a one-loss team that isn't the champion.

Your conference at a glance

Each verdict describes the conference's best-placed contender at that record. A team with a weaker schedule in the same conference may do a little worse.

Find your conference

Conference	1 loss	2 losses	3 losses	Can it have a 1-loss non-champion?

Why twelve conferences show "Can't happen" for one loss

A team can only finish with one loss and not win its conference if the champion ends up with a conference record at least as good. In twelve conferences, the games already played make that impossible, however the rest of the season goes:

Every contender already has two losses: the ARC, Middle Atlantic (MAC), UMAC and USA South.
The only one-loss team lost outside the conference, so any conference loss would be its second: Susquehanna (Landmark), Ithaca (Liberty League), Linfield (Northwest) and Hampden-Sydney (Old Dominion).
CNE: Endicott is unbeaten, and every other team already has at least four losses, so Endicott wins the title even if it loses a game.
MASCAC: Bridgewater State is the only one-loss team, and its last game is against Massachusetts Maritime: win and it takes the title, lose and it has two losses.
SCAC and SCIAC: both decide their titles with a championship game between the top two teams, and every eligible contender left already has a loss, so reaching that game and losing it would be a second loss. (Whittier, restarting its program, is not eligible for the SCIAC title or an at-large bid this year.)

In six more conferences (Centennial, Heartland, Minnesota, New Jersey, Ohio and SAA), a one-loss non-champion can happen only if two teams tie for the title and the tiebreaker sends the automatic bid to the other one.

How the 11 at-large spots are projected to split

In our projections for 2026, one-loss teams fill most of the 11 spots and two-loss teams fight over what's left. These numbers come from playing out the rest of this season, not from past seasons.

Middle projection from 400 simulated finishes.

Finished with	Non-champions expected	At-large spots they get	Range in 8 of 10 simulations
1 loss	About 9	About 8	7 to 11 teams; 7 to 10 get in
2 losses	About 20	About 3	17 to 24 teams; 1 to 4 get in
3 losses	About 28	None	Not once in 400 simulated seasons

So a two-loss team is competing with about 20 others for roughly three spots, and conference reputation usually decides which three. Even one loss isn't a guarantee: in about 2 of every 3 simulations, at least one one-loss team from a lower-rated conference was passed over for a two-loss team from a stronger one. Conference tiebreakers add to that: head-to-head can hand the automatic bid to the team with the worse overall record, which sends a one-loss team into the at-large pool.

How a two-loss team can jump a one-loss team

A second loss hurts, but playing in a well-rated conference can be worth more than one win. Every week a WIAC team's opponents carry a better grade than an SCAC team's, and that adds up by November.

Each dot is a contender at that record, projected to season's end: UW-River Falls winning out, and the best-placed team from each other conference that does not win its automatic bid. The dashed line is the projected cutoff for the 11th and last at-large spot. Green: usually in. Yellow: a coin flip. Orange: usually out.

A two-loss WIAC team such as UW-River Falls is projected ahead of the best one-loss runner-up from the MWC, NACC, Heartland and Presidents’ Athletic, and about even with one from the NESCAC, OAC or SAA.
A two-loss Empire 8 team is about even with the best one-loss runner-up from the Heartland and Presidents’ Athletic, and ahead of one from the NACC and MWC. Two-loss teams from the next tier (ASC, CCIW, North Coast, Michigan, Minnesota, New Jersey) project at about 64 to 64.6, just under the cutoff.
It doesn't work the other way: a two-loss team from a low-rated conference does not pass a one-loss team from a well-rated one.

The Mount Union question is now one game

Can Mount Union still make the field? Yes, but only by beating Baldwin Wallace. Win that game and the one after it, and Mount Union takes the OAC's automatic bid, with Baldwin Wallace in as an at-large team. Lose it, and Mount Union is out.

How it got here

Heidelberg's 24-13 home loss to Marietta in Week 9 ended the three-way race. Baldwin Wallace (8-0, 6-0 OAC) beat Wilmington 66-13, and Mount Union (6-2, 5-1 OAC) won 58-14 at Ohio Northern. They meet at Mount Union on Nov. 7. Then Baldwin Wallace hosts Muskingum, and Mount Union plays at Otterbein.

If Mount Union wins both, the two finish tied at 7-1 in the OAC, and Mount Union holds the head-to-head win. If Baldwin Wallace wins at Mount Union, it clinches the title outright.

If both win their last game and…	OAC automatic bid	Baldwin Wallace	Mount Union
Mount Union beats Baldwin Wallace	Mount Union, 8-2	In every simulation At-large at 9-1: about 67.7, #23, 2.3 above the cutoff. Projected seed around 23 (20 to 26).	In Automatic bid at about 65.1, #30. Projected seed around 30 (27 to 32), so a first-round game.
Baldwin Wallace wins at Mount Union	Baldwin Wallace, 10-0	In About 75.4, #6. Projected seed around 6 (4 to 8).	Out in every simulation At 7-3: about 60.7, #53, about 4 points below the cutoff.

Based on NPI grades and home field alone, our projection model gives Mount Union about a 1 in 4 chance of winning the game. It knows nothing else about either team.

What about Heidelberg?

Heidelberg (6-2, 4-2 OAC) is a long shot either way. It made the field in about 3% of simulations if Mount Union wins and 11% if Baldwin Wallace wins. The difference: when Mount Union takes the automatic bid, Baldwin Wallace uses one of the 11 at-large spots, and the cutoff rises by about a third of a point.

Is #1 UW-Platteville's to lose?

Yes. A 10-0 UW-Platteville finished #1 in all 300 simulated seasons. It won 35-7 at UW-Stevens Point in Week 9.

UW-Platteville (8-0) leads the NPI at 80.5, ahead of Wheaton (77.3), Cortland (75.4), Endicott (75.0) and Christopher Newport (74.9). Beating UW-Whitewater and winning at UW-La Crosse would lift it to about 82.3.

UW-Platteville's lead when both it and the named team win every remaining game: the dot is the middle outcome, the bar spans the middle 80% of 80 simulated finishes of everyone else's games. Left of zero, the challenger finishes #1.

What changed from last week: Cortland is still the closest challenger and finished second in 209 of the 300 simulated seasons, Wheaton in 48. In a separate set of 80 seasons for each challenger where both finished unbeaten, none passed UW-Platteville; Cortland came closest, at about 0.5 behind. Christopher Newport, now #5 after beating Rowan 13-12, replaces Hope in this chart.

Can the defending champs get in at 8-2?

Very likely. If UW-River Falls wins out to finish 8-2 and doesn't win the WIAC, it made the field as an at-large team in all 300 simulated seasons.

UW-River Falls is 6-2 after losing 24-16 at UW-Whitewater and 42-0 at UW-Platteville, then beating UW-Eau Claire 59-22 and UW-Stout 35-30. Getting to 8-2 means winning at UW-La Crosse and beating UW-Stevens Point. With both losses in conference, it can't finish ahead of whichever of UW-Platteville or UW-Whitewater wins the WIAC, so its road runs through an at-large bid.

How far an 8-2 UW-River Falls finishes above or below the cutoff for the 11th and last at-large spot, across 300 simulated finishes of everyone else's games. Each bar counts the simulations in a half-point range.

Why two losses don't sink it

It lost to the #1 and #7 teams. A loss earns no win credit, but it still earns 60% of the opponent's grade. Losing to UW-Platteville (80.5) or UW-Whitewater (74.4) grades out far better than losing to an average team. The 42-0 score doesn't matter; NPI ignores margin.
It already has two quality wins. Beating teams rated above 54 adds a bonus, and UW-Eau Claire (62.5) and UW-Stout (59.4) both qualify. Its last two opponents don't. The WIAC went 17-4 against other conferences, so its teams carry the best ratings of any league.
The WIAC usually gets two at-large spots. In 90% of these simulations the WIAC sent exactly two non-champions to the field, and three in another 9%. An 8-2 River Falls is typically one of them.

Where it lands: about 67.6, around #23 nationally (#20 to #26 in 8 of 10 simulations), and about 7th among the teams competing for at-large spots, roughly 2.6 points above the cutoff. That's up from about 2.1 last week. It cleared the cutoff every time, though in 8 of the 300 seasons by less than half a point, when upsets elsewhere pushed the cutoff up. It also helps when its nonconference opponents keep winning: when Wartburg finished with two losses, River Falls cleared the cutoff by about 2.8 points, against about 2.1 when Wartburg lost four.

What a third loss would do

It would end the at-large case. A 7-3 River Falls drops four to five points, to around 63, and from roughly #23 to about #40 nationally. It missed the field in all 160 simulations.

River Falls finishes	Projected grade	National rank	Versus the cutoff	Made the field
8-2 (wins out)	67.6	#23	+2.6	300 of 300
7-3, loses at UW-La Crosse	63.4	#38	−1.2	0 of 80
7-3, loses to UW-Stevens Point	62.8	#41	−1.8	0 of 80

Which game it loses barely matters. Losing at La Crosse costs a little less: a road loss counts 10% less than a home loss, and La Crosse (49.3) is rated a bit higher than Stevens Point (45.1), so a loss to it still earns a little more credit. Either way, the third loss turns a near-lock into a near-miss.

Merchant Marine at 8-1 or 7-2

One loss is fine. Two losses almost certainly isn't. Merchant Marine (7-1) won 40-26 at WPI in Week 9. It plays a nine-game schedule and has one game left: Coast Guard in Washington, D.C. (Week 11).

Merchant Marine finishes	Projected grade	National rank	Versus the cutoff	Made the field
8-1 (beats Coast Guard)	69.2	#19	+4.2	Every simulation
7-2 (loses to Coast Guard)	63.5	#37	−1.0	About 1 in 25

At 8-1, Merchant Marine is safely in, about four points above the cutoff, ranked about #15 to #22. It was inside the top 24 in all but one of 150 simulations, the range that usually gets a first-round bye. It wins the NEWMAC only if Springfield (5-0 in the conference, with the head-to-head win) loses both of its last games, to Norwich and at SUNY Maritime; that didn't happen in any of 150 simulations.
At 7-2, it falls almost six points, into the high 30s nationally, and gets in only when the cutoff happens to drop.
Why a second loss hurts it more than River Falls: with nine games instead of ten, each loss is a bigger share of the season, and NEWMAC opponents are rated below WIAC opponents, so Merchant Marine's wins earn less.

Can the NESCAC get two teams in? Three?

Two is now very likely. Three is possible but needs one specific finish. Across 400 simulated seasons, the NESCAC put two teams in the field 85% of the time, one team 8% of the time and three teams 7.5% of the time.

Amherst (24-21 at Tufts) and Trinity (Conn.) (26-23 at Middlebury) both survived close games in Week 9 and are 7-0. Wesleyan (6-1) won 28-0 at Bowdoin. Two games among them are left: Trinity at Amherst (Week 10) and Wesleyan at Trinity (Week 11). The only other games left for the three are Amherst at Williams and Wesleyan against Williams. If Amherst and Wesleyan both beat Williams, the two games among the three decide everything.

If the top three beat everyone else and…	Finish	NESCAC automatic bid	At-large result	Chance of this finish
Amherst beats Trinity, Trinity beats Wesleyan	Amherst 9-0, Trinity 8-1, Wesleyan 7-2	Amherst	Two in Trinity in every simulation (about 67.2, 2.3 above the cutoff); Wesleyan out	About 28%
Trinity beats Amherst, Trinity beats Wesleyan	Trinity 9-0, Amherst 8-1, Wesleyan 7-2	Trinity	Two in Amherst in 79 of 80 (about 66.7, 1.8 above); Wesleyan out	About 19%
Amherst beats Trinity, Wesleyan beats Trinity	Amherst 9-0, Wesleyan 8-1, Trinity 7-2	Amherst	Two in Wesleyan in every simulation (about 67.4, 2.3 above); Trinity out	About 10%
Trinity beats Amherst, Wesleyan beats Trinity	All three 8-1, each with a loss to another	Decided by draw	Three in in 75 of 80; two in the rest	About 6%

The other 37% of finishes are ones where Williams beats Amherst, Wesleyan or both. Those usually leave the NESCAC with one or two bids.

Why a NESCAC 8-1 can clear the line

A closed league rates itself. The NESCAC plays no outside games, so the formula pins its average near 50, and the league's wins pile up at the top. An 8-1 team beat seven or eight teams below it and lost only to another contender.
The loss is to a strong team. A loss to a fellow contender rated in the high 60s or 70s still earns 60% of that team's grade, which is about as good as a loss gets.
It lands about 1.5 to 2.5 points above the cutoff. That was enough in all but one of 240 simulations of these three finishes, but it is not a wide margin.
Two losses ends it. A 7-2 Trinity or Wesleyan projects around 61, about four points short. With nine games and no outside schedule to lean on, a second loss costs more than it would a WIAC team.

Why three is so hard

Three bids needs the circle: Trinity beats Amherst and Wesleyan beats Trinity, with Amherst and Wesleyan also beating Williams. That's about a 6% chance. Even then, the NESCAC's tiebreaker can't separate three teams that are 1-1 against each other and beat everyone else, so the automatic bid goes by draw. The other two need at-large spots at about 66.5 to 67.5. They both made it in 75 of 80 simulations, but having two NESCAC teams in the pool also raises the cutoff to about 65.3, so the margin is only about one to two points.

Addendum

North Central after the Wheaton loss, and where a 9‑1 North Central lands

North Central's slide was temporary, but the loss to Wheaton is not. The Week 6 loss dropped North Central to #35. After beating Elmhurst 74-2, winning 62-7 at Carroll and beating Carthage 55-3, it is back up to #18 (68.5). If it wins out to finish 9-1, it projects at about 70.1, around #15 nationally, and made the field in all 150 simulations. Had it beaten Wheaton, the same season would have it near #2.

After Weeks 6 to 9: actual NPI. Projected: middle result of 150 simulated finishes in which North Central wins its last two games. The dashed line is the projected cutoff for the last at-large spot.

Why one loss hurt so much in October

After Week 6, the grade was an average of only five games. North Central's wins over Bethel, Wartburg and Augustana each grade out around 71 to 73. The 13-10 home loss to Wheaton grades out at 43, because a loss earns no win credit at all; only the 60% credit for Wheaton's strength remains. One game out of five is a big slice of any average, which is how North Central fell to #35.
A home loss counts a little extra. The formula weights a home loss 10% more than a neutral-site game, and a home win 10% less.
Weak wins still count, for now. Beating a weak team like North Park or Elmhurst grades out around 55 to 60. The formula lets a team set aside wins like that, but only once it has more than five wins' worth to count. At 5-1, North Central can't drop them yet.
A strong schedule can't fully cover a loss. Opponent strength is worth 60% of each opponent's grade. The win itself is worth a flat 40 points. Losing to the #4 team still costs most of the value of beating it.

How it resolves

Once North Central has more than five wins, the formula keeps its best five wins and sets aside the weak ones: North Park, Carthage, Elmhurst and nearly all of Millikin in a typical finish. Wins over Carroll and Illinois Wesleyan replace them. That lifts North Central from 62.2 (#35) after Week 6 to about 70.1 (around #15) at 9-1; the climb to #18 is most of the way there.

One typical 9-1 finish: each game's grade, with the games the formula counts in solid color and the ones it sets aside faded. The grade is the weighted average of the counted games.

The loss never shrinks

It is tempting to think one loss matters less over ten games than over five. In NPI it doesn't work that way. Because weak wins get set aside, a 9-1 North Central's grade rests on about six counted games: its five best wins and the loss, which can't be dropped. The loss is worth about 8 points now and about 8 points in November. That is the gap between #2 and #15.

What it means for the playoffs: a 9-1 North Central made the field in all 150 simulations, as an at-large team about five points above the cutoff. It never won the CCIW's automatic bid in these simulations; that would take Wheaton losing both of its last two games.

The fine print

These are projections, not promises, made after Week 9 of the 2026 season.

How: we recreated the NPI formula and checked it against the NCAA's official Week 9 NPI: every team's grade and adjusted win-loss record match to three decimals. These projections also include Sunday's Delaware Valley–King's result. We applied it to every game played so far, then played out the rest of the season hundreds of times, with better-rated teams usually winning but upsets built in.
Coin flip means the team lands within about a point of the cutoff, close enough that results elsewhere decide it.
What changes things: the cutoff moves a point or two with upsets around the country, and a conference's reputation can still shift slightly with the 5 outside games left on the schedule. Expect the verdicts to firm up each week through mid-November.
Three losses: that column stops at "Long shot." No three-loss team got in across 400 simulated seasons; the WIAC's very best case lands about a point and a half below the cutoff.
Conference champions: each conference's automatic bid follows its own tiebreaker where we found it: head-to-head first almost everywhere, then steps such as record against the rest of the standings, NPI or a draw. Head-to-head counts only when every tied team played every other one, and the Presidents' Athletic ranks teams by the loss column. The WIAC's “Rose Bowl” rule sends the bid to the team whose last conference title is longest ago. Steps that need point margins, first-half scoring or yardage are skipped, and the rule moves to its next step. Where a rule isn't yet confirmed, we use head-to-head, then NPI. The SCAC and SCIAC title games are played out.
Runner-ups only: each conference’s verdict uses its best team that does not win the automatic bid. An earlier version mistakenly counted a few seasons where that team won its conference, which overstated the MWC, USA South, Northwest, Presidents’ Athletic and Middle Atlantic.
Data corrections: Pacific Lutheran's Week 7 win at Pacific (30-14) is corrected, and Centre at Hanover (Week 1) is now a Hanover home game, per the NCAA.
Excluded games: games against Azusa Pacific and St. Francis, which are not playoff eligible, are left out, the same way games against non-Division III opponents are.