NFL 2026 · a ratings desk, not a fan page

Who’s better, and by how much.

We rate every NFL team against a league-average club. Then we turn that into a predicted gap for each game and compare it to the sportsbook. If we disagree by about a field goal, we look at a bet. If not, we skip. No favorite teams.

Wednesday 19 Aug 2026
Preseason
Week 1 not open
TNF decision window

How to read this

Teams — plus is better than average. Minus is worse. Tap a club to see why.
Games — tap a matchup. Open is the first street we logged. Current is today’s sportsbook. B$ Line is the gap we expect.
Copper means we disagree by about a field goal, or we landed on opposite sides of 3 or 7. That is a look, not an automatic bet.
Skip is normal. Most games we do not bet.

Book we grade the last number against

This week’s $1,000

One unit is $50. A full play is 3 units. 1u parlays are for fun. Leftover stays unspent.

Week 1 · desk is open

Week 1 · desk is open. NE@SEA + Melbourne finals locked (Donald Out).

Process fires only with ~1.5 / key 3/7. Pass is fine. Street is live; skip is normal.

How the week runs

Mon First look
Write down the sportsbook numbers. Don’t bet yet unless the gap is huge.
Wed Thursday game
Wait for the official injury report (4pm ET). Bet or skip.
Fri Sunday games
Same idea after Friday’s injury report. Skip is fine.
Sun Underdogs
Only if the public actually moved the number. Otherwise leave it.

Words a friend might not know

Line: The point handicap. Chiefs −3 means Kansas City must win by more than 3.
Open: The first sportsbook number we logged for the game, with the time we recorded it.
Current: What the sportsbook is offering right now, with the time that number was last pulled.
B$ Line: The gap we expect, from the home team’s view.
Street: Same as Current — the live sportsbook number.
Edge: How far those two numbers are apart.
Skip: We looked and did not bet. That counts as a decision.
Unit: $50. The week is 20 units ($1,000). 1u look, 2u standard, 3u full, 4u max.
Straight: One game, sized in units. This is what we grade.
Parlay: Several games at once, 1 unit. For fun, not the score.
3 and 7: NFL games land on field goals and touchdowns a lot. A half-point around those numbers matters.

0.0 is a league-average club · plus is better

Teams

Every club is built from this season’s results, Madden’s equal 22, PFF, and injuries. Last season was only the Week 1 starting point and is out of the number from Week 5. Tap a team to see the pile-up. Plus means better than average. The number on the card is not “they will win by this many.”

Sort Import profiles

Coordinators · names and history · not a line

Staff

OC, DC, and special teams for all 32 clubs. Copper stars are Coach, OC, DC, or ST of the Year. More stars means more of those awards. Scheme only when a source named it. This does not move the number.

Our guess vs the sportsbook · copper is a look, not a ticket

Games

Tap a game. Open is the first street we logged. Current is today’s sportsbook. B$ Line is the gap we expect. Copper means we disagree enough to look.

Home-field points

Shadow board · chips are not on the B$ line

ATS

One game per card. The B$ line is the only bet number. B$ DVOA and the stat chips explain the matchup. They do not move the line. A dot is not a betting signal.

Our B$ Line · Mon → Sun · empty days stay blank

Line log

One row per game. Each day is the desk B$ Line we actually froze — not a live recompute. A dash means we have no snapshot. Copper means the number moved from the last filled day. Tap the delta for what happened.

Week

After the games · we type the last number

After the games

Did our number beat the sportsbook’s last number? Wins and losses bounce around. The last number is the grade.

week	game	our spread	mkt spread	spread edge	our total	mkt total	total edge	close spread	close total	result	res vs close	res vs result

Club rollup · we vs the street

Mean of (edge / 2) credited to home and the opposite to away. Close replaces market when you type it.

club	n	mean credit

Why a half-point matters

3 and 7

NFL games land on field goals (3) and touchdowns (7) a lot. If we sit on one side of that number and the sportsbook sits on the other, the half-point is the whole bet.

When we look at each game

The week plan

A calendar for when a number is worth deciding. Old stories like “Thursday is always under” are labeled folklore. We do not run those.

VERIFIED NFL clock or a desk rule we can audit
MIXED industry timing + our filter
FOLKLORE story, not a trigger
DESK RULE how this book is run

What we run, what we ignore

Rules

COPY = we actually do this. ENTERTAINMENT = the $50 parlay, not the grade. ADAPT = keep the idea, change the mechanic. IGNORE = a story we do not bet on.

Week 2 on · the line we published

Bet History

What the B$ line was before the game, the final score, and the against-the-spread result. The same games as Overall Record.

Same games · behind sign-in

Overall Record

The record, the closing line value, and the outright upsets. Opening and closing lines are ESPN DraftKings. Week 1 is not in this record.

My tickets

Personal tickets. Stakes, open bets, and profit stay here. They are not the official 32-game record.

week	window	type	purpose	game	market	pick	line	juice	stake	book	close	CLV	result	profit	notes	grade

Admin · accounts

Users

Create a login for a friend. They can see the desk. They cannot change lines, tickets, injuries, weather, or anything else you type.

login	name	role	last sign-in	password

How the number is built · not the recipe

Methodology

The B Line is our number for a game: the gap we expect between the two teams, from the home side. It is an opinion we freeze before the kick. It is not a ticket, a price at a book, or a promise.

The record

Bet History and Overall Record show the same finished games from Week 2 on. Week 1 is not in that record. The Weeks 2 and 3 B$ line is the game-day number the site showed at kickoff, kept as it was recorded. From Week 4 on, the official line is the locked projection rounded to the nearest half point before it is published. Exact .25 and .75 round away from zero. The Games page and the B$ Daily show that rounded line, and the raw number stays beside it. A later week is added after the score and the DraftKings close are on file. We do not rewrite a locked line, and we do not recompute a finished row from today’s model.

What the desk looks at

A lot of things are judged together. None of them is the model by itself.

How the teams played in prior seasons.
How they have played recently.
Offense and defense, not just the final score.
Players, and whether the people who matter are available.
Injuries and the rest of the personnel picture.
The matchup itself.
Situational spots, including travel, rest, and short weeks.
The market: where the number opened, where it went, and that a price exists.
Historical trends that have shown up often enough to respect.
Quantitative indicators.
Qualitative reads that a box score will not hold.

Last season, then this one

Last season is a starting point, not a life sentence. As the new season piles up real games, that older body of work carries less of the conversation, and once enough of this season is in the books it is out of the number. The taper is part of the judgment. The schedule of it is not published here.

One number

Those inputs are evaluated together. The output is one number, the B Line. We compare it with the market. A disagreement is a look, not an automatic bet.

What this page will not do

This is the public account of the work. It is not the combination. There are no weights, formulas, or scaling rules on this page, and no list of which source is multiplied by what. The next unpublished game stays on the desk.

Admin only · not the production model

Testing

Rerun frozen inputs from games that are already final. The production B Line in the table is the pinned lock. The official record grades the game-day line. An experimental number lives only in this test.

Forgot password

Who’s better.