NFL EDGES

Live game edges — model vs. Kalshi, ranked

Gridiron Edge · 5-min refresh

COMBO EDGE

Correlation-aware Kalshi combos — free

SILVER EDGE

Options vs. Kalshi on weekly silver brackets

NFL 2026

Power rankings, futures & win totals — all 32 teams

NFL Week 3 Picks: The Best Play From Every Game, and Where to Get In

One call for all 15 remaining NFL Week 3 games, taken off each PMP game page: six props, three moneyline Leans and six passes, each with a rest level, a scale-in ladder and a walk-away price.

Justin Herbert beside The 7 Oracles headline Fifteen games, one play each, the NFL Week 3 card with nine positions and six passes
Justin Herbert beside The 7 Oracles headline Fifteen games, one play each, the NFL Week 3 card with nine positions and six passes
The 7 Oracles
The PredictionMarketsPicks Oracles
September 26, 2026

Fifteen games are left on the Week 3 slate. Here's one call for each, taken straight off its own game page, and for every position, how to get in: where to rest the order, how to scale in, the price where you walk away, and whether to go tonight or wait for inactives. Nine of the calls are positions. Six are passes, and on each pass we name what we're watching and why it isn't a play. Every game links to its full breakdown, so you can check the work.

The card

Prices are what Kalshi showed at about 7:20 PM ET Saturday. Now is what it costs to take the position this second: the ask for a YES, 100 minus the YES bid for a NO. Rest at is where we'd post our first order. Walk away is the most we'd pay on a resting order. Above it, the edge is too thin to be worth the risk.

Kick (ET)GameThe playNowRest atWalk awayStatus
1:00LAC @ BUFHerbert 2+ pass TDs NO60¢59¢65¢PLAY · HIGH
1:00CAR @ CLEPanthers to win56¢55¢56¢ · maker onlyLEAN
1:00NYJ @ DETLaPorta 5+ rec NO53¢52¢55¢PLAY · MOD
1:00HOU @ IND————PASS
1:00NE @ JAXLawrence 2+ pass TDs NO51¢50¢56¢PLAY · HIGH
1:00KC @ MIAE. Johnson 2+ rec YES47¢46¢51¢PLAY · HIGH
1:00TEN @ NYG————PASS
1:00CIN @ PIT————PASS
1:00SEA @ WAS————PASS · fork
4:05ARI @ SFCardinals to win23¢21¢21¢ · maker onlyLEAN · small
4:05MIN @ TBVikings to win53¢52¢52¢ · maker onlyLEAN
4:25BAL @ DALAndrews 5+ rec NO60¢59¢62¢PLAY · HIGH
4:25LV @ NO————PASS
8:20LAR @ DENNix 2+ pass TDs NO59¢58¢61¢PLAY · MOD
Mon 8:15PHI @ CHI————PASS · fork

How the system picks. A prop at PLAY status on our board comes first. PLAY means it cleared every publish gate: direction agreement, the confidence band, no unmodelled absence in the position room, and a real two-sided book. If a game has no PLAY, the next choice is a moneyline the game page's Kickoff Board marks as a Lean. If it has neither, the game is a pass. We don't fill a slot to make the card look full. A pass is a call too.

Five of the six props are NOs. We didn't go looking for that. It's what cleared the gates. It's still worth knowing, because when a card leans that hard one way, the misses tend to come together. Size with that in mind.

How to get in: make the market, don't take it

Every level on this card follows the same four rules.

  1. Be the bid. Post a resting order on your side and let the market come to you. To buy a NO at 58¢, you post a NO bid at 58, which is the same order as offering the YES at 42. On the thin prop books below, your order will often be the best bid. That's what making the market means here. A resting (maker) fill pays no fee on most standard Kalshi markets. Crossing the spread (taking) costs the taker fee, 7% × price × (1 − price), which is about 1.7¢ a contract near a coin flip.
  2. Scale in, don't lump in. Split your planned size into tranches. The first goes at the Rest at level. The rest go 2¢ to 4¢ better, and only fill if the market drifts your way. If only the first tranche fills, you hold a smaller position at a better price. That's fine.
  3. The walk-away is a hard stop. It's the most we'd pay on a resting order, meaning our number minus our minimum edge (5 points on a prop, 2.5 on a moneyline Lean). If you have to cross the spread to get filled, knock another 2¢ off it for the fee. Above the walk-away, you're paying for our opinion instead of being paid for it.
  4. News before size. Inactives post about 90 minutes before each kick. Any play that depends on a teammate's status (Andrews, Johnson, LaPorta) gets a starter tonight and the rest after inactives. The quarterback touchdown unders don't carry that risk, so they can be built tonight.

The three moneyline Leans are maker-only. Each is a 3-to-4-point gap between Kalshi and the rest of the market. Take the ask, pay the fee, and you're left with about a point, which is inside our own noise. Rest the bid or skip it. And don't wait on them: the edge is Kalshi lagging the other venues. When Kalshi catches up, the Lean is gone.

The 1 PM window

Chargers at Bills: Justin Herbert under 2 touchdown passes

NO on 2+ pass TDs · 60¢ now (book 40/42 on the YES, about 2,400 contracts open) · PLAY, HIGH tier.

Herbert has two touchdowns against three interceptions through two games. His offensive line had eight players miss Wednesday's practice, and he's walking into Buffalo. The risk is the other side of that: the Bills have the best offense in football, and a Chargers team chasing the game throws more. The board clears the under anyway, and it's the HIGH-tier row on the page. The Kickoff Board also shows a small Lean on the Chargers moneyline at 24.5¢ (0.8% quarter-Kelly). That's a price read on a big underdog. The prop is the play. Full breakdown: Chargers vs. Bills Week 3 prediction · Justin Herbert's prop ladder.

Panthers at Browns: Panthers to win

YES on Carolina · 56¢ now (book 55/56, 193,000 contracts open) · LEAN, 1.8% quarter-Kelly.

To be clear up front, this is a price play. Our ratings model has the Browns by 1.5. The Lean exists because Kalshi had Carolina at 55.5¢ against a 59% fair value built from the rest of the market's line. You're buying the same team cheaper on Kalshi than it trades anywhere else. You're not buying our opinion of the game. If that distinction bothers you, skip it. It's the reason this is a Lean and not a PLAY. Full breakdown: Panthers vs. Browns Week 3 prediction.

Jets at Lions: Sam LaPorta under 5 receptions

NO on 5+ receptions · 53¢ now (book 47/48 on the YES, about 860 open) · PLAY, MODERATE.

Kalshi has five LaPorta catches as a coin flip, and our projection sits just under five. That makes this a small gap on a fair-looking price, which is why it's MODERATE and not HIGH. The page also shows LaPorta 40+ receiving yards on the same side. We're taking the catches, because receiving yards is a stat family we don't promote to a headline play right now. Full breakdown: Jets vs. Lions Week 3 prediction · Sam LaPorta.

Texans at Colts: Pass

The interesting number in this game is a trap. Dalton Schultz at 6+ catches is the only Kalshi-vs-books gap of 5¢ or more on the entire Week 3 prop board: Kalshi YES at 42¢ against a five-book consensus near 47%. But our own read on Schultz points the other way, he hadn't practised midweek, and the Kickoff Board flags his row as a conflict. When the price and our number disagree and the player's status is open, we don't have a play. Houston is also without Nico Collins. Full breakdown: Texans vs. Colts Week 3 prediction.

Patriots at Jaguars: Trevor Lawrence under 2 touchdown passes

NO on 2+ pass TDs · 51¢ now (book 49/50 on the YES, about 230 open) · PLAY, HIGH tier.

The Jaguars' run game hasn't broken 4.7 yards a carry, and New England just held Pittsburgh to three points. This one is decided in the middle of the Jacksonville line, and Lawrence's touchdown count is the most direct way to trade that. Ignore the Parker Washington under you'll see on the page's board. It's flagged as a conflict, not a play. Full breakdown: Patriots vs. Jaguars Week 3 prediction · Trevor Lawrence.

Chiefs at Dolphins: Emmett Johnson 2+ receptions

YES on 2+ receptions · 47¢ now (book 46/47, about 1,150 open) · PLAY, HIGH tier.

The game page withholds its moneyline call on this one while the model is under review, and nothing here overrides that. This is a prop, priced on its own. Our projection clears two catches, and Kalshi has it under a coin flip. It's the only YES among the six props, and it's the only one where you need something to happen rather than not happen. Full breakdown: Chiefs vs. Dolphins Week 3 prediction · Emmett Johnson.

Titans at Giants: Pass

This is a QB-adjusted game. Jameis Winston has the Giants job with Jaxson Dart out for the rest of the regular season, the market moved for it days ago, and the board publishes no position off a corrected rating. The closest look is Cam Skattebo at 2+ catches, which sits on WATCH because the gap is bigger than we trust. Full breakdown: Titans vs. Giants Week 3 prediction.

Bengals at Steelers: Pass

This is a great game at a bad price. The Bengals moneyline is priced dead on, 62 against 62. The board's only prop row, Pat Freiermuth under 3.5 catches, is flagged as a conflict. The league's best defense by EPA against a Steelers offense that hasn't reached 300 yards in either game is worth watching. It isn't worth a position. Full breakdown: Bengals vs. Steelers Week 3 prediction.

Seahawks at Commanders: Pass (Oracle Fork)

Seattle hasn't named its quarterback. Sam Darnold was a full participant Thursday, and Drew Lock is the other branch. The page publishes both lines, and a fork doesn't carry a position until it collapses to one. Washington's half is settled: Jayden Daniels is out and Marcus Mariota starts. The Stefon Diggs 4+ catches row is on WATCH. Full breakdown: Seahawks vs. Commanders Week 3 prediction.

The 4 PM window

Cardinals at 49ers: Cardinals to win (small)

YES on Arizona · 23¢ now (book 22/23, 97,000 open) · LEAN, 0.8% quarter-Kelly.

This is the one Lean where our ratings model and the price agree on direction. Our line has San Francisco by 7.4 against a market line of 8, and Kalshi had the Cardinals at 21.5¢ against a fair value near 25%. It's small for a reason: Arizona's secondary is coming apart, with one corner on IR, another who didn't practise, and a third still working back from an Achilles. It's a small position on a big underdog at the slowest-paced team in football. Size it like one. Full breakdown: Cardinals vs. 49ers Week 3 prediction.

Vikings at Buccaneers: Vikings to win

YES on Minnesota · 53¢ now (book 52/53, 266,000 open) · LEAN, 1.5% quarter-Kelly.

Same caveat as Carolina: our ratings model has Tampa Bay by 2.6. The Lean is Kalshi at 52.5¢ against a 55% fair value from the rest of the market. It's Kyler Murray's first real start as a Viking, behind a defense allowing 12.5 points a game, against an 0-2 team. That's the market's story, and Kalshi is selling it a little cheaper than everyone else. Full breakdown: Vikings vs. Buccaneers Week 3 prediction.

Ravens vs. Cowboys (Rio): Mark Andrews under 5 receptions

NO on 5+ receptions · 60¢ now (book 40/41 on the YES, about 4,700 open) · PLAY, HIGH tier.

This is the deepest prop book on the card, so you can get size down without moving it. The risk is Zay Flowers. Baltimore's leading receiver is questionable with a hamstring after only individual drills Friday, and the timing plan above is built around his status. One more note: the game page leads with the 4+ rung because it's closest to our projection. The rung that clears the publish gate is 5+. Full breakdown: Ravens vs. Cowboys Week 3 prediction · Mark Andrews.

Raiders at Saints: Pass

New Orleans lost its rookie left tackle, and Las Vegas is 2-0 against two teams with a combined 0-4 record. Neither fact is a price. The one prop that interests us, Travis Etienne at 2+ catches, sits in a backfield with an absence our model doesn't account for, so it stays on WATCH. Full breakdown: Raiders vs. Saints Week 3 prediction.

Sunday night

Rams at Broncos: Bo Nix under 2 touchdown passes

NO on 2+ pass TDs · 59¢ now (book 41/42 on the YES, about 500 open) · PLAY, MODERATE.

Neither quarterback is on the injury report, and the board has the under at MODERATE. The Kickoff Board also shows a Lean on the Rams moneyline (Kalshi 54.5¢ against a 57% fair value, 1.4% quarter-Kelly). Both point the same way, toward a quieter night for Denver's passing game. Treat them as one idea, not two positions, and remember Puka Nacua is doubtful for Los Angeles. Full breakdown: Rams vs. Broncos Week 3 prediction · Bo Nix.

Monday night

Eagles at Bears: Pass (Oracle Fork)

Chicago has three quarterbacks, and two of them are hurt. Caleb Williams strained a hamstring, his backup is in the concussion protocol, and Ben Johnson hasn't named a starter. The page publishes a line with Williams and a line without him, and no position comes off either until the starter is confirmed. Saquon Barkley at 70+ rushing yards is on WATCH, and that book is too thin to trade anyway: a 48/53 spread with about 70 contracts open. Full breakdown: Eagles vs. Bears Week 3 prediction.

Before you take any of these

Our probability and the edge on every prop above are on each game page's prop table for Pro members. The contract, the side and the price are free, here and on the full prop board.

Every model line for the week, every forked quarterback and every Kalshi price is on the NFL Week 3 predictions hub. If you want the Kalshi-vs-books version of this card, the Oracle Ladders board has it.

Prediction markets are not sportsbooks. Kalshi is a CFTC-regulated exchange, and contracts trade between 1¢ and 99¢ as market-implied probabilities and settle on the real outcome. Availability varies by state. Everything here is analysis with a stated position attached, recorded at a stated price on a stated date. Nothing here is advice. Trade responsibly.

Frequently Asked Questions

How was one play picked for each game?

Each game's play comes off its own PMP game page, in a fixed order. First choice is a player prop at PLAY status on our prop board, meaning it cleared every publish gate. Next is a moneyline the page's Kickoff Board marks as a Lean. If neither exists, the game is a Pass, and we name the contract we're watching and say why it isn't a play. Every play here is the same call its game page makes.

Why are six of the fifteen games a pass?

Because their game pages say pass. Two are Oracle Forks, where the starting quarterback isn't confirmed. One is a QB-adjusted game running on a corrected rating. In the other three, every candidate row is either flagged as a conflict or sits on WATCH. Putting a play in those slots would mean this article disagrees with the page it links to.

Why take the Panthers when your model has the Browns?

Because it's a price play, not a side play. The Panthers Lean is Kalshi at 55.5 cents against a 59% fair value built from the rest of the market's line. Our own ratings model has Cleveland by 1.5. The Vikings Lean works the same way. The Cardinals Lean is the one where the ratings model and the price point in the same direction.

Are these prices still live?

No. Every price here was read off Kalshi at about 7:20 PM ET on Saturday, Sept. 26, 2026. Props move fastest on Sunday morning, when inactives post. Check the game page, and the contract itself, before you take a position.

Should I take these now or wait for a better price?

It depends on the play, and each one says which. The three moneyline Leans should be posted now as resting orders and never taken at the ask, because the fee eats the whole gap. The quarterback touchdown unders carry no status risk, so they can be built tonight. The receiver and back props (Andrews, Johnson, LaPorta) get a small starter tonight and the rest after inactives post, about 90 minutes before kick.

What does walk away mean?

It's the most we'd pay on a resting order: our number minus our minimum edge, which is 5 points on a prop and 2.5 on a moneyline Lean. If you have to cross the spread to get filled, take roughly another 2 cents off for the taker fee. Above the walk-away, the position isn't worth having.

Where are the model probabilities and edges?

On each game page's player-prop table, for Pro members. This article gives you the contract, the side, the price to pay and the tier for free.

Read Next

Go Pro · 26% off your first year

The read is free. The position is Pro.

Pro turns the read into a position: the cross-market arb scanner, Gridiron Edge NFL against live Kalshi prices, Thee Oracle, and edge alerts in Discord the moment a gap opens.

Claim 26% off →

$111 for the first year, then $150/yr. Code applies automatically. Ends Dec 31. Monthly is $14.99/mo.

Every signal graded in public, losers left on the board. See the record →

Prediction-market edges, weekly. The only newsletter built for Kalshi & Polymarket traders.

No spam. Unsubscribe anytime.

7O

The 7 Oracles

Prediction Markets Desk

The 7 Oracles is the analyst collective behind PredictionMarketsPicks — calibrated, data-first coverage of Kalshi, Polymarket, DraftKings, and FanDuel markets.

About the Oracles →
NFL Week 3 picksNFL Week 3 predictions 2026NFL Week 3 player propsKalshi NFL Week 3Justin Herbert passing touchdowns propTrevor Lawrence passing touchdowns propBo Nix passing touchdowns propSam LaPorta receptions propMark Andrews receptions propEmmett Johnson receptions propPanthers vs Browns predictionVikings vs Buccaneers prediction

Want more analysis like this?

Get The 7 Oracles' daily prediction market breakdown — free, no fluff, straight to your inbox.

Get the daily edge in your inbox →