Graded, 8/15Buffalo led 13-7 at halftime and was in front by three scores late in the fourth. The first-half call below lost. What follows is the read as it stood at 11:07 a.m. ET, two hours before kickoff, with the prices that were actually on the board, plus the grade wherever it lands. We post the misses the same way we post the hits.
Preseason handicapping is one question wearing a lot of costumes. Who is actually on the field, and for how long. Everything else comes after that. Season stats mean nothing when the guy who produced them is standing on the sideline in a baseball cap, and a spread means nothing when you cannot tell whether a starter gets one series or ten.
Both coaches answered the question this week. They answered it with wildly different amounts of precision, and that gap was the whole bet.
The Quick Answer
Buffalo was a 3.5-point home favorite for Saturday's 1 p.m. ET kickoff at the new Highmark Stadium, NFL preseason Week 1 and Buffalo's first game of the summer. The total sat on three different numbers depending on where you looked, and the Bills moneyline ran from -166 to -190. The play was Carolina's first-half moneyline at +163 at NoVig and +162 at FanDuel, a bet you could also find at +135 if you were not paying attention. Below: what each coach said about his rotation, why Carolina's second unit had a head start, the four books still hanging the key number of 3, and the under that was a full point cheaper than the one everybody else was taking.
Carolina Gave A Number. Buffalo Gave Nothing.
Dave Canales told reporters on August 9 that Bryce Young and the Panthers starters would play 10 to 20 plays at Buffalo, similar to the workload Kenny Pickett got in the Hall of Fame Game. Joe Person, Alex Zietlow and Sheena Quick all filed it the same day.
A count is the most useful thing a coach hands you in August. You can build a first-half projection around 10 to 20 plays. You cannot build one around a vibe.
Buffalo never gave the vibe either. Go back through two weeks of Bills camp reporting and there is no snap number on Josh Allen anywhere. Joe Brady talked about C.J. Gardner-Johnson's calf, the club extended right guard O'Cyrus Torrence, Matt Parrino filed his practice observations, and across all of it nobody ever put a workload on the franchise quarterback for the first game in the new building. Maybe nobody asked. Either way, you were handicapping the length of Allen's afternoon with nothing on the record, and that is worth pricing: a coach with a full house, a brand new stadium, every reason to protect the guy and none to show him off.
Dave Loughran framed the method on Friday's preseason show about as cleanly as it can be framed: you are not psychoanalyzing these guys, you are reading who a coach is working with, where he sits in his own tenure, and whether the teams held joint practices. Two of those three were available here. The third was not, and it mattered, because these two never got on a practice field together. No joint sessions in the camp wire, no beat writer telling you one offense beat up the other for two days. Losing the read that usually settles a preseason side left a hole nothing else fills. Less information than normal, which is exactly why the price you paid was doing more of the work than usual.
One Team Had Already Played A Football Game
Carolina won the Hall of Fame Game 33-30 over Arizona on August 6, on a walk-off touchdown run by rookie Haynes King, who went 21-of-34 with two scores through the air and added 39 yards on the ground. Zietlow's note afterward was that the 63 combined points were the second-most in the history of that game. Whatever you think of the level of football, Carolina's backups had live reps, a plan they had already run, and a week of corrections since. Buffalo's had an open practice on August 8, where the gates opened to a band playing "Welcome to the Jungle" and Josh Allen jogged out in the Nickel City uniform for the first time.
That was the asymmetry, and it is why this was a first-half bet and not a full-game bet. In a game where the second unit plays most of the snaps, the second unit that has already played a game is worth something.
The quarterback rooms said the same thing from a different angle. Behind Young, Carolina had Pickett, who has started NFL games and started this preseason, and Kyle Trask, signed on August 9 after King pulled a hamstring, and who spent 2023 in Tampa Bay working under Canales. Behind Allen, Parrino's August 7 notebook still had the QB2 job listed as up for grabs, and nothing in the following week resolved it publicly. One team knew who its backups were.
Carolina was not walking in healthy, and the holes are worth naming:
- Chuba Hubbard Was Out, week to week with a hamstring per Canales via Mike Kaye on August 13, after an MRI. That left Jonathon Brooks as the back most likely to see the early work, and Adam Schefter reported on August 12 that the third-year man was appearing in the first preseason game of his NFL career. Canales said he was "ready to rock." Etienne remained ahead of Dillon on Quick's unofficial depth chart.
- Haynes King Was Week To Week with a hamstring of his own, which is why Trask is on the roster at all. The guy who won the Hall of Fame Game was not part of this one.
- Darren Waller Signed A One-Year Deal on August 12. Mike Kaye reported the move was not injury related and that Waller had played under new Panthers associate head coach Darrell Bevell in Miami. A name for the fall, not a factor on a Saturday afternoon.
Buffalo had its own subtractions. Schefter reported Friday that DT Phidarian Mathis was suspended three games for violating the league's substances of abuse policy, and Gardner-Johnson had been working back from the calf strain that put him on a cart on August 1. Neither moves a preseason first half much. Both were worth knowing before deciding how seriously to take the Bills' front in the second quarter.
The First Half Was Where The Number Lived
Here is the payoff to that opening promise, and it comes with a caveat you should read before the table.
Eighteen sportsbooks posted a first-half moneyline on this game and they did not agree on Buffalo. Strip the vig out of each pair and you can see what each one actually believed about Carolina. Every price came off the OddsShopper live odds screen, which stacks every book posting a market on one page instead of the one your app happens to open on.
| Venue | Carolina 1H | Buffalo 1H | Hold | De-vigged Carolina |
|---|---|---|---|---|
| NoVig | +163 | -194 | 4.01% | 36.6% |
| FanDuel | +162 | -225 | 7.40% | 35.5% |
| BetRivers, Bally, Unibet, BetParx, SugarHouse, NorthStarBets | +155 | -195 | 5.32% | 37.2% |
| Caesars | +154 | -190 | 4.89% | 37.5% |
| BookMaker | +152 | -180 | 3.97% | 38.2% |
| BetMGM, SportsInteraction, Borgata | +150 | -185 | 4.91% | 38.1% |
| Bet365 | +150 | -180 | 4.29% | 38.4% |
| Fanatics | +150 | -180 | 4.29% | 38.4% |
| DraftKings | +145 | -188 | 6.09% | 38.5% |
| OnyxOdds | +145 | -188 | 6.09% | 38.5% |
| TheScore | +135 | -180 | 6.84% | 39.8% |
FanDuel is the fun row in there. It was the most Buffalo-leaning venue in the group and it still handed out the second-best Carolina price, because it ran a 7.40% hold, the widest on the board, which shoves both sides outward. Buffalo went to an unplayable -225. Carolina went past almost everybody.
Now the part most previews skip. Average those de-vigged views and the field's fair price on Carolina's first half was about 37.7%. NoVig's +163 breaks even at 38.0%. So this was never a mispricing. The market had it right, and what you were buying at +163 was your own read on the rotation at the best price anyone would sell it to you.
Real bet, small bet. If you thought a defined 10-to-20-play Carolina starter package plus a second unit with a game already under it was worth more than 38% in a 30-minute market, you took it at the top of the board and nowhere else. At +145 you were paying DraftKings 18 cents for the privilege of having an opinion. At +135 you were donating. Twenty-eight cents separated the best and worst price on the same 30 minutes of football, which is more than most people's handicap is worth in the first place.
Take 20% off OddsShopper Pro with code PANTHERSBUF20and stop finding out after kickoff that the number you took was the worst one posted.
The first quarter was louder still, if your book took it. Carolina's first-quarter moneyline ran +155 at BetRivers, Bally, Unibet, BetParx, NorthStarBets and SugarHouse and +154 at FanDuel, down through +135 at DraftKings and OnyxOdds and +120 at Fanatics, all the way to +102 at Rebet. Call it a 53-cent range on one side of 15 minutes of football, and Buffalo's side was scattered right along with it, from -160 at Fanatics out to -210 at FanDuel. A spread that wide is a market admitting nobody has any idea. Eleven books is thin and first-quarter limits are thinner, so call it a look and not a lean. But if the announcements were what you were betting, the opening quarter was the purest version of them.
The Full Game, And The Four Books Still Sitting On 3
Back to the whole game, where the story was the number rather than the side.
Three is the most common margin of victory in football, so a half-point that crosses it is one of the few point purchases worth making. Our NFL key numbers guide puts the cheap threshold at roughly ten cents of juice, and this board was offering that trade in the direction people usually have to pay for.
| Where | Carolina | Cost of crossing 3 |
|---|---|---|
| ProphetX, WynnBET, CoolBet, Bet99 | +3 at -103 to -110 | reference price |
| Bet365, Fanatics, Circa | +3.5 at -110 | about 7 cents |
| Kalshi | +3.5 at -126 | 23 cents |
| Rebet | +3.5 at -133 | 30 cents |
Twenty-nine of the 33 books posting a spread had moved to Buffalo -3.5. Four were still hanging 3: ProphetX at -103, WynnBET and CoolBet at -105, Bet99 at -110. Off ProphetX's -103, buying up to the +3.5 that Bet365, Fanatics and Circa had at -110 cost about seven cents. Take it.
Bet99's row is the one to stare at. Its +3 was priced at -110, the same -110 three other books wanted for +3.5. Identical price, and one of them hands you the most common margin in football for free. There is no argument for the shorter number there. If that was your only book, you were giving away a key number to nobody.
Just do not treat that half-point as worth what it is worth in November. Preseason games run with backup kickers and long stretches of clock, and they land on exactly 3 a little less reliably than real ones. The math still favors crossing. It favors it less emphatically than the spread pricing mechanics would suggest in a regular-season game.
The moneyline was the same shopping lesson in dollars. Circa paid +165 on Carolina, NoVig +160, BookMaker +158, ProphetX +155. Rebet paid +138 for the identical ticket. Buffalo backers could take -166 at DraftKings or FanDuel, or hand Circa, BetMGM, PointsBet, SportsInteraction or Borgata -190 for the same result. Call it 27 cents of spread on the Carolina side and 24 on the Buffalo side, decided entirely by which app you happened to have open.
As for the full-game moneyline, the market had already settled and there was nothing left to argue about. Pinnacle had Carolina +153 and Buffalo -183, which de-vigs to a 37.9% Carolina win. Kalshi, an order book rather than a sportsbook, had Carolina +146, which is a shade over 40 cents in exchange terms, against Buffalo -175, and de-vigs to 39.0%. Two completely different price-discovery mechanisms landing a point apart is the signal to stop hunting for a full-game moneyline edge, and it is why this piece spent its energy on the first half. Our guide to trading NFL on Kalshi covers the fee math that eats into a quoted exchange price, and the wider board for the week sits on our NFL preseason odds page.
The Total Was Three Different Numbers
Preseason totals usually carry a public-over lean. Casual money floods overs, books know it, and August offenses are conservative anyway. This board did not show one.
Twenty books had this at 38.5, seven had it at 38 including Pinnacle, and five had it at 37.5. De-vig Pinnacle's 38 and you get 49.7% over. De-vig the 38.5 consensus and you get almost exactly 50.0%. A coin flip, no shading in either direction, which means there was nothing to fade at the posted number and the totals work was entirely about what you paid.
So the total was a shopping exercise and nothing else, and the money in it needed no opinion at all. Wanting the under at 38.5, six books had it at -108 while DraftKings charged -120 for the identical bet. Twelve cents for nothing. Wanting the over at 38.5, NoVig paid +104 while Fliff charged -115, a nineteen-cent gap on the same ticket. Moving off the number was a real trade rather than a free one: OnyxOdds would sell the under at 37.5 for even money, but 37.5 is a full point harder to win than 38.5, and the 20 cents you saved is what you paid for that point. ProphetX's under 38 at -101 sat in between. Nobody handicapped anything to collect any of that.
One angle to ignore entirely was the building. This was the first game ever played in the new Highmark Stadium, so there was no venue scoring history and no established wind pattern to lean on. Anyone selling you a stadium effect on this one was selling you a guess with a nice graphic on it.
Nobody Would Post A Touchdown Prop
Not one sportsbook on our screen had anytime or first touchdown markets up for this game. Correct call, and worth copying. With Hubbard out, Brooks making his first preseason appearance, and both starting offenses leaving early, goal-line volume was unknowable. A market nobody will price is not a market you should be trying to beat.
Where The Read Actually Broke
Carolina's first half at +163 was the play, taken small, and it lost 13-7. Worth being precise about how, because the how is the part that transfers.
The pregame announcement was clear enough. Canales said 10 to 20 plays. What did not hold up was the inference stacked on top of it, that a defined starter package plus a second unit with a Hall of Fame Game already behind it was worth more than 38% over 30 minutes. Buffalo scored ten in the first quarter and Carolina did not answer until the second. The half was effectively decided before the part of the game I had the most confidence in.
That points at the one line in this piece I would move. The first-quarter market got waved off here as a look rather than a lean, on the grounds that 11 books and thin limits are not worth chasing. It was also the 15 minutes that beat the bet, and a 53-cent range across those 11 books says the market was no more sure of it than I was. Next Saturday I want the quarter priced properly instead of shrugged at.
And the side that beat us was the quiet one. Carolina gave a number, Buffalo gave nothing, and the team with nothing on the record won the half by six. What I cannot tell you yet is why, because I do not have Buffalo's snap counts. A long Josh Allen afternoon and a short one produce very different lessons here, and 13-7 is consistent with both. Blaming the silence read on a scoreline alone would be the same results-oriented thinking this piece spent two thousand words avoiding. So that is Monday's job: pull the participation report, find out how long Allen actually played, and only then decide whether treating an undisclosed workload as a soft input was the error or just the excuse.
Every price above came off one screen. OddsShopper Pro puts all 30-plus books on a game in a single view with the best number flagged, which on this board was the difference between +163 and +135 on a bet you were making anyway. Use code PANTHERSBUF20 for 20% off.
For the rest of the card, the free expert picks today page carries where our staff landed. If a 33-book board is new to you, the odds screen walkthrough shows how to cut it down to the number you want in a few clicks, our NFL spread betting primer covers the pricing underneath all of it, and the preseason betting guide explains why rotations beat rosters in August.
Odds pulled Saturday, August 15 at approximately 11:07 a.m. ET, roughly two hours before the 1 p.m. ET kickoff. Please bet responsibly.
More on this: Browns Vs Bears Preseason Odds: The Best Price On Cleveland -3 · Cowboys Vs Seahawks Preseason Odds: Cowboys +2.5 Pays +108 · Eagles Vs Ravens Preseason Odds, Predictions And Picks · Rams Vs Chiefs Preseason Odds: Rams First Half Pays +125 · Vikings Vs Giants Preseason Odds, Predictions And Picks
