Falcons @ Packers: a quarterback comes back, the number moves two points toward him, three ratings say it should not have — and the props are about which quarterback's tape you believe
Green Bay (1-1) hosts Atlanta (0-2) at Lambeau in perfect weather, and the whole handicap is one man: Michael Penix Jr. makes his first start in ten months, named Monday, full practice Tuesday. The Falcons with backups scored 16 points in two games (32nd in scoring, 32nd in passing, 3.5 turnovers a game); the Falcons with Penix last October and November threw to Drake London for 158, 118, 104 and 119. The market moved the number from Atlanta +6 to +4.5 on the news and the total from 46.5 to 43 on the two weeks of tape. Three published rating systems still make it Green Bay by eight to nine. That is the argument on the side, and it is printed in full below. The props follow the quarterback change: London's tape with Penix says over, Pitts's tape with Penix says under, and Bijan's median says the line forgot who touches the ball. Every rated player is on the NFL Insights Board.
Game-state groups on this card (Lesson 10): Penix-throws-to-London → London O69.5 + Dotson U17.5 (two legs, one script; Zaccheaus's under is the third and sits on WATCH by the cap); Atlanta-trails-and-Bijan-catches → Bijan O119.5 R+R (one leg); Packers-control-it → GB −4.5 (one leg — and note it needs the opposite of the Bijan script in the second half). Two groups of one, one group of two.

Falcons @ Packers — Thursday Night
Thu 8:15 PM ET · Lambeau Field · Green Bay, WI
Venue · weather
Lambeau · outdoors · 62°F at kickoff, wind 4–5 mph, 0% rain (read 6:17 PM)
Line
GB −4.5 (−107) / ATL +4.5 (+104) on our data platform · −4.5 at three of four Nevada books, −5 at one; the public book still shows −5.5 (the worse number for the favorite is the better one for the dog) · opened −6 / −6.5
Total
43.5 (O +106 / U −108) · 43 at every Nevada book · opened 46.5 · ratings' projected totals 43.2 / 46.1
Ratings
GB −8.0 / −9.1 / −8.5 — all three outside the number by 3.5 to 4.6
⚡ THE JUICE. Atlanta's 2026 is two games of Cooper Rush and Jack Strand: 8 points a game, 123.5 passing yards a game, a 0.0% red-zone rate, 3.5 turnovers a game — every one of them 31st or 32nd. The defense has been the opposite: 3.8 yards a play allowed (2nd), 63 rushing yards a game allowed (1st), 175 passing (8th), 3.5 sacks (4th). Penix returns to that. Green Bay's problem is the receiver room: Jayden Reed was carted off on the second play at the Jets and is OUT with a neck injury, and the coach said it is too early to know if he plays again this year. Watson (27.5%) and Golden (26.1%) already had the targets; Kraft is the third option. Green Bay passes 67% of the time and has allowed 28 a game (T22nd) — a defense the Falcons' backups still could not score on. The matchup tab's first-two-weeks summary reads 16–42 in points and 247–489 in passing yards, Atlanta's way in rushing 277–129. The platform's own money: moneyline 79% of bets / 59% of money Green Bay; spread 67% / 66% Green Bay; total 51/49.
💸 THE CROWD VS THE CASH. The public book (5:57 PM): Atlanta +5.5 at 46% of the money on 36% of tickets (+10 — the dollars lean the dog while the tickets lean the favorite); moneyline Atlanta 60% of money on 23% of tickets (+37 on a +200 dog); total Over 69% money / 60% tickets. The market-maker (+4.5): Atlanta 55% of the money on 29% of tickets (+26, SHARP on the dog); Green Bay 45% money / 71% tickets; total Over 58/51; moneyline 94% of money Green Bay. The tracker: opened Atlanta +6 (market-maker) and +6.5 elsewhere, now +4.5 at three books and +5 at one — a two-point move toward the dog against 64–71% of tickets on the favorite: RLM by definition, and news by explanation (Penix). The total fell 46.5 → 43, three and a half points, on two weeks of Atlanta scoring eight. Three rating systems: Green Bay by 8.0, 9.1 and 8.5 — every one of them 3.5 to 4.6 outside the current number, projected scores 24.5–27 to 17–20. Systems: the network's weekly splits-systems report was not captured this run — none printed. Expert room (44 posts on this game): sides — Atlanta +6.5 twice, Atlanta +4.5 once, Atlanta moneyline +288 once (two units), Under 43.5 once, nothing on Green Bay; props — London overs by six experts (five on 5.5 receptions, three on yards 59.5–64.5), Pitts overs by five, Watson over 64.5 by four, Bijan rush over 83.5 and attempts over 18.5, Love under 229.5 and under 31.5 attempts, Kraft over 43.5 and a touchdown, Lloyd over 26.5 AND under 26.5 (two experts, opposite sides), Kaleb Johnson under 38.5 (the network's model, 25.9% edge).
🎯 THE FINAL VERDICT: PACKERS −4.5 (−107 on the platform; −4.5 at three Nevada books, take nothing worse) — conviction 53. The three ratings are the thesis: they make this an eight- to nine-point game, and the current number is four and a half. The public-book tickets agree. The argument against, in full, and it is loud: the sharp dollars are on Atlanta at BOTH books (+26 at the market-maker, +10 at the public book), the line moved two points toward the dog against the tickets (RLM — our own record grades RLM leans 1-for-9, a tiebreaker not a thesis, and here it is explained by a returning quarterback the ratings may not fully carry), the room has four experts on Atlanta and zero on Green Bay, and Atlanta's defense is the second-best in the league by yards per play. That is why this is 53 and not 58: models against money, and Lesson 6 says the money is the tiebreaker, not the thesis — but the money here is not a whisper. If you want the dog, the +5 at one Nevada book and the +5.5 at the public book are the numbers; we are not publishing it. Total: STAY AWAY — the ratings' totals straddle the number (43.2 / 46.1), the Over money is public-shaped at both books, and the one number that decides it (does Penix make Atlanta an offense) is the one thing no window measures.
The positional board — Green Bay (QB · WR1 · WR2 · WR3 · RB · TE)platform panels captured Thu Sep 24, 6:13–6:15 PM ET · 17 rated · WR/CB tool not captured this run
Love U224.5 pass yds★★★☆☆ 6.0+111
Projection vs linemedian 205 · avg 216.5 vs 224.5
Fantasy median vs DFSFS median ~17 vs PP 17 / UD 17.15 — on the number
Hit windowsL10 50 · L5 60 · 2026 50 (387 @ MIN, 145 @ NYJ) · 2025 56
Opponent · roleATL 24th vs pass yds on the platform's rank, but 8th in 2026 yards allowed (175/g) · 30.6 att, 55% pass rate · Reed OUT
The median is 205 and the line is 224.5 — twenty yards of cushion on the under, at plus money. The reason it is a WATCH: the windows are coin flips (5 of 10), the platform ranks Atlanta 24th against the pass, and Green Bay just threw 46 times at the Jets with Reed gone after two plays. The network's props writer has the under at 229.5; the model in the same building has the attempts under 31.5 at an 8.8% edge — Atlanta faces 38.5 attempts a game (29th), which is the wrong way for both. Rated, not bet.
Median 205 vs 224.5Windows 50%ATL faces 38.5 att/gPlus price
WR1GBNO PLAYChristian Watson
Watson O71.5 rec yds−104
Projection vs linemedian 47.5 · avg 64.7 vs 71.5
Fantasy median vs DFSFS median ~12 vs PP 14 / UD 11.95
Hit windowsL10 40 · L5 40 · 2026 50 (147, 41) · 2025 36
Opponent · roleATL 25th vs rec yds · 21% target share (27.5% in 2026, Reed OUT) · WR/CB read not captured this run
Four experts in the room have him over 64.5 and the case is real — 27.5% of the targets with Reed gone, Atlanta 25th against receiving yards. The tape says no: a 47.5 median against a 71.5 line, four of ten over, and a 2026 that is one 147 and one 41. The under has the cushion and the opponent rank runs the wrong way for it (Conflicts). Neither side.
Median 47.5 vs 71.5Reed OUT → share upUnder is Conflicts (ATL 25th)WR/CB read not captured this run
Golden U59.5 rec yds★★★☆☆ 6.0−122
Projection vs linemedian 18 · avg 34.5 vs 59.5 — 2026: 95, 58
Fantasy median vs DFSFS median ~7 vs PP 11.5 / UD 8.85
Hit windowsL10 80 under · L5 60 · 2026 50 · 2025 87
Opponent · roleATL 25th vs rec yds · 13% target share in the window, 26.1% in 2026 · WR/CB read not captured this run
The screen loves the under — eight of ten, an 18 median, an 87% rookie season. Every one of those numbers was earned as the fourth receiver; in 2026 he is the second target on the team at 26% and has gone 95 and 58. The continuity gate says the 2025 pie is a different pie, and the 2026 pie says the line is about right. WATCH.
2025 windows = WR4 role2026 share 26%Median 18WR/CB read not captured this run
Moore U9.5 rec yds+117
Projection vs line2-target role · 2026 100% under
Fantasy median vs DFSFS PP 3.5
Hit windowsL10 90 under · 2026 100 · 2025 74
Opponent · roleATL 25th vs rec yds · 5.8% target share · WR/CB read not captured this run
A two-target receiver whose under cashes nine of ten — priced at +117 because one catch beats it. Not a bet, and the WR3 slot with Reed out is a tile, not a leg.
Plus price on a 2-target underNo role
Lloyd U24.5 rush yds★★☆☆☆ 5.0−113
Projection vs line2026: 37, 20 · avg 28.5 vs 24.5 · 3.1 ypc
Fantasy median vs DFSFS median ~6 vs PP 6.5 / UD 5.45
Hit windows2 games only (1 of 2 over)
Opponent · roleATL 1st vs rush yds (63/g allowed) · 48.7% rush share of a 33%-run-rate offense (19 att in two games)
Atlanta is the No. 1 run defense in the league and Green Bay runs a third of the time, so the over is Conflicts by rule. The under has the matchup and no cushion: the line opened 31.5 on Monday and has been cut to 24.5 — seven yards of the news already spent. Two experts in the room took opposite sides of 26.5. WATCH.
ATL 1st vs rushLine 31.5 → 24.5 (spent)Two-game sample
Kraft O49.5 rec yds+111
Projection vs linemedian 49.5 · avg 56.9 — the line IS the median
Fantasy median vs DFSFS median ~9 vs PP 11 / UD 8.75
Hit windowsL10 50 · L5 60 · 2026 50 (65, 15) · 2025 50
Opponent · roleATL 25th vs rec yds · 16% target share (13% in 2026) · 57% of his targets land where this defense is tough
Half his targets go to zones Atlanta defends well, the median sits exactly on the line, and every window is a coin flip. The props writer has him over 43.5 — six yards lower than the board — and that is the honest price. No play at 49.5.
Median = lineCoin-flip windowsReed OUT → share up
The positional board — Atlanta (QB · WR1 · WR2 · WR3 · RB · TE)platform panels captured Thu Sep 24, 6:13–6:15 PM ET · 17 rated · WR/CB tool not captured this run
QBATLWATCHMichael Penix Jr.
Penix O204.5 pass yds★★★☆☆ 5.5−106
Projection vs linemedian 221 · avg 212.8 vs 204.5 (five 2025 starts: 250, 241, 221, 177, 175)
Fantasy median vs DFSFS median ~13 vs PP 13 / UD 12.65
Hit windowsL5 60 (2025 only) · no 2026 window — first start since the ACL (Nov 2025)
Opponent · roleGB 8th vs pass yds — wrong way for the over (opp 0, Conflicts) · named the starter Monday, full practice Tuesday
The first start in ten months, against the 8th-ranked pass defense, behind an offense that ranks 32nd in scoring (8 a game) and 32nd in passing (123.5) with two backups. The median from his five 2025 starts clears the line by 16, and none of those were off an ACL. The rank makes it Conflicts; the situation makes it a WATCH. One expert has the interception under.
First start in 10 monthsGB 8th vs pass — Conflicts2025 median 221
London O69.5 rec yds★★★★☆ 8.0+105
Projection vs linemedian with Penix 118 (158, 42, 118, 104, 119) · full-window median 64.5 vs 69.5
Fantasy median vs DFSFS median ~14 vs PP 14 / UD 10.55
Hit windowsL10 50 · L5 20 (backup QBs) · Penix starts 4 of 5 over · 2026 0 for 2 (29, 51 with Rush/Strand)
Opponent · roleGB 10th vs rec yds · 26% target share, 55% of the team's air yards · WR/CB read not captured this run
Two quarterbacks, two receivers. With Penix under center last season London went 158, 42, 118, 104 and 119 — four of five over this number by thirty-plus; with Rush and Strand he went 27, 4, 78, 29 and 51. The 2026 pie is the backups' pie and the line was set from it; the pie that starts tonight is Penix's, where London owns 55% of the air yards. Argument against: a quarterback in his first game off an ACL, Green Bay 10th against receiving yards, the plus price on a prop (we prefer minus), and no WR/CB read this run. Three experts in the room have London yards overs at 59.5–64.5.
Penix window 4 of 5 over55% air-yard sharePenix first start off ACLWR/CB read not captured this run
Dotson U17.5 rec yds★★★☆☆ 6.5−104
Projection vs linemedian 7 · avg 9.8 vs 17.5
Fantasy median vs DFSFS PP 4
Hit windowsL10 80 under · L5 60 · H2H 100 · 2026 50 (19, 11) · 2025 72
Opponent · roleGB 10th vs rec yds · 8% target share, 8% air-yard share · WR/CB read not captured this run
Ten and a half yards of cushion on a receiver who gets eight percent of the targets and has failed to reach 24.5 in nine straight road games. The under is the same bet as the London over — Penix throws to London — and it is counted with it. Argument against: the QB change resets every window (his 2026 games are 19 and 11 under backups), and one 40-yard catch beats it.
Median 7 vs 17.58% shareQB change resets windowsWR/CB read not captured this run
WR3ATLWATCHOlamide Zaccheaus
Zaccheaus U9.5 rec yds★★★☆☆ 6.5−104
Projection vs linemedian 6.5 · avg 9 vs 9.5
Fantasy median vs DFSFS PP 3
Hit windowsL10 80 under · L5 100 · H2H 67 · 2026 100 · 2025 50
Opponent · roleGB 10th vs rec yds · 6% target share · WR/CB read not captured this run
Eight of ten and five straight under, on a 6% target share. It would be the third leg on the same script (Penix throws to London) and Lesson 10 caps that group at two, so it sits on WATCH with the better number already published.
L5 100% underThird leg on one script — capped
B. Robinson O119.5 rush + rec yds★★★★☆ 7.5−113
Projection vs linemedian 155.5 · avg 140.6 vs 119.5
Fantasy median vs DFSFS median ~21 vs PP 20.5 / UD 18.75
Hit windowsL10 60 · L5 60 · 2026 50 (173, 81) · 2025 53 · road 4 of 5 over 120.5
Opponent · roleGB 9th vs rush yds (87/g), 10th vs rec yds · 63% rush share (58.7% in 2026), 6.3 targets at 75% catch
Thirty-six yards of median cushion on the one Atlanta player the offense runs through — 63% of the carries, six targets a game, and 173 total yards two weeks ago at Pittsburgh. Green Bay is 9th and 10th against the run and the catch, so the opponent is neutral; the script does the work: Atlanta trails, Bijan catches. Argument against: 81 last week, a 32nd-ranked offense that has scored 16 points all season, and a Green Bay run defense holding backs to 3.8 yards a play. One expert has the rushing over at 83.5 and another the attempts over 18.5.
Median 155.5 vs 119.563% share · 6 targetsATL offense 32nd
Pitts U39.5 rec yds★★☆☆☆ 4.0−125
Projection vs linemedian 41 · avg 52.3 vs 39.5
Fantasy median vs DFSFS median ~8 vs PP 8.5 / UD 6.05
Hit windowsL10 50 · L5 60 under · 2026 100 under (0, 15) · 2025 47
Opponent · roleGB 10th vs rec yds · 20% target share · his big games (82, 90, 166) came with Rush, his Penix games were 14 and 25
Five experts in the room have Pitts overs at 35.5–49.5, and the tape they are reading is the Rush tape — 82, 90 and 166 in December. With Penix at quarterback last fall he went 14 and 25. Tonight's quarterback is Penix. Neither side has a cushion (median 41 vs 39.5), so it is a WATCH, but the direction of the QB change points under, not over.
Median 41 vs 39.5Penix games: 14, 25Five experts over — dissent
Also rated — the second backs, the second tight end, the kickerplatform panels captured Thu Sep 24, 6:13–6:15 PM ET · 17 rated · WR/CB tool not captured this run
RB2ATLNO PLAYBrian Robinson
B. Robinson U24.5 rush yds−111
Projection vs linemedian 21.5 · avg 22.4 vs 24.5 · 2026: 31, 51
Fantasy median vs DFSFS PP 4
Hit windowsL10 60 under · 2026 0 for 2 · 2025 68
Opponent · roleGB 9th vs rush yds · 19% share in the window, 31.7% in 2026 (20 carries in two games) · full practice (ankle)
The window says under and the 2026 role says the window is stale — a third of the carries now, 31 and 51 the last two. No side.
Role grew in 20262026 0-for-2 under
K. Johnson U39.5 rush yds−110
Projection vs line8 carries in two games (20.5% share) against a 39.5 line
Fantasy median vs DFSFS PP 6
Hit windowsunder 100% in every window — he has never reached it
Opponent · roleATL 1st vs rush yds · third back behind Lloyd (48.7%) and Brooks (28.2%)
A 39.5-yard line on a back with eight carries in two games against the No. 1 run defense — the network's model calls it a 25.9% edge to the under and the board agrees. We did not open his panel this run, so it is printed as a WATCH, not published; a panel-less leg is not a leg.
No panel opened8 carries / 2 gamesATL 1st vs rush
J. Smith O8.5 rec yds−104
Projection vs linemedian 4 · avg 10.8 vs 8.5
Fantasy median vs DFSFS PP 3
Hit windowsL10 40 · L5 40 · 2026 50 (50, 0)
Opponent · roleATL 25th vs rec yds · 6% target share · half his targets land where this defense is tough
A 4-yard median against an 8.5 — one expert took the over on the 50-yard game at Minnesota. Four of ten. No.
Median 4L10 40%
Brooks U4.5 rush att−161
Projection vs line28% rush share · 11 carries in two games
Fantasy median vs DFSFS PP 5
Hit windowsL10 70 under · 2026 50
Opponent · roleATL 4th vs rush att
Five and a half carries a game on a −161 under with a half-carry cushion. The room has his yards under 16.5. Not at that price.
Price −161Cushion 0.5
Folk U6.5 kicking pts−127
Projection vs lineATL 8 points a game, 0% red-zone efficiency (32nd)
Fantasy median vs DFSFS PP 7
Hit windowsL10 60 under · 2026 50
Opponent · roleGB 7th vs kicking points
An offense scoring eight a game with a 0.0% red-zone rate makes the kicker's under tempting at −127 — but Penix's return is exactly the variable that changes it. Printed, not bet.
ATL 0% RZPenix returnsPrice −127
Screened and declined, on the record: Love O224.5 / O231.5 (−125 / −113; median 205 — the under is the side with the cushion and it sits on WATCH); Love U31.5 att (−110; ATL faces 38.5 attempts a game, 29th — Conflicts); Watson O71.5 / U71.5 (median 47.5 vs 71.5 says under, ATL 25th vs rec yds says Conflicts — neither); Golden U59.5 (−122; a WR4's windows priced against a WR2's 2026 role); Lloyd O24.5 (−100; ATL 1st vs rush — Conflicts) and U24.5 (−113; 31.5 → 24.5, spent); Kraft O49.5 (+111; median on the line); Penix O204.5 (−106; GB 8th vs pass — Conflicts; first start off an ACL) and U204.5 (+104; no cushion); London U69.5 (the backup-QB tape, not tonight's); Zaccheaus U9.5 (−104; the third leg on one script, capped); Bijan U78.5 rush (−102; median 84.5) and O78.5 (−113; the R+R line carries the read); Pitts O39.5 / U39.5 (median 41 — coin flip, QB change points under); B. Robinson U24.5 (−111; the 2026 role outgrew the window); K. Johnson U39.5 (−110; no panel opened — WATCH, not published); Jonnu Smith O8.5 (−104; median 4); Brooks U4.5 att (−161); Folk U6.5 (−127); every quarter, half, longest and touchdown market (unread by rule). Not captured this run and printed as absent: the WR/CB matchup tool (every receiver tile says so), the AI QB projection, the weekly splits-systems report.
The ledger
Side: Packers −4.5 (−107; 53) — models over money, dissent printed. Total: STAY AWAY. Props (three): London O69.5 rec yds (+105), Dotson U17.5 rec yds (−104), Bijan O119.5 rush + rec yds (−113). WATCH (rated, reasons printed, not bet): Love U224.5, Golden U59.5, Lloyd U24.5, Penix O204.5, Zaccheaus U9.5, Pitts U39.5, K. Johnson U39.5. Price note: every price is our data platform's board at 6:11–6:15 PM ET; a published leg grades at its first-published price. Stars print, do not size (Sep 22 promotion). Inactives post ~6:45 PM ET; if Penix is not the starter at kickoff, the London and Dotson legs are VOID by their own terms (the QB is the leg) and this card says so now. Sources: our data platform (games board 6:11 PM, props board 6:11–6:13 PM, 15 panels 6:13–6:15 PM, Matchup tab public betting and 2026 rankings), a Las Vegas betting network's two-book splits (5:57 PM), eight-book tracker with openers (6:15 PM), weekly ratings (6:17 PM) and expert room (6:16 PM), Open-Meteo at Lambeau (6:17 PM), the Falcons' Tuesday practice report and the Monday starter announcement (mainstream coverage Sep 23–24). NFL Insights Board.
Bet what you can afford to lose. If gambling is a problem for you or someone you know, call 1-800-GAMBLER.
What changed on this card
1 update · newest first
Thu Sep 24 · 6:22 PM ET First publish — Thursday Night card (v3 positional board, 12 + 5 extras). Our data platform: games board 6:11 PM ET, Props board filtered to ATL @ GB 6:11–6:13 PM (1090/1089 rows, 213 core-market rows), 15 Market Details panels 6:13–6:15 PM (Reed OUT note on the Injuries tab; Matchup tab public betting and 2026 rankings); a Las Vegas betting network's two-book splits 5:57 PM, eight-book tracker with openers 6:15 PM, weekly ratings 6:17 PM, expert room 6:16 PM (44 posts); Open-Meteo at Lambeau 6:17 PM. One side, three props, seven WATCH tiles. WR/CB tool, AI QB projection and the weekly systems report NOT captured this run — printed as absent. Inactives post ~6:45 PM ET: if Penix is not the starter, the London and Dotson legs VOID by their own terms. Stars print but do not size (Sep 22 promotion).
The card, in your inbox, before first pitch.
One email a day: every leg, every Edge Gap, and the argument against our own
top pick. When a leg misses, that's in there too — same email, same prominence.
No spam. Unsubscribe in one click. 18+ only — informational and
entertainment purposes, not betting advice.