Past Picks

San Francisco hosts Denver as a 3-point favorite with a 48.5 total. Christian McCaffrey's receiving line is set high at 43.5, and even though the matchup grades as great for him, every projection we track lands below it. He was a full participant on the final injury report.

Our model projects 31.5 receiving yards and makes the under a 67% play at 45.5. We're getting 43.5 at -115, two yards lower, and the outside sources still clear it comfortably: Sleeper 28.8, the expert consensus 36.6, another projection model 39.5 and ESPN 40.9. Big receiving lines on running backs are often priced off the ceiling; we're betting on the typical game. The play: McCaffrey under 43.5 receiving yards.

Dallas visits Houston as a 3-point underdog with a 48.5 total, and the matchup grades as poor for Javonte Williams in the passing game. Three catches beats this under.

Our model projects 1.7 receptions and makes the under a 61% play. The expert consensus (2.4) and another projection model (2.1) agree; Sleeper and ESPN sit just over at 2.6. Those are fractions of a catch, and at -105 we need only 51.2%, so the model's read plus a poor matchup gives us the edge. The play: Williams under 2.5 receptions.

Philadelphia hosts the Rams as a 4.5-point underdog with a 42.5 total, and the matchup grades as poor for tight ends. Dallas Goedert was ruled out with a knee injury, which puts more of the tight end work on Zach Ertz, and that's the risk for this under.

Even with that, the projections sit below the line. Our model projects 14.5 receiving yards and makes the under a 66% play at 23.5; we're getting 21.5 at -110. ESPN (16.6), Sleeper (18.7) and another projection model (10.8) are under the number, and the expert consensus is just over at 23.1. A poor matchup and a seven-yard gap between the model and the line keep us on the under. The play: Ertz under 21.5 receiving yards.

Philadelphia is a 4.5-point home underdog to the Rams, the matchup grades as poor for tight ends, and Zach Ertz needs three catches to beat this under. Dallas Goedert was ruled out with a knee injury, which adds to Ertz's role, but every projection still has him short of three catches.

Sleeper (2.0), the expert consensus (2.4), ESPN (1.9) and another projection model (1.1) are all under 2.5, and our model projects 1.6 receptions with a 61% read on the under. The price is -125 (break-even 55.6%), so the edge is moderate, but four of four outside sources under the number is what we want with a teammate out. The play: Ertz under 2.5 receptions.

Minnesota hosts Miami as a 10-point favorite with one of the lowest totals of Week 4 (38.5). A big favorite runs the ball, which is why Aaron Jones's rushing line sits at 67.5, but the matchup grades as poor for him on the ground, and he was limited on the final injury report.

Our model projects 54.5 rushing yards and makes the under a 63% play at 66.5; we're getting 67.5 at -110, a yard better. Sleeper (66.6) and another projection model (62.9) are under the line, while the expert consensus (72.2) and ESPN (82.5) are over. A 13-yard gap between the model and the number, a poor matchup and a limited practice status are enough for the under. The play: Jones under 67.5 rushing yards.

Seattle hosts the Chargers as a 7-point favorite, and its backfield is thin: Zach Charbonnet and Jadarian Price were both ruled out, and George Holani was limited on the final injury report. That points more carries Holani's way, which is the obvious risk for this under, but the matchup grades as poor for running backs.

Our model projects 26.5 rushing yards and makes the under a 57% play at 31.5. Sleeper (27.5) and another projection model (27.2) also sit under the number, while the expert consensus (32.2) and ESPN (37.5) are over. The line already reflects the extra work; the question is whether it overshoots, and the model and two of the four projections say it does. The play: Holani under 31.5 rushing yards.

Miami visits Minnesota as a 10-point underdog in a game totaled at just 38.5. A big underdog throws to catch up, and Malik Washington's rushing line of 5.5 depends on gadget carries that tend to disappear in that script. The matchup also grades as poor for him on the ground.

Our model projects half a yard and makes the under a 70% play, and every outside projection agrees: Sleeper 4.1, the expert consensus 4.1, ESPN 3.8 and another projection model 2.8. Wide receiver Caleb Douglas was ruled out, which could mean more snaps for Washington, but more of that work should come as a receiver. At -120 the break-even is 54.5%. The play: Washington under 5.5 rushing yards.

Chicago hosts the Jets as a 3.5-point favorite, but Caleb Williams was ruled out on the final injury report. That's the main risk for any Chicago receiving over.

Even so, every projection we track has Colston Loveland clearing 34.5: Sleeper 43.9, the expert consensus 40.8, ESPN 36.6 and another projection model 48.6. The matchup grades as good for tight ends, and short throws to the tight end are a common way to support a quarterback change. Our model calls it close to even (52% at 33.5, a yard better than the line we're getting), and at -120 we need 54.5%, so this is a consensus play rather than a model edge. The play: Loveland over 34.5 receiving yards.

Chicago hosts the Jets as a 3.5-point favorite without Caleb Williams, who was ruled out on the final injury report. With Williams out, short throws to the tight end are a natural way to move the ball, and Colston Loveland draws a matchup that grades as good for his position. Four catches beats the line.

Our model makes the over a 60% play at 3.5 and projects 3.6 receptions. The expert consensus (3.6), another projection model (4.2) and Sleeper (3.52) are over the line, and ESPN is just under at 3.3. At even money we need 50% to break even, so the model's 60% is a big cushion. The quarterback change is the risk to watch. The play: Loveland over 3.5 receptions at +100.

Detroit visits Carolina as a 3.5-point favorite in the highest-totaled game of Week 4 (52.5). High totals mean more plays, and Jameson Williams draws a matchup that grades as good for receivers. Four catches beats the line.

Sleeper (3.7), the expert consensus (3.6) and another projection model (3.9) all project him over 3.5, with ESPN just under at 3.3. Our model makes the over a 56% play, and at -105 we need 51.2%. A high total, a good matchup and three of four outside sources on our side is a clean over. The play: Williams over 3.5 receptions.

Denver visits San Francisco as a 3-point underdog in a 48.5-total game, and Jaylen Waddle draws a matchup that grades as good for receivers. He was a full participant on the final injury report. Five catches beats the line.

All four outside projections are over 4.5: Sleeper 4.7, the expert consensus 4.8, ESPN 4.8 and another projection model 5.4. Our model calls it a coin flip, so this one rests on the consensus and the script: a road underdog in a higher-total game should be throwing. At -110 we need 52.4%. The play: Waddle over 4.5 receptions.

Atlanta visits New Orleans as a 1.5-point underdog in a 48.5-total divisional game, and Kyle Pitts draws a matchup that grades as good for tight ends. His receiving line sits at 32.5.

The expert consensus (35.8), ESPN (35.1) and another projection model (38.6) all project him over the number, and Sleeper is just under at 30.9. Our model projects 31.5 yards and has the over as a coin flip even at 31.5, a yard better than the line we're getting, so this is purely a consensus play, at -115 (break-even 53.5%). A close game with a high total should keep both offenses throwing. The play: Pitts over 32.5 receiving yards.

San Francisco hosts Denver as a 3-point favorite with a 48.5 total, and Mike Evans draws a matchup that grades as great for receivers. He's listed as questionable with a rib injury after a limited final practice, so this is a bet to place once he's confirmed active.

If he plays, the projections sit well above the line: Sleeper 65.4, ESPN 57.2, the expert consensus 54.9 and another projection model 59.4, all 10 to 21 yards over 44.5. Our model is more conservative (43.5 projected yards) and calls it close to even at 43.5, a yard better than our line, so the edge comes from the outside consensus and the matchup. At -115 we need 53.5%. The play: Evans over 44.5 receiving yards, once he's confirmed active.

Mike Evans draws a matchup that grades as great for receivers, with San Francisco a 3-point home favorite over Denver. He carried a questionable tag with a rib injury after a limited final practice, so confirm he's active before betting.

Four catches beats the line, and every outside projection is over it: Sleeper 4.2, the expert consensus 3.9, ESPN 3.9 and another projection model 4.7. Our model makes the over a 55% play. At -120 (break-even 54.5%) the model alone is close to the price, so the full agreement of four outside sources is what makes this a bet. The play: Evans over 3.5 receptions, once he's confirmed active.

Atlanta visits New Orleans as a 1.5-point underdog in a 48.5-total game, and Zachariah Branch's receiving line is tiny at 5.5. One catch can clear it, and the matchup grades as good for him.

Our model projects 16.5 yards and gives the over a 64% chance even at 10.5, so the 5.5 line we're getting is five yards of extra cushion. Every outside projection agrees by a wide margin: Sleeper 11.3, ESPN 12.4, another projection model 14.3 and the expert consensus 15.3. At -120 we need 54.5%, and a line this far below every projection is as good as it gets. The play: Branch over 5.5 receiving yards.

Denver visits San Francisco as a 3-point underdog with a 48.5 total. A road underdog in a higher-scoring game is more likely to be throwing than running its quarterback, and Bo Nix's rushing line sits at 4.5.

Our model projects 3.6 attempts and makes the under a 58% play. Sleeper (4.3), the expert consensus (4.4) and another projection model (4.1) are all under the line; ESPN is just over at 4.8. At -125 the break-even is 55.6%, so the edge is modest, but three of four outside sources back the model. The play: Nix under 4.5 rush attempts.

Carolina hosts Detroit as a 3.5-point underdog in the highest-totaled game of Week 4 (52.5). A shootout script means the Panthers should be throwing, and Bryce Young's rushing line is a low 2.5. He was a full participant on the final injury report.

Our model projects just 1.6 attempts and makes the under a 61% play. The outside projections disagree, ranging from 3.1 (ESPN) to 4.3 (another projection model), with Sleeper at 4.1 and the expert consensus at 3.4. That's a real split, and we're siding with the model on a quarterback the line already expects to stay in the pocket. At -105 we need 51.2%. The play: Young under 2.5 rush attempts.

New England visits Buffalo as a 7-point underdog with a 49.5 total. A road underdog chasing points is usually throwing, which works against Drake Maye's rushing volume, and his line sits at 5.5. He was a full participant on the final injury report.

Our model projects 4.5 attempts and makes the under a 62% play. The outside sources are tight to the number: another projection model has 5.4, the expert consensus sits exactly on 5.5, and Sleeper (5.8) and ESPN (6.0) are slightly over. At -125 we need 55.6%; the model clears it, and the trailing script supports it. The play: Maye under 5.5 rush attempts.

Philadelphia hosts the Rams as a 4.5-point underdog with a 42.5 total. An underdog is less likely to be kneeling out the clock, and a team playing from behind leans on the pass, both of which work against Jalen Hurts's rushing volume. His line sits at 6.5.

Every source we track has him under it: Sleeper 6.3, the expert consensus 6.1, ESPN 6.0 and another projection model 5.8, with our model at 5.5 and a 62% read on the under. The price is steep at -130 (break-even 56.5%), but full agreement and a model read five points above break-even make it playable. The play: Hurts under 6.5 rush attempts.

The Chargers visit Seattle as 7-point underdogs with a 43.5 total. A road underdog likely to be throwing to catch up isn't calling a lot of quarterback runs, and Justin Herbert's rushing line sits at 4.5.

Our model projects 3.6 attempts and makes the under a 57% play. Sleeper (4.0), the expert consensus (4.2), another projection model (4.25) and ESPN (4.46) are all under the line. At -115 we need 53.5%, so the edge is modest but backed by all four outside sources. The play: Herbert under 4.5 rush attempts.

Baltimore hosts Tennessee as an 11.5-point favorite with a 41.5 total. A big favorite can grind the clock with its running backs, and our model expects Lamar Jackson to run less than the market does: 4.6 attempts, with a 54% read on the under at 5.5. He was a full participant on the final injury report.

The outside projections are well above the line (Sleeper 9.2, another projection model 8.5, the expert consensus 7.1, ESPN 6.9), so this is a contrarian, model-only play, and late kneel-downs as a big favorite are a real risk. What makes it a bet is the price: at +105 we need just 48.8% to break even. The play: Jackson under 5.5 rush attempts at +105.

Miami visits Minnesota as a 10-point underdog in a game totaled at 38.5. A big underdog is usually throwing to catch up, and Malik Willis's rushing line of 6.5 asks for seven attempts.

Our model projects 5.5 and makes the under a 63% play. Sleeper (5.2), ESPN (5.4) and the expert consensus (6.3) are all under the line, and another projection model is just over at 6.55. At -120 the break-even is 54.5%, and the model clears it by eight points with three of four outside sources in agreement. The play: Willis under 6.5 rush attempts.

Jacksonville visits Cincinnati as a 2.5-point underdog in a 51.5-total game, one of the highest of Week 4. A high-scoring, close game should keep Trevor Lawrence throwing, and his rushing line sits at 3.5.

Our model projects 2.7 attempts, but its read on the under is only 51%, and the outside projections lean the other way (Sleeper 4.3, ESPN 4.0, another projection model 4.0, the expert consensus 3.9). This is one of the thinnest plays on the card: the case is a model projection nearly a full attempt below the line and a throwing script, at a price (-115, break-even 53.5%) that asks the model to be right about the volume. It's a lean, not a strong play. The play: Lawrence under 3.5 rush attempts.

New Orleans hosts Las Vegas as a 3.5-point favorite, and the Saints have depth at running back: Travis Etienne and Kendre Miller were both full participants on the final injury report. Shared work and a favored script both cut into Alvin Kamara's receiving role, and this line asks for three catches.

Our model makes the under a 74% play, one of the strongest reads of Week 3, and projects 1.5 receptions. Every outside projection agrees: Sleeper 2.1, the expert consensus 2.1, ESPN 2.0 and another projection model 2.2. At -110 the break-even is 52.4%. The play: Kamara under 2.5 receptions.

Las Vegas visits New Orleans as a 3.5-point underdog with a 44.5 total, and the matchup grades as poor for running backs in the passing game. Ashton Jeanty's receiving line sits at 22.5.

Our model projects 17.5 yards and makes the under a 55% play at 20.5, so the 22.5 we're getting is two yards of extra cushion. Another projection model (17.6) and Sleeper (22.1) are under the line, while the expert consensus (23.6) is just over and ESPN (34.2) well over. With a poor matchup and two extra yards, we'll take it at -115. The play: Jeanty under 22.5 receiving yards.

Kansas City visits Miami as a 10-point favorite, and a big favorite with a lead tends to keep its backs running rather than catching. Emmett Johnson's receiving line sits at 10.5.

Our model projects just 2.5 receiving yards and makes the under a 72% play at 11.5. Sleeper (10.3) and the expert consensus (9.3) are under the number we're getting; ESPN (11.5) and another projection model (16.2) are over. The matchup grades as great for him and the outside sources split, but the model's projection sits eight yards below the line. At -110 we need 52.4%. The play: Johnson under 10.5 receiving yards.

Kansas City is a 10-point road favorite in Miami, and Emmett Johnson's receiving role projects as small: two catches beats us, and the projections expect about one.

Our model projects 0.7 receptions and makes the under a 67% play. Sleeper (1.1), the expert consensus (1.0) and ESPN (1.4) are all under 1.5, with another projection model just over at 1.6. The price is -125 (break-even 55.6%), but a big-favorite script and three of four outside sources under the number give the model's read plenty of support. The play: Johnson under 1.5 receptions.

Minnesota visits Tampa Bay as a 1.5-point underdog with a 42.5 total. Josh Oliver's receiving line is small at 11.5, and our model sees an even smaller role: 4.5 projected yards and a 67% read on the under.

Sleeper (10.5) and another projection model (11.0) are under the number as well; ESPN is over at 13.0. A tight end projected for a limited receiving role in a low-total game is our kind of under, and at -120 (break-even 54.5%) the model clears the price comfortably. The play: Oliver under 11.5 receiving yards.

Washington hosts Seattle as an 8.5-point underdog, and Jayden Daniels was ruled out on the final injury report, which adds uncertainty to Washington's passing game. Rachaad White's receiving line sits at 15.5, and fellow back Kaytron Allen was a full participant, so the backfield work should be shared.

Our model projects 8.5 receiving yards and makes the under a 66% play at 16.5; we're getting 15.5 at -110. Sleeper (13.7), the expert consensus (12.5) and another projection model (9.8) all sit under the number; ESPN is the outlier at 20.0. A trailing script is the risk, but with Daniels out and three of four outside sources under, we like it. The play: White under 15.5 receiving yards.

Kansas City is a 10-point favorite in Miami, and Emmett Johnson's rushing line of 26.5 sits well above what most projections expect from him. The matchup grades as poor for him on the ground.

Our model projects 14.5 rushing yards and makes the under a 74% play, one of the strongest reads of Week 3. Sleeper (15.3), ESPN (20.7) and the expert consensus (21.7) all agree; another projection model is slightly over at 27.9. A big favorite usually spreads carries around late, and the model's projection is 12 yards below the line. At -110 the break-even is 52.4%. The play: Johnson under 26.5 rushing yards.

Cleveland hosts Carolina as a 2.5-point underdog with a 41.5 total, and KC Concepcion's rushing line sits at 4.5. Wide receivers get the occasional designed carry, but our model projects almost nothing on the ground: 0.4 yards, with a 77% read on the under at 5.5, one of the strongest of Week 3.

We're getting 4.5 at -110, a yard lower than the model's number. Sleeper (3.3) and another projection model (1.9) are under it; the expert consensus (5.0) and ESPN (8.5) are over. The matchup grades as poor for him as a runner. A single carry can beat a line this small, which is the risk, but the model's near-zero projection is a strong signal. The play: Concepcion under 4.5 rushing yards.

Green Bay hosts Atlanta as a 4.5-point favorite with a 43.5 total, and the matchup grades as poor for Kaleb Johnson on the ground. His rushing line sits at 35.5.

Our model makes the under a 56% play at 38.5, and we're getting 35.5 at -115, three yards lower, so the cushion is thinner than it looks. The outside consensus does the heavy lifting: Sleeper (22.0), another projection model (25.4) and the expert consensus (30.8) all sit well under 35.5, and only ESPN (40.6) is over. Three sources averaging about 26 yards against a 35.5 line is a strong signal. The play: Johnson under 35.5 rushing yards.

Arizona visits San Francisco as a 7.5-point underdog, and a team likely to be trailing tends to lean away from the run. Tyler Allgeier's rushing line sits at 33.5.

Our model projects 20.5 rushing yards and makes the under a 71% play at 34.5. The expert consensus (32.1), ESPN (32.1) and another projection model (31.6) are all under the 33.5 we're getting, with Sleeper just over at 34.3. At -120 the break-even is 54.5%, and the model's 13-yard gap to the line is a big cushion. The play: Allgeier under 33.5 rushing yards.

Seattle visits Washington as an 8.5-point favorite with a 40.5 total, and AJ Barner draws a matchup that grades as good for tight ends. Quarterback Sam Darnold was a full participant on the final injury report. Three catches is the target.

Sleeper (2.9), the expert consensus (2.8) and another projection model (2.6) all project him over 2.5, with ESPN just under at 2.48. Our model calls it a coin flip, so this is an agreement play: a good matchup and three of the four outside sources over a short line. At -110 we need 52.4%. The play: Barner over 2.5 receptions.

Las Vegas visits New Orleans as a 3.5-point underdog, and Brock Bowers draws a matchup that grades as great for tight ends. The caution flag is health: Bowers did not practice on the final injury report and was listed as questionable with a knee injury, and the Raiders also listed quarterback Aidan O'Connell as questionable. This is a bet to make only once Bowers is confirmed active.

If he plays, the numbers are strong. Every outside projection has him over 4.5: Sleeper 5.5, the expert consensus 5.2, ESPN 6.0 and another projection model 5.2. Our model makes the over a 59% play. At -125 the break-even is 55.6%, and a great matchup for a team likely to be throwing from behind adds to the case. The play: Bowers over 4.5 receptions, once he's confirmed active.

Miami hosts Kansas City as a 10-point underdog, and a team that far behind on the spread usually throws a lot in the second half. Greg Dulcich draws a matchup that grades as good for tight ends and needs three catches.

All four outside projections have him there: Sleeper 2.8, the expert consensus 3.0, ESPN 2.8 and another projection model 3.2. Our model makes the over a 54% play, and at +120 the break-even is just 45.5%, a nine-point cushion. Fellow tight end Seydou Traore was a full participant on the final injury report, so the position's snaps may be shared, which is the main risk. The play: Dulcich over 2.5 receptions at +120.

Cleveland hosts Carolina as a 2.5-point underdog with a 41.5 total, and Harold Fannin Jr. draws a matchup that grades as good for tight ends. Four catches beats the line.

Every outside projection clears it: Sleeper 3.8, the expert consensus 4.1, ESPN 4.1 and another projection model 4.8. Our model makes the over a 57% play. The price is at the top of our range at -130 (break-even 56.5%), so the edge is narrow on the model alone, but full agreement from four outside sources is what pushes it over. The play: Fannin over 3.5 receptions.

Atlanta visits Green Bay as a 4.5-point underdog, and both quarterbacks, Michael Penix Jr. and Tua Tagovailoa, were full participants on the final injury report. Kyle Pitts draws a matchup that grades as good for tight ends and needs four catches.

The expert consensus (3.8), ESPN (4.1), another projection model (4.0) and Sleeper (3.54) all project him over 3.5. Our model makes the over a 54% play. At -120 we need 54.5%, so the edge comes from the outside consensus and the script: a road underdog should be throwing. The play: Pitts over 3.5 receptions.

Kansas City visits Miami as a 10-point favorite, and Rashee Rice draws a matchup that grades as great for receivers. Rice and Patrick Mahomes were both full participants on the final injury report.

Five catches beats the line. The expert consensus (5.1) and another projection model (5.3) project him over 4.5, while ESPN (4.46) and Sleeper (4.2) are just under. Our model makes the over a 56% play, and the price is the bonus: at +115 the break-even is 46.5%, almost ten points below the model. The play: Rice over 4.5 receptions at +115.

Tennessee visits the Giants as a 2.5-point underdog in the lowest-totaled game of Week 3 (37.5). Low totals usually mean fewer plays, but Wan'Dale Robinson's line is a short 3.5 catches and the matchup grades as good for him.

All four outside projections have him over the number: Sleeper 3.8, the expert consensus 3.9, ESPN 3.8 and another projection model 4.0. Our model calls it close to even at 52%. At -105 we need 51.2%, so this is an agreement play on a short line rather than a big model edge. The play: Robinson over 3.5 receptions.

Denver hosts the Rams as a 1.5-point favorite with a 44.5 total, a close game where we don't expect a long kneel-down sequence or a lot of designed quarterback runs. Bo Nix's rushing line sits at 4.5.

Our model projects 3.6 attempts and makes the under a 58% play. The expert consensus (4.4) and another projection model (4.2) are under the line; Sleeper (4.6) and ESPN (4.8) are just over. At -110 we need 52.4%, and the model's projection sits nearly a full attempt below the number. The play: Nix under 4.5 rush attempts.

Philadelphia visits Chicago as a 3-point favorite with a 42.5 total. Jalen Hurts's rushing line sits at 6.5, and every source we track puts him at or under it.

Our model projects 5.6 attempts and makes the under a 60% play. The expert consensus (6.4), ESPN (6.3) and another projection model (5.8) are all under 6.5, and Sleeper sits exactly on it. At -125 the break-even is 55.6%, so the edge is moderate, but the agreement is complete. The play: Hurts under 6.5 rush attempts.

New Orleans visits Baltimore as an 8.5-point underdog, a script that usually adds targets for running backs, and the matchup grades as good for Alvin Kamara. Those are real risks for an under, so the case has to be in the numbers, and it is.

Our model projects 4.5 receiving yards and makes the under a 68% play at 12.5. Every outside projection we track sits below the line: Sleeper 9.4, the expert consensus 10.6, ESPN 10.2 and another projection model 8.8. Kamara was a full participant on the final injury report, so his role isn't in question; the projections just don't see a 13-yard receiving day. At -115 the break-even is 53.5%. The play: Kamara under 12.5 receiving yards.

Las Vegas visits the Chargers as a 6.5-point underdog in a 43.5-total game. A trailing script can push checkdowns to running backs, but the matchup grades as poor for backs in the passing game, and Ashton Jeanty's receiving line at 26.5 is set above where most projections land. He was a full participant on the final injury report.

Our model projects 20.5 receiving yards and makes the under a 58% play at 25.5; we're getting 26.5 at -115, a yard of extra cushion. Sleeper (17.7), the expert consensus (24.7) and another projection model (20.9) all sit under the number, with ESPN the outlier at 34.5. Three of four outside sources plus the model, in a poor matchup, is our kind of under. The play: Jeanty under 26.5 receiving yards.

Atlanta hosts Carolina as a 2.5-point underdog with a 43.5 total, and the quarterback room is unsettled: Michael Penix Jr. was listed out and Tua Tagovailoa doubtful on the final injury report. Bijan Robinson's receiving line is big at 42.5, and it asks a lot of a passing game with that much uncertainty at quarterback.

Our model projects 29.5 receiving yards and makes the under a 68% play at 43.5. Sleeper (30.4) and the expert consensus (42.2) sit under the 42.5 we're getting; ESPN (49.5) and another projection model (43.4) are over, so the outside sources split. The matchup grades as good for Robinson, which is the risk, but a 13-yard gap between the model and the line is too big to pass up at -115. The play: Robinson under 42.5 receiving yards.

The Chargers host Las Vegas as 6.5-point favorites with a 43.5 total. A team protecting a lead leans on the run, and David Njoku's receiving line of 23.5 is set above where most projections land.

Our model projects 13.5 yards and makes the under a 67% play. Sleeper (14.8), another projection model (21.8) and the expert consensus (22.3) all sit under the number; ESPN is the lone source over at 27.8. At -115 (break-even 53.5%) the model's edge is large, and three of four outside sources agree. The play: Njoku under 23.5 receiving yards.

David Njoku needs three catches to beat this under, and in a game where the Chargers are 6.5-point home favorites over Las Vegas, the projections have him closer to two. A favored team running the ball late is the script we want for a tight end under.

Our model projects 1.6 receptions and makes the under a 67% play. Sleeper (1.6), the expert consensus (2.3), another projection model (2.4) and ESPN (2.47) are all under 2.5. At even money we only need 50% to break even, which makes this one of the best prices on the board for a read this strong. The play: Njoku under 2.5 receptions at +100.

The Rams host the Giants as 7-point favorites, and the matchup grades as poor for Kyren Williams in the passing game. A big home favorite tends to feed its back on the ground rather than through checkdowns, and his receiving line sits at 16.5.

Our model projects 12.5 yards and makes the under a 56% play at 15.5; we're getting 16.5 at -105, a yard better than the model's number. Sleeper (14.5), the expert consensus (15.7) and ESPN (15.8) all sit under the line, with another projection model just over at 17.7. At -105 the break-even is 51.2%. The play: Williams under 16.5 receiving yards.

The Jets host Green Bay as 3.5-point underdogs, and the matchup grades as poor for Mason Taylor. Three catches beats us, and every projection we track has him under that.

Our model projects 1.6 receptions and makes the under a 64% play. Sleeper (2.2), the expert consensus (2.4), ESPN (2.1) and another projection model (1.9) are all under 2.5. At -115 the break-even is 53.5%, and the model clears it by more than ten points with full agreement from the outside sources. The play: Taylor under 2.5 receptions.

Pittsburgh visits New England as a 4.5-point underdog with a 41.5 total. The matchup grades as poor for Rico Dowdle as a receiver, and his receptions line asks for three catches.

Our model projects 1.6 receptions and makes the under a 62% play at 2.5. Every outside source agrees: Sleeper 2.4, the expert consensus 2.0, ESPN 2.0 and another projection model 2.2. At -105 we need just 51.2% to break even. A trailing script is the one thing that could push extra checkdowns his way, but four projections and the model all landing under is enough. The play: Dowdle under 2.5 receptions.

Tennessee hosts Philadelphia as a 7-point underdog in a game totaled at just 39.5. Trailing scripts are the risk for a running-back receiving under, but in a low-total game the volume isn't expected to be there, and most projections have Tyjae Spears well short of 15.5.

Our model projects 7.5 receiving yards and makes the under a 67% play. Sleeper (11.8), the expert consensus (13.9) and another projection model (9.8) are all under the line; ESPN is the lone source over at 20.1. At -110 the break-even is 52.4%. The play: Spears under 15.5 receiving yards.

Tennessee is a 7-point home underdog to Philadelphia in a 39.5-total game, and Tyjae Spears needs three catches to beat this under. Our model and most of the outside projections have him short of that.

The model projects 1.6 receptions and makes the under a 66% play. Sleeper (2.3), the expert consensus (2.2) and another projection model (1.7) sit under 2.5, with ESPN above at 2.8. The price is -120 (break-even 54.5%), and the model clears it by more than ten points. The play: Spears under 2.5 receptions.

San Francisco hosts Miami as a 12.5-point favorite, the biggest spread of Week 2. A big lead should mean a run-heavy second half, but those carries usually go to the backs, not to Deebo Samuel, whose rushing line sits at 6.5. Wide receiver De'Zhaun Stribling was ruled out with an ankle injury, which could add to Samuel's role, though more of that should come through the air.

Our model projects just half a yard rushing and makes the under a 67% play at 5.5; we're getting 6.5 at -120, a yard better. Sleeper (5.0) and another projection model (4.6) are under, while the expert consensus (6.6) is just over and ESPN (7.9) is over. On a number this small, the model's near-zero projection is the strongest signal we have. The play: Samuel under 6.5 rushing yards.

Cleveland visits Tampa Bay as an 8.5-point underdog, and a trailing team doesn't hand the ball to its wide receivers. KC Concepcion's rushing line sits at 7.5, and the matchup grades as poor for him on the ground.

This is the strongest read on the Week 2 board: our model projects half a yard and makes the under a 77% play. Sleeper (3.2), the expert consensus (3.1) and another projection model (1.6) are all well under the line; only ESPN (7.8) sits above it. At -115 the break-even is 53.5%. The play: Concepcion under 7.5 rushing yards.

Philadelphia visits Tennessee as a 7-point favorite in a 39.5-total game. A favorite with a lead runs the ball, which is why Saquon Barkley's line is set as high as 80.5, but in a low-scoring game we think that number asks for too much.

Our model projects 64.5 rushing yards and makes the under a 65% play. Another projection model is even lower at 61.8, and Sleeper (78.1) also sits under the line; the expert consensus is just over at 80.9 and ESPN (88.7) is clearly over. The matchup grades as good for Barkley, which is the risk, but a 16-yard gap between the model and the number is a lot of cushion at -115. The play: Barkley under 80.5 rushing yards.

Arizona hosts Seattle as a 3.5-point underdog in a 40.5-total game, and a team likely to be playing from behind won't lean heavily on its run game. Tyler Allgeier's rushing line sits at 35.5.

Our model projects 22.5 rushing yards and makes the under a 70% play at 36.5. We're getting 35.5 at -115, a yard lower, and every outside projection still sits under it: Sleeper 28.0, the expert consensus 31.8, ESPN 33.9 and another projection model 34.8. At -115 the break-even is 53.5%. The play: Allgeier under 35.5 rushing yards.

San Francisco hosts Miami as a 12.5-point favorite, and Deebo Samuel draws a matchup that grades as great for receivers. Fellow wideout De'Zhaun Stribling was ruled out with an ankle injury, which concentrates targets among the remaining receivers.

Four catches beats the line. The expert consensus (3.9) and another projection model (4.2) project him over 3.5, while ESPN (3.45) and Sleeper (3.4) are just under; our model calls it a coin flip at 51%. The case is the price and the setup: at +110 the break-even is 47.6%, and a great matchup with a receiver out of the lineup tilts this our way. The play: Samuel over 3.5 receptions at +110.

Atlanta hosts Carolina as a 2.5-point underdog, and Drake London draws a matchup that grades as good for receivers. The catch is the quarterback: Michael Penix Jr. was listed out and Tua Tagovailoa doubtful on the final injury report, so the quarterback situation is uncertain.

Even so, three of the four outside projections put him over the number: the expert consensus 4.9, ESPN 5.0 and another projection model 4.9, with Sleeper lower at 4.0. Our model projects 4.6 catches and makes the over a 56% play at 4.5. At -125 (break-even 55.6%) this is a thin edge built on volume, so the quarterback news is the factor to watch. The play: London over 4.5 receptions.

Dallas hosts Washington as a 4.5-point favorite with one of the highest totals of Week 2 (51.5). High totals mean more plays and more targets, and the matchup grades as good for Jake Ferguson. Four catches is the target.

All four outside projections clear it: Sleeper 3.9, the expert consensus 4.0, ESPN 3.9 and another projection model 4.2. Our model sits at 3.5 and calls the over a 53% play, and at -105 we need 51.2%. A high total, a good matchup and full agreement among the outside sources is enough. The play: Ferguson over 3.5 receptions.

Denver hosts Jacksonville as a 2.5-point favorite, and Jaylen Waddle draws a matchup that grades as great for receivers. Wide receiver Marvin Mims Jr. was ruled out with a foot injury, which leaves more of the target share for Waddle and the rest of the room.

The projections cluster right around the line: Sleeper (4.56) and another projection model (5.5) are over 4.5, the expert consensus sits exactly on it, ESPN is a hair under at 4.49, and our model calls it a coin flip. That's why the price matters: at +105 we need 48.8% to break even, and a great matchup with a receiver out tilts a coin flip in our favor. The play: Waddle over 4.5 receptions at +105.

Atlanta hosts Carolina as a 2.5-point underdog, and Kyle Pitts draws a matchup that grades as good for tight ends. Quarterback is the uncertainty: Michael Penix Jr. was listed out and Tua Tagovailoa doubtful on the final injury report.

Four catches beats the line, and all four outside projections expect him to get there: Sleeper 3.6, the expert consensus 4.0, ESPN 4.3 and another projection model 4.0. Our model sits right on 3.5 and calls it close to even (52%), but at +105 we need only 48.8% to break even. A home underdog likely to throw, a good matchup and plus money make this a value over despite the quarterback questions. The play: Pitts over 3.5 receptions at +105.

Kansas City hosts Indianapolis as a 6-point favorite with a 46.5 total, and Rashee Rice draws a matchup that grades as good for receivers. Both Rice and Patrick Mahomes were full participants on the final injury report.

Five catches is the target. Sleeper (5.3), the expert consensus (5.6) and ESPN (5.4) all project him over 4.5; another projection model is slightly under at 4.4. Our model makes the over a 55% play, which by itself falls just short of the -125 break-even (55.6%). The three outside projections at 5.3 to 5.6 are what carry this one. The play: Rice over 4.5 receptions.

Denver hosts Jacksonville as a 2.5-point favorite with a 45.5 total. Bo Nix's rushing line sits at 4.5, and our model expects fewer: 3.6 attempts, with a 57% read on the under.

The outside sources are tighter to the number. Another projection model (4.2) and Sleeper (4.46) are under the line, while the expert consensus (4.7) and ESPN (5.0) are slightly over. Those are fractions of an attempt either way, so this is a model-driven play at -110 (break-even 52.4%). In a close game there's less reason for a long kneel-down sequence at the end. The play: Nix under 4.5 rush attempts.

San Francisco hosts Miami as a 12.5-point favorite. Big favorites can pick up kneel-downs late, and those count as rush attempts, so this under needs Brock Purdy to be a pocket passer for most of the game, which is what our model expects: 2.6 attempts and a 57% read on the under at 3.5.

The outside projections are higher, between 3.6 and 4.4 (Sleeper 4.2, the expert consensus 4.4, ESPN 4.1, another projection model 3.6), so this is a model-versus-consensus spot. The price is what makes it: at +110 the break-even is just 47.6%, and the model sits nine points above that. The play: Purdy under 3.5 rush attempts at +110.

Cleveland visits Tampa Bay as an 8.5-point underdog with a 41.5 total. A team chasing points throws, and Deshaun Watson's rushing line of 5.5 is a full attempt above where our model sets it.

The model projects 3.5 attempts and makes the under a 64% play at 4.5, so at 5.5 we're getting an extra attempt of cushion. Every outside source is below the number: Sleeper 3.1, the expert consensus 3.8, ESPN 4.5 and another projection model 2.4. The price is -130 (break-even 56.5%), but the line is the edge here. The play: Watson under 5.5 rush attempts.

The Giants visit the Rams as 7-point underdogs with a 47.5 total. A big road underdog is usually throwing to catch up, and Jaxson Dart's rushing line asks for seven attempts.

Our model projects 5.6 attempts and makes the under a 54% play. Another projection model (5.8) and ESPN (6.49) are under the line, and Sleeper is slightly over at 6.7. At even money we need 50%, so the model's read plus the trailing script gives us a small edge. The play: Dart under 6.5 rush attempts at +100.

Green Bay at Minnesota is a tight divisional opener, with the Vikings laying 2.5 at home and a 46.5 total. That isn't a script that forces Minnesota to throw to its backs all afternoon, and Aaron Jones's receiving yards have to come against a matchup that grades as poor for running backs catching passes. We'd rather bet against a big receiving day than for one.

Our model projects just 9.5 receiving yards and makes the under a 59% play at 14.5. Sleeper (13.3) and another projection model we track (12.0) also sit below the number, while the expert consensus (16.5) and ESPN (19.9) lean higher. It isn't unanimous, but the model's edge is clear: at -115 we need about 53.5% to break even. The play: Jones under 14.5 receiving yards.

Minnesota opens at home as a 2.5-point favorite over Green Bay, and the matchup grades as poor for running backs in the passing game. Aaron Jones needs three catches to beat this under, and nothing in the projections says a three-catch day is the expectation in a close, divisional game where the Vikings aren't expected to be chasing.

Our model has him at 1.7 receptions and makes the under a 56% play at 2.5, a thin but real edge at -110, where break-even is 52.4%. Sleeper projects 2.2 catches and another projection model we use has 1.9. The expert consensus sits right on the number at 2.5, and ESPN is the lone source above it at 2.8. Two outside projections agree with the model and the others sit within a fraction of a catch. The play: Jones under 2.5 receptions at -110.

Seattle opens at home as a 3-point favorite over New England with a modest 44.5 total, and George Holani projects for a small receiving role. The matchup grades as poor for running backs catching the ball, and Emanuel Wilson, the only other Seattle back on the final injury report, was a full participant.

This is one of the cleanest unders on the Week 1 board. Our model makes it a 70% play at 11.5 and projects Holani for only 2.5 receiving yards, and every outside projection we track agrees: Sleeper 8.3, the expert consensus 8.8, ESPN 8.4 and another projection model 8.2. We're getting 10.5 at -110, a yard below the model's number, and all four sources still sit under it. The play: Holani under 10.5 receiving yards.

Seattle hosts New England as a 3-point favorite, and George Holani's role in the passing game projects as minimal. The matchup grades as poor for running backs as receivers, and no projection we track has him above 1.3 catches. One catch or fewer cashes this, and we get plus money.

Our model projects 0.7 receptions and gives the under a 63% chance at 1.5. Sleeper and the expert consensus both have him at 1.3, ESPN at 1.1 and another projection model at 1.1, so all four outside sources are under the number. Getting +120 on a 63% event is the kind of price we look for: break-even is 45.5%, which leaves a wide cushion even if the model is a little optimistic. The play: Holani under 1.5 receptions at +120.

Buffalo opens on the road in Houston as a 1.5-point favorite in a game totaled at 44.5, and the matchup grades as poor for running backs as receivers. This line asks James Cook for steady checkdown work in a spot the matchup grades don't like. Fellow back Ty Johnson carried a questionable tag with a hamstring after a limited final practice, which could keep more of the backfield on Cook's plate, but most of that would show up on the ground.

Our model projects 12.5 receiving yards and makes the under a 63% play at 19.5. Three of the four outside projections agree: Sleeper 17.1, the expert consensus 18.0 and another projection model 19.1, with ESPN the lone source above at 20.9. At -115 the break-even is 53.5%, so there's real room. The play: Cook under 19.5 receiving yards.

Washington opens in Philadelphia as a 6-point underdog with a low 43.5 total. Trailing scripts can push targets to running backs, and that is the risk here, but our model still sees a modest receiving role for Rachaad White: 8.5 projected yards, with fellow back Jacory Croskey-Merritt (limited on the final injury report, no game status) also in the mix.

The model makes the under a 65% play at 15.5, and we're getting 14.5 at -110. The expert consensus (13.4) and another projection model (11.0) are below our number, while Sleeper (14.8) and ESPN (14.9) sit just above it, within half a yard. That's a split among the outside sources but a clear lean from the model, and receiving lines this short usually come down to one or two catches. The play: White under 14.5 receiving yards.

The Rams and 49ers open the season in Melbourne, with Los Angeles a 3.5-point favorite and a 47.5 total. Tyler Higbee was a full participant on the final injury report, so there's no health angle here. The case is simply that his receiving line is set above what every projection we track expects.

Our model projects 8.5 yards and makes the under a 65% play at 16.5. Sleeper has him at 9.2, ESPN at 11.5, the expert consensus at 14.9 and another projection model at 15.9, all below the number. An international neutral-site opener adds some uncertainty to everyone's usage, but when four independent projections and the model all land under, we'll take -115. The play: Higbee under 16.5 receiving yards.

Houston hosts Buffalo as a 1.5-point underdog in a game totaled at 44.5, and the matchup grades as poor for running backs in the passing game. Woody Marks needs two catches to beat this, and the outside projections all sit between 1.3 and 1.5.

Our model projects 0.7 receptions and makes the under a 63% play. Sleeper (1.41), ESPN (1.47) and another projection model (1.3) are all under the line, and the expert consensus sits exactly on 1.5. None of the outside sources project two catches. At -120 we need 54.5% to break even, and the model clears that by almost nine points. The play: Marks under 1.5 receptions.

San Francisco is a 3.5-point underdog to the Rams in Melbourne, and Deebo Samuel's rushing line is the kind of small number we like to fade. It only takes a handful of designed touches to clear, but the expectation is far lower, and the matchup grades as poor for him on the ground.

Our model projects 3.5 rushing yards and makes the under a 70% play at 9.5. We're getting 7.5 at -120, two yards lower than the model's number, but every outside projection still sits underneath it: Sleeper 4.1, the expert consensus 5.4, ESPN 5.6 and another projection model 6.1. On a line this small one carry can swing it, which is why we want every source on our side before we bet it. They are. The play: Samuel under 7.5 rushing yards.

Indianapolis hosts Baltimore as a 3-point underdog in a 48.5-total game, the kind of script that keeps the ball in the air, and the matchup grades as good for Alec Pierce. He was limited on the final injury report but carried no game designation into the weekend.

Our model gives the over a 57% chance at 29.5, and the outside numbers are far more bullish than the line: Sleeper projects 51.7 yards, ESPN 49.5 and another projection model 51.0. When three independent projections sit roughly 20 yards above the number and the matchup is favorable, we'll lay -115 on the over. The practice tag is the only caution flag. The play: Pierce over 29.5 receiving yards.

Alec Pierce needs three catches against a Baltimore defense that grades as a good matchup for him, in a game where the Colts are 3-point home underdogs and should be throwing. He was limited on the final injury report but carried no game status.

Our model makes the over a 57% play at 2.5 and projects 2.6 catches. Sleeper (2.7), ESPN (3.1) and another projection model (2.9) all sit above the line. At +110 the break-even is 47.6%, so a 57% read gives us real value at plus money. The play: Pierce over 2.5 receptions at +110.

George Kittle opens the season in Melbourne with the 49ers as 3.5-point underdogs to the Rams, a script that should keep San Francisco throwing, and the matchup grades as good for the tight end. He was limited on the final injury report but had no game status.

This is a price play as much as a projection play. Our model has the over as a 53% shot at 3.5, and at +105 we only need 48.8% to break even. Sleeper (3.8), the expert consensus (3.7) and ESPN (3.53) all project him over 3.5, while another projection model is lower at 3.1. A good matchup, a trailing script and plus money on a better-than-even event is a bet we make. The play: Kittle over 3.5 receptions at +105.

Tennessee hosts the Jets as a 1.5-point favorite in the lowest-totaled game of Week 1 (39.5). Low totals don't have to mean low volume for a tight end, though, and the matchup grades as good for Gunnar Helm. He needs three catches.

All four outside projections are over 2.5: Sleeper 3.0, the expert consensus 2.9, ESPN 3.3 and another projection model just over at 2.52. Our model calls it close to a coin flip at 52%, which is exactly why the price matters: at +110 the break-even is 47.6%, so we're being paid more than the true odds on an event the projections lean toward. The play: Helm over 2.5 receptions at +110.

Las Vegas opens at home as a 3-point favorite over Miami with a low 40.5 total. Jack Bech's line is modest, the matchup grades as good for him, and every projection we track has him clearing it.

Our model projects 19.5 yards and gives the over a 55% chance at 17.5. We're getting 18.5 at -115, a yard higher, but the outside numbers sit well above it: Sleeper 25.0, ESPN 27.9 and another projection model 25.2. When the independent projections average about 26 yards against an 18.5 line, a couple of catches gets us home. The play: Bech over 18.5 receiving yards.

Seattle is a 3-point home favorite over New England, and Jaxon Smith-Njigba draws a matchup that grades as great for receivers. Fellow wideout Tory Horton was questionable with a hamstring after a limited final practice, which could funnel more targets to the remaining receivers.

The line asks for seven catches, a big number, but Sleeper (6.8), the expert consensus (6.9) and ESPN (6.8) all project him above it, with another projection model lower at 5.8. Our model sees it as close to even (52%), so this is a matchup-and-volume play rather than a model-edge play. A great matchup plus three of four projections over the number is enough at -110. The play: Smith-Njigba over 6.5 receptions.

Houston is a 1.5-point home underdog to Buffalo, and Jaylin Noel only needs two catches against a matchup that grades as good for him. Small receiving lines like this are about having a role, not a big one, and every projection we track gives him enough of one.

Our model makes the over a 58% play at 1.5. Sleeper projects 1.7 catches, the expert consensus 2.2, ESPN 2.9 and another projection model 2.2, so all four outside sources clear the number. At -110 the break-even is 52.4%. The play: Noel over 1.5 receptions.

Kayshon Boutte's receiving line sits at 21.5, and the outside projections are nearly double it: Sleeper 43.7, the expert consensus 38.2, ESPN 43.8 and another projection model 40.2. That's a 17-to-22-yard cushion across four independent sources, in a game where Houston is a 1.5-point home underdog to Buffalo and the matchup grades as good for him.

Our model is more conservative at 25.5 projected yards but still makes the over a 58% play at 20.5; we're getting 21.5 at -110. When the model and every outside projection land on the same side and the line sits this far below the consensus, it's an easy over. The play: Boutte over 21.5 receiving yards.

The Jets open in Tennessee as 1.5-point underdogs in the lowest-totaled game of Week 1 (39.5), but Kenyon Sadiq's receiving line is low enough that a couple of catches clears it. He was a full participant on the final injury report, and the matchup grades as good for the tight end.

Our model makes the over a 63% play at 12.5 and projects 18.5 yards; we're getting 14.5 at -115. Every outside projection agrees: Sleeper 18.0, ESPN 21.4, the expert consensus 24.4 and another projection model all the way up at 35.2. Even the most conservative source clears the number by three and a half yards. The play: Sadiq over 14.5 receiving yards.

Two catches is all Kenyon Sadiq needs, and every projection we track has him above the 1.5 line. The Jets are 1.5-point underdogs in Tennessee, he was a full participant on the final injury report, and the matchup grades as good for him.

Our model gives the over a 70% chance at 1.5, one of the strongest reads on the Week 1 board. Sleeper projects 1.8 catches, ESPN 2.1, the expert consensus 2.3 and another projection model 3.0, so all four outside sources are over the number. At -120 we need 54.5% to break even, and the model clears it by more than 15 points. The play: Sadiq over 1.5 receptions.

Arizona opens in Los Angeles as an 8.5-point underdog, the kind of script that usually means plenty of passing in the second half, and the matchup grades as good for Michael Wilson. Four catches is the target.

Sleeper (3.7), the expert consensus (3.8) and ESPN (4.2) all project him over 3.5, with another projection model at 3.2, and our model makes the over a 56% play with a 3.6-catch projection. The price is the drawback: at -125 we need 55.6% to break even, so this is a thin edge that leans on the game script. Big underdogs throw, and we'll take the over. The play: Wilson over 3.5 receptions at -125.

Carolina hosts Chicago as a 3-point underdog in a 47.5-total game, and Tetairoa McMillan draws a matchup that grades as good for receivers. Five catches is a real number, but it's right where the projections land.

All four outside sources project him over 4.5: Sleeper 4.7, the expert consensus 4.7, ESPN 5.0 and another projection model 4.7. Our model sits right at 4.5 and makes the over a 55% play. The edge is modest at -115 (break-even 53.5%), so the case rests on agreement: a good matchup, a home underdog likely to throw, and every independent projection on the same side. The play: McMillan over 4.5 receptions.

Indianapolis is a 3-point home underdog to Baltimore in a 48.5-total game. Teams chasing points throw, and a quarterback passing to catch up isn't usually adding designed runs. Daniel Jones needs four rushing attempts to beat this, kneel-downs included, and a trailing script also lowers the odds of a late kneel sequence.

Our model projects 2.5 attempts and makes the under a 68% play at 3.5. Sleeper (3.2) and the expert consensus (3.1) agree; ESPN (4.3) and another projection model (3.7) are higher, so the outside sources split. The model's read is strong enough to carry it at -120, where break-even is 54.5%. The play: Jones under 3.5 rush attempts.

Philadelphia opens at home as a 6-point favorite over Washington with a 43.5 total. Jalen Hurts's rushing volume is the question, and our model is lower on it than the market: it projects 5.6 attempts and makes the under a 69% play at 7.5.

We're getting 6.5 at even money, a full attempt lower than the model's number, so the cushion is thinner, and as a favorite there's always the chance of kneel-downs late. ESPN (6.4) and another projection model (6.1) sit under 6.5, while Sleeper (7.4) and the expert consensus (6.6) are over. At +100 we only need 50% to break even, and the model's read at the higher line gives us room to give up the attempt. The play: Hurts under 6.5 rush attempts at +100.

The Giants host Dallas as 3-point underdogs in a 47.5-total game, and Jaxson Dart's rushing line sits at 6.5. Seven attempts is a lot to ask of a quarterback whose team figures to be trailing and throwing, and a home underdog is less likely to be kneeling out the clock late.

Every projection we track has him under the number: Sleeper 5.9, the expert consensus 5.6, ESPN 5.9 and another projection model 6.0, with our model at 5.6 attempts and a 57% read on the under. The price is steep at -130 (break-even 56.5%), so this is a narrow edge, but four independent sources landing between 5.6 and 6.0 against a 6.5 line is the kind of agreement we want. The play: Dart under 6.5 rush attempts.

Washington goes to Philadelphia as a 6-point underdog in a 43.5-total game. A trailing script tends to turn a running quarterback into a passer, and Jayden Daniels needs eight rushing attempts to beat this number.

Our model projects 6.6 attempts and makes the under a 59% play at 7.5. Sleeper (7.1), the expert consensus (6.7) and another projection model (7.2) all sit below the line, with ESPN just over at 7.6. At -105 the break-even is 51.2%, so the model's 59% leaves a solid cushion, and three of the four outside sources back it. The play: Daniels under 7.5 rush attempts.

Green Bay visits Minnesota as a 2.5-point underdog, and the matchup grades as poor for Jordan Love as a runner. A 2.5 line means three attempts beats us, so this is a bet on Love staying in the pocket in a game where the Packers are likely to be throwing.

Our model projects just 1.6 attempts and makes the under a 67% play, a big edge at -115 (break-even 53.5%). The outside projections are less certain: the expert consensus has 2.4, while Sleeper (3.0), ESPN (2.9) and another projection model (2.7) sit just above the line. Those are fractions of an attempt either way, and we side with the model, which sees the clearest gap. The play: Love under 2.5 rush attempts.

Buffalo opens in Houston as a 1.5-point favorite with a 44.5 total. Josh Allen's rushing line is set high at 6.5, and our model doesn't expect him to need that many carries in a close game where the Bills are favored only slightly.

The model projects 5.5 attempts and makes the under a 63% play. Sleeper (5.7) and the expert consensus (6.3) are under the line; ESPN (6.8) and another projection model (6.6) are just over. At -110 we need 52.4%. The edge is the full attempt between the model's projection and the line, and in a one-score game script we'll bet on Allen landing closer to five or six. The play: Allen under 6.5 rush attempts.

The Chargers open at home as 8.5-point favorites over Arizona, and Justin Herbert's rushing line sits at 4.5. Big favorites can pick up late kneel-downs, which count as rush attempts, so this one needs Herbert to be a pocket passer for most of the afternoon, which is what the projections expect.

Our model projects 3.7 attempts and makes the under a 52% play. The expert consensus (4.1), another projection model (4.3) and Sleeper (4.46) are all under the line, with ESPN just over at 4.7. At -110 we need 52.4% to break even, so this is close to an even proposition that we take on the model's projection and an outside consensus sitting just under the number. The play: Herbert under 4.5 rush attempts.

Minnesota hosts Green Bay as a 2.5-point favorite, and Kyler Murray's rushing line sits at 5.5. A line that high tells you the market expects him to run, but the projections expect something closer to four or five attempts in a tight divisional game.

Our model projects 4.5 attempts and makes the under a 63% play. The expert consensus (5.1), ESPN (4.7) and another projection model (4.6) all sit under 5.5, with Sleeper the exception at 5.9. The price is steep at -130 (break-even 56.5%), but the model clears it by almost seven points with three of the four outside sources in agreement. The play: Murray under 5.5 rush attempts.

Baltimore opens in Indianapolis as a 3-point favorite with a 48.5 total. Lamar Jackson's rushing line is 6.5, and this is a spot where our model breaks from the crowd: it projects 5.5 attempts and makes the under a 63% play.

To be clear, the outside projections lean the other way: Sleeper 8.1, the expert consensus 6.8, ESPN 7.2 and another projection model 7.6. Our quarterback rush-attempt unders are built on the model's number, and it's confident here even with the outside sources leaning over. At -120 (break-even 54.5%) the model's read still clears the price. The play: Jackson under 6.5 rush attempts.

Miami visits Las Vegas as a 3-point underdog with one of the lowest totals of the week (40.5). Malik Willis's rushing line is 6.5, and a team chasing points tends to lean on the pass.

Our model makes the under a 62% play at 7.5 and projects 6.5 attempts. We're betting the 6.5 line, a full attempt lower, which costs real cushion, but at plus money (+105) the break-even drops to 48.8%. ESPN (6.3) and another projection model (6.3) are under the number, while Sleeper (8.0) and the expert consensus (7.8) are over it, so this is a split decision outside the model, and the plus price is what makes it playable. The play: Willis under 6.5 rush attempts at +105.

Jacksonville opens at home as an 8.5-point favorite over Cleveland with a 40.5 total. Big favorites can collect kneel-downs at the end, which count as rush attempts, but Trevor Lawrence's projections sit comfortably under this number even so.

Every source we track has him below 4.5: Sleeper 4.4, the expert consensus 3.9, ESPN 4.4 and another projection model 3.9, with our model at 3.6 attempts and a 57% read on the under. At -105 we need 51.2% to break even, and that is a cushion with full agreement behind it. The play: Lawrence under 4.5 rush attempts.

New Orleans opens in Detroit as a 7-point underdog with one of the highest totals of the week (49.5). A big road underdog in a high-scoring setup is likely to be throwing, which works against Tyler Shough's rushing volume.

Our model projects 3.6 attempts and makes the under a 63% play at 4.5. The expert consensus (4.2) and ESPN (3.8) agree; Sleeper (5.0) and another projection model (4.9) are slightly over. At -130 the break-even is 56.5%, and the model clears it by six points. The play: Shough under 4.5 rush attempts.