NFL RESEARCH

Our score: PIT 25, NO 22

STAT: New Orleans have allowed a touchdown on 23 of the 72 drives they have faced, 31.9%, the fourth-highest rate in the league. This is a dome in Paris with no weather to argue about, the Saints' six games have averaged 55.0 points apiece and Pittsburgh's last three produced 51, 56 and 45. Against it: 44.5 is already the highest total Pittsburgh have been given all season — their previous high was 42.5 — and the Steelers score only 1.90 points a drive, so the under needs nothing more than the Weeks 1 and 2 version of this offence, whose games produced 33 and 23. No side: the drive model has Pittsburgh by three on a 3.5-point line, and their defence (1.86 points a drive allowed, a 14.5% drive touchdown rate behind only Minnesota and level with Kansas City) is the best unit on the field without being worth laying the number at a neutral site. One unverified source lists Chris Olave questionable; treat that as a re-check, not a fact.

Our score: SF 27, ATL 21

STAT: San Francisco score a touchdown on 36.7% of their offensive drives, 22 in 60, and allow one on 16.1% — both comfortably the best marks in this game. Atlanta are 0-3 at home, scoring 16.3 and allowing 31.7 in those three, and both their wins came on the road. All six San Francisco margins clear this number: 20, 22, 6, 10, 3 and 6, and a 6-0 team is getting only 2.5 on the road. Against it, and it caps the grade at C+: four of those six margins are single digits and the last three were 6, 3 and 6, the home/road split that makes Atlanta look broken is three games each way, Atlanta run for better than five yards a carry, and San Francisco have a day less rest. The moneyline is graded a notch lower — -155 on a team we project to win by five or six is fair, not generous. No total; San Francisco's games have averaged 47.8 and Atlanta's 49.2, which is the line.

Our score: CLE 17, TEN 13

STAT: Tennessee have scored 10, 20 and 3 in three home games and lost all three. Last week's 26 at Indianapolis reads like a corner turned and is not — four Slye field goals from 29, 56, 49 and 44, an 8-yard Ward touchdown and a blocked punt scooped at the Indianapolis 14 and taken in, all on 102 passing yards. Tennessee reach the end zone on 11.9% of their drives and Cleveland on 14.1%, the two worst rates in this game, which is how you get the lowest total on the board. Against the under: both teams are field-goal machines, which pushes games toward a number rather than under it, and one special-teams score decides a game this small, exactly as happened to Tennessee last week. No side — 2.5 points of projected edge on a 1.5-point line is not enough to lay it, Cleveland score 15.0 on the road and put up 12 and 13 in two of their last three, and Quinshon Judkins did not appear in a Week 6 box score and has no posted market this week. Re-check him Sunday.

Our score: IND 18, MIN 23

STAT: Minnesota have scored eight offensive touchdowns in five games and kicked 15 field goals — only 14.3% of their drives reach the end zone. Correct the record while we are here, because it has been quoted wrong: they are 5-0 straight up but 4-1 against the spread, and the one miss is the only time anyone laid a big number with them, a 15-10 home win as a 10-point favourite. Their margins are 17, 6, 7, 5 and 10, so just twice have they cleared this number by more than a field goal. It is a D and not better because everything underneath points the other way: Minnesota allow 1.16 points a drive, the fewest in the league, Daniel Jones has four passing touchdowns against six interceptions, Indianapolis have given it away on 12 of 66 drives, and gone out on downs on five more, and the Vikings are home off a bye with 14 days' rest to seven. No total — it has already been bet from 47.5 to 41.5 and our projection lands on 41, so there is nothing left to take.

Our score: MIA 14, NYJ 21

STAT: the Jets have scored 24 points or fewer in all six games — 23, 17, 24, 12, 17, 19 — and their two home dates have averaged 33.0 total points. Miami score 12.0 a game on the road and reach the end zone on 14.9% of their drives, with Malik Willis having taken every Dolphins attempt this season for 849 yards, two touchdowns and four interceptions. Against it, and it is why this is only a C: Miami allow a touchdown on 31.3% of the drives they face, among the five worst rates in the league, they come off a bye with 14 days to the Jets' seven, and 39.5 is already the second-lowest number on the board — the market is here ahead of us. No side at all: two and a half points between our projection and a 4.5-point line is noise, and -230 is no price for a team that has not scored 25 all year. Breece Hall has not appeared in a box score since Week 3 and no Jets running back has a posted market this week, which is its own hint; re-check the backfield.

Our score: TB 20, CAR 24

STAT: Tampa Bay have lost five games by 6, 4, 7, 3 and 7 points and are still only 1-4-1 against the spread. That is the trap on the other side of this game — "they keep it close" is already in the price, and the margin they keep losing by is one a field goal does not rescue. Carolina are the better offence by a distance, 0.116 EPA a play to Tampa Bay's -0.130, with a 69.6% red-zone touchdown rate against the Buccaneers' 50.0%. It is a D because the edge is about a point on a three-point line, Carolina's rate stats lean on a 37-point night in a 59-37 loss, they allow 5.22 yards a carry, and the quarterback is unresolved: Baker Mayfield has Week 7 props posted and Jalon Daniels has none, with Carolina's moneyline drifting from -205 on Monday night out to -166 by Wednesday afternoon, which is money coming back to Tampa Bay. Tampa Bay scored 20.7 a game in Mayfield's three starts and 19.0 in Daniels' three, so the swing is smaller than it reads — confirm the starter anyway. No total: our two methods disagree by five points on the same game.

Our score: NYG 20, HOU 26

STAT: Houston have beaten exactly one team by more than five points all season — 26-3 at Tennessee — and their other five margins are -5, -14, -2, -4 and +3. They are 2-4 with a plus-one point differential and 10.5 is by far the largest number they have been asked to carry. The honest problem, and it is what caps this at C: the line opened Houston -5.5 and has been bet to -10.5, and the Giants' Week 7 quarterback props all belong to J.J. McCarthy with none posted for Jameis Winston. The market is pricing a first start we cannot size, and when our own blend reads Houston by six against a -10.5 line, the right response is suspicion of the model, not confidence in the dog. No total for the same reason — the 60, 54 and 61 points New York's last three games produced were Winston throwing. Cam Skattebo also missed Week 6 and his rushing prop is 31.5 against a 58.8-yard average, so the backfield may be compromised too. No moneyline: +390 is the same opinion with none of the cushion.

Our score: DEN 22, ARI 19

STAT: Denver have not won a game by more than seven points this season — 20-13, 30-26, 23-17 and 17-14. That is why there is no side at 4.5, even against a team that has lost both of its home games by exactly 24; our projection lands Denver by three against a 4.5-point line, which leans Arizona by too little to post. The total is the half worth having: Denver's six games have averaged 39.8 points and gone under in five of six, and neither offence finishes drives, Denver scoring a touchdown on 16.7% of them and Arizona on 18.2% with 14 field goals. Against it — Arizona give up 7.33 yards a pass play, fourth-most in the league, and their two home dates produced 38 and 66 points, so the live way this goes over is Arizona's defence, not Denver's pace. One to re-check: no Arizona running back has a posted prop market this week, and Jeremiyah Love, their leading rusher, missed Week 6.

Our score: LAR 26, LV 23

STAT: Las Vegas have scored 27.0 points a game and allowed 27.0 — a dead-even margin through six games, against the Rams' 22.5 and 23.0 — and the visitor is laying 4.5. The Raiders are 5-1 against the number and both home losses came by three, 27-30 to Kansas City and 30-33 to Buffalo. It is a D because that is a market-price argument and not a quality one: the Rams allow a touchdown on 19.0% of the drives they face against Las Vegas' 29.4%, and the Raiders give up 5.18 yards a carry to a team that runs at 4.76. No total — our number is 49 on a 48.5 — and a warning before anyone touches this game: Davante Adams has no posted prop market for Sunday while eight other Rams do, and he is 49 targets and 480 yards of that offence. The line has been bet from LAR -7 down to LAR -3.5, so the sheet carries the live number.

Our score: KC 23, SEA 21

STAT: Kansas City have outscored their opponents by 11.4 points a game, the best margin in the league, and they are the side getting points. What keeps this at C+ and nowhere near higher: those five opponents are a combined 9-20, Seattle allow 1.52 points a drive, second only to Minnesota, and 4.31 yards a pass play, the stingiest mark in the league, no quarterback has thrown for 190 on them in six games, and the Seahawks have ten days' rest to Kansas City's seven. The model that likes this side carries a 10.35-point mean error on margin in backtest, so a four-and-a-half-point disagreement with the market sits inside its own noise. One correction before anyone reads it off the sheet: our schedule table lists Drew Lock for Seattle and it is wrong again — Sam Darnold has taken every meaningful snap since Week 3 and the books have posted Darnold's props. No moneyline, because +124 is the same opinion at a worse price than the points.

Our score: DAL 22, PHI 19

STAT: no Philadelphia game this season has produced more than 46 points, and five of their six would have stayed under this 44.5. The Eagles punt on 51.6% of their drives, sit 30th of 32 in EPA per play, have scored 14.5 a game through four straight losses, and Jalen Hurts was ruled out for the rest of the season on Oct 19 after doctors found blood clots in his lungs (ESPN's Adam Schefter; the Eagles and his agent describe a DVT with bilateral pulmonary emboli) — the books have posted Tanner McKee's props for Monday, not his. We are telling you plainly that this is a trend pick and not a model pick. The argument against is real either way — Dallas have the worst defence in the league by EPA per play, quarterbacks have thrown 11 touchdowns against one interception on them, and Dallas' first four games produced 48, 57, 65 and 64 points. No side: Dallas are the better team but scored 16 and 13 the last two weeks, and the 4.5 already has the quarterback news in it.