Week 5 previews

Thu Oct 8, 7:15 PM CT

Tampa Bay Buccaneers at Dallas Cowboys

Tampa Bay keeps it within 8.5

Vegas says Dallas by 8.5, 47.5 total points. Our model says Dallas by 3.5, 45.6 total points.

We think Dallas wins, but by less than Vegas does. The Cowboys' offense has been excellent (about +0.20 EPA/play, 2nd of 32) and the defense has been leaky (+0.19 allowed, 31st of 32, where higher is worse). Tampa's offense has been bad (−0.16, 32nd of 32). With Baker Mayfield out, our model only has Dallas winning by 3.5.

Baker Mayfield is Out (dislocated thumb; ESPN return date Oct 25). The depth chart still lists him as QB1, and the nflverse week-5 report leaves his status blank. The schedule lists Jalon Daniels, so we use Daniels at quarterback. Dak Prescott starts for Dallas.

Newcomer tip: EPA/play Every snap has a fair point value before the ball is snapped. EPA (points added per play) is how much the play moved that number. Dallas's +0.20 offense means it is gaining about two-tenths of a point per snap on average.

We like Tampa Bay to keep it within 8.5. The gap is 5.0 points.

Official pick:

Sun Oct 11, 8:30 AM CT

Philadelphia Eagles at Jacksonville Jaguars

Over/under pulled Oct 8 (bug)

Too close to call

Vegas says Jacksonville by 7, 42.5 total points. Our model says Jacksonville by 5.6, 47.0 total points.

Jacksonville should win. The Jaguars rate as one of the better teams so far (+0.046 opponent-adjusted, 4th of 32) and our model has them by about 5.6 at home with Trevor Lawrence. Philadelphia is near average (−0.005, 17th of 32). Who wins by how much is close to Vegas at 7.

Newcomer tip: opponent-adjusted rating Raw stats get credit, or blame, taken away for soft or tough schedules. JAX at +0.046 (4th of 32) is near the top of this slate. PHI at −0.005 (17th of 32) is in the middle.

Official pick:

Sun Oct 11, 12:00 PM CT

Chicago Bears at Green Bay Packers

Too close to call

Vegas says Chicago by 2.5, 45.5 total points. Our model says Chicago by 3.1, 46.8 total points.

Chicago should win by about a field goal. The Bears rate better than Green Bay (+0.036, 7th of 32, vs −0.004, 15th of 32), and with Caleb Williams starting our model has Chicago by 3.1. Vegas already has the Bears by 2.5. That gap is about half a point.

The schedule still lists Tyson Bagent. The depth chart has Caleb Williams at QB1, and the ESPN injury report does not list him. We use Williams against Jordan Love.

Newcomer tip: success rate The share of plays that stay on schedule (enough yards for the down). Chicago's offense succeeds on about 49% of snaps (5th of 32). Green Bay's succeeds on about 39% (28th of 32).

Official pick:

Sun Oct 11, 12:00 PM CT

Cincinnati Bengals at Miami Dolphins

Over/under pulled Oct 8 (bug)

Miami keeps it within 6.5

Vegas says Cincinnati by 6.5, 42.5 total points. Our model says Cincinnati by 1.4, 45.5 total points.

We think Cincinnati wins a close game. Vegas has the Bengals by 6.5 on the road. Our model has them by about 1.4, because playing at home does not make up the difference in the team ratings. Joe Burrow against Malik Willis is a real talent gap. We still like Miami to keep it within 6.5.

Newcomer tip: the margin Vegas has Cincinnati winning by 6.5, a touchdown short a half-point. Miami keeps it within 6.5 if Miami wins or loses by 6 or fewer. We look at the difference between that number and our model's margin, and we only speak up at 2 points or more.

We like Miami to keep it within 6.5. The gap is 5.1 points.

Official pick:

Sun Oct 11, 12:00 PM CT

Cleveland Browns at New York Jets

Over/under pulled Oct 8 (bug)

Cleveland wins, or loses by 1 at most

Vegas says the Jets by 1.5, 39.5 total points. Our model says Cleveland by 1.1, 44.6 total points.

We think Cleveland wins narrowly. Vegas has it the other way. Both teams rate poorly (CLE −0.025, 26th of 32, NYJ −0.056, 31st of 32). Vegas has the Jets winning at home by 1.5. Our model slightly prefers Cleveland with Deshaun Watson over Geno Smith.

Newcomer tip: opponent-adjusted rating Raw EPA gets corrected for who you played. The Jets are 31st of 32 at −0.056, near the bottom, which is why our model will not treat them as the team that should win at home.

We like Cleveland to win or lose by 1 at most. The gap is 2.6 points.

Official pick:

Sun Oct 11, 12:00 PM CT

Houston Texans at Tennessee Titans

Over/under pulled Oct 8 (bug)

Tennessee keeps it within 7.5

Vegas says Houston by 7.5, 37.5 total points. Our model says Houston by 4.0, 44.7 total points.

Houston should win. Tennessee is the lowest-rated team on this slate (−0.071, 32nd of 32), and Vegas also has the Texans winning, by 7.5. Our model is less sure about the margin (Houston by 4.0). C.J. Stroud and Cam Ward are the starters.

Newcomer tip: score vs margin 37.5 total points means a low-scoring game. The projected total is 44.7. Who wins is a separate question from the score.

We like Tennessee to keep it within 7.5. The gap is 3.5 points.

Official pick:

Sun Oct 11, 12:00 PM CT

Indianapolis Colts at Pittsburgh Steelers

Indianapolis keeps it within 2.5

Vegas says Pittsburgh by 2.5, 44.5 total points. Our model says Pittsburgh by 1.3, 45.9 total points.

We think Pittsburgh wins a close one at home. Our model has the Steelers by 1.3, and Vegas has them by 2.5. Daniel Jones against Aaron Rodgers barely changes the quarterback part of the number. The Wednesday gap is 1.2 points on the margin and 1.4 on the score, closer to Vegas than it was on Monday. The official pick locked Monday still stands.

Newcomer tip: home field Home field in our model is about 1.5 points. Pittsburgh's advantage here is mostly that home field, on two teams rated near each other.

Official pick:

Sun Oct 11, 12:00 PM CT

Las Vegas Raiders at New England Patriots

Too close to call

Vegas says New England by 3.5, 45.5 total points. Our model says New England by 3.1, 45.8 total points.

New England should win by about a field goal. Our model has the Patriots by 3.1 at home with Drake Maye, and Vegas has them by 3.5. Kirk Cousins is the Raiders' starter. The gap on the margin is 0.4 and the gap on the score is 0.3.

Newcomer tip: early-down EPA Scoring drives are usually won on first and second down. New England's early-down offense is about −0.08 EPA/play (points added per play, 26th of 32). Las Vegas is about −0.03 (20th of 32).

Official pick:

Sun Oct 11, 12:00 PM CT

Minnesota Vikings at New Orleans Saints

We pulled our over/under call for this game on Oct 8 after a data error.

Mid-week check (not an official pick)

Minnesota wins by 2 or more

Vegas says Minnesota by 1.5, 42.5 total points. Our model says Minnesota by 4.3, 46.0 total points.

Minnesota should win by more than a field goal. The Vikings rate better (+0.027, 10th of 32, vs NO −0.034, 27th of 32), and Kyler Murray against Tyler Shough pushes the margin to Minnesota by 4.3. Vegas only has the Vikings by 1.5. On Monday the gap on the margin was under 2 points. It is big enough now.

Tyler Shough is Questionable (hand) on the ESPN injury report. We assume he starts. If he does not, the quarterback piece of this number no longer applies.

Newcomer tip: defense EPA allowed Lower is better. Minnesota's defense has allowed about −0.21 EPA/play (points added per play, 1st of 32). That is a big reason they rate above the Saints.

We like Minnesota to win by 2 or more. The gap is 2.8 points.

Official pick:

Sun Oct 11, 12:00 PM CT

New York Giants at Washington Commanders

Too close to call

Vegas says Washington by 3.5, 42.5 total points. Our model says Washington by 5.1, 44.4 total points.

Washington should win at home. Both teams still sit near the bottom on team rating (WAS −0.040, 29th of 32, NYG −0.050, 30th of 32), but Jayden Daniels as the starter swings the quarterback part toward the Commanders. Our model has Washington by 5.1. Vegas has them by 3.5, a 1.6-point gap, and the scores are 1.9 apart.

Jayden Daniels is the depth-chart QB1 and he is not on the ESPN injury report. ESPN's Wednesday note says he fully practiced and is expected to play. Jameis Winston is the Giants' starter. Monday's preview used Athan Kaliakmanis because Daniels was still listed out. This one uses Daniels.

Newcomer tip: why the quarterback note matters The team rating blends whoever has played. The quarterback adjustment re-aims the margin at the starter going into this game. Monday's number and this number are different because the starter changed.

Official pick:

Sun Oct 11, 3:05 PM CT

Denver Broncos at Los Angeles Chargers

Over/under pulled Oct 8 (bug)

Too close to call

Vegas says Denver by 3.5, 41.5 total points. Our model says Denver by 2.6, 46.7 total points.

Denver should win a close road game. The Broncos rate clearly better (+0.040, 5th of 32, vs LAC −0.010, 20th of 32) and Vegas has them by 3.5. Bo Nix against Justin Herbert is close to that number (our model: Denver by 2.6).

Newcomer tip: explosive play rate The share of plays that gain 20 or more passing or 15 or more rushing. Denver's offense gets a big play on about 5.3% of snaps (24th of 32). The Chargers get one on about 6.1% (15th of 32).

Official pick:

Sun Oct 11, 3:25 PM CT

Detroit Lions at Arizona Cardinals

Over/under pulled Oct 8 (bug)

Arizona keeps it within 5.5

Vegas says Detroit by 5.5, 54.5 total points. Our model says Detroit by 1.3, 45.9 total points.

Detroit should win a one-score game. The Lions rate a bit better than Arizona (+0.011, 11th of 32, vs −0.023, 24th of 32), and Jared Goff against Jacoby Brissett fits that. Vegas has Detroit by 5.5. Our model wants Detroit by 1.3. That margin gap is large enough for a pick.

Newcomer tip: when the score looks high 54.5 still means a track meet. The projected total is 45.9, built from both teams' offensive and defensive EPA.

We like Arizona to keep it within 5.5. The gap is 4.2 points.

Official pick:

Sun Oct 11, 3:25 PM CT

San Francisco 49ers at Seattle Seahawks

San Francisco wins, or loses by 2 at most

Vegas says Seattle by 2.5, 45.5 total points. Our model says San Francisco by 0.3, 48.0 total points.

We think San Francisco is the better team, so we like them to keep it close. The 49ers are the highest-rated team on the slate (+0.068, 1st of 32), with an offense at about +0.27 EPA/play (1st of 32). Seattle is solid (+0.029, 8th of 32) and Vegas has the Seahawks winning at home by 2.5. With Sam Darnold starting, our model only has the 49ers by 0.3, which is still far enough from Vegas to like San Francisco keeping it within 2.5.

The schedule still lists Drew Lock. The depth chart and ESPN have Sam Darnold as the starter. We use Darnold. Brock Purdy starts for San Francisco.

Newcomer tip: pass EPA San Francisco's passing sits around +0.54 EPA per pass play (points added per pass play, 1st of 32). That is the unit carrying a top rating after four weeks.

We like San Francisco to win or lose by 2 at most. The gap is 2.8 points.

Official pick:

Sun Oct 11, 7:20 PM CT

Baltimore Ravens at Atlanta Falcons

Mid-week check (not an official pick)

Baltimore keeps it within 3.5

Vegas says Atlanta by 3.5, 43.5 total points. Our model says Baltimore by 0.1, 46.6 total points.

This is a toss-up on our model, with Baltimore a tenth of a point in front. Team ratings still favor the Ravens (+0.037, 6th of 32, vs Atlanta −0.012, 21st of 32). Before quarterbacks, our model had Baltimore by about 2.5. Lamar Jackson and Michael Penix Jr. pull it to Baltimore by 0.1. Vegas now has Atlanta by 3.5. On Monday it was Baltimore by 2.5.

Lamar Jackson is not on the ESPN injury report. On Monday he was listed questionable. Michael Penix Jr. is Atlanta's depth-chart QB1 and is not listed either. The schedule and the depth chart agree on both starters.

Newcomer tip: only games already played Each week's rating uses only games already played. Wednesday's ratings include Monday night's completed week-4 game, Falcons 45, Saints 24. That game is why Atlanta's rating rose and New Orleans's fell. Opponent adjustment then shifted the other teams a little.

We like Baltimore to keep it within 3.5. The gap is 3.6 points.

Official pick:

Mon Oct 12, 7:15 PM CT

Buffalo Bills at Los Angeles Rams

Over/under pulled Oct 8 (bug)

Too close to call

Vegas says the Rams by 3, 54.5 total points. Our model says Los Angeles by 4.3, 47.9 total points.

The Rams should win at home. They are near the top of the ratings (+0.061, 2nd of 32) and our model has them by about 4.3 with Matthew Stafford. Buffalo is strong too (+0.028, 9th of 32) with Josh Allen. Vegas only has them by 3, a 1.3-point gap, too small for a pick on the winner.

Newcomer tip: a big score Our model's total is offensive and defensive EPA turned into points. It has no separate bump for a Monday night game.

Official pick: