Key takeaways
- Thursday-night short-week trap: Patriots at Bears, Chicago -3. ESPN's FPI gives the Bears a 60% win probability and Pete Prisco's cover line is a flat "no way I'm picking against the Bears." The market agrees. A fine confidence-pool floor play, not a difference-maker.
- Our model's biggest lock: Houston -9.5 at home against the Giants, grading out at 75.8% confidence. The one game on the board where you can basically close your eyes.
- Our model's biggest contrarian call: Dallas +4.5 at Philadelphia on Monday night, 71% confidence. Every public betting splits sheet has the Eagles. Our model does not.
- The upset the experts are quietly circling: Browns at Titans. CBS Sports picks Cleveland on the road. Our model picks Tennessee at home but only at 50.5% — one of two true coin flips of the week.
- Full board, every injury-weighted lean, every sharp-money read, lives on Pick Insights.
You've got about 48 hours to lock in Week 7 picks and the slate did not come to make life easy. Fourteen games, and only two of them — Houston against the Giants and Minnesota against Indianapolis — show a market win probability higher than 65%. The rest is coin-flip country, which is exactly the kind of week where pools get won by the person who dared to call one upset right, and lost by the person who "just picked home favorites."
Here's where everybody actually stands going into Thursday night, who the experts are backing, and the two calls our model is willing to put its name on in writing.
Who's hot, who's cold going into Week 7
You don't need advanced stats to know the Giants are 9.5-point road dogs at Houston for a reason — that spread IS the stats. The important thing for a pick'em entry is not where every team ranks on some power ranking, it's which teams are trending in a direction the market has not fully caught up to yet.
Trending up
The Texans are the clearest story on the board. A 9.5-point home favorite is a line that only shows up when a team has strung together enough wins that the market is scared to NOT price them like a contender. There's steam on Houston's number this week, meaning books kept moving the Texans even as public money piled in — the betting world's politely-phrased way of saying "the sharps agree."
The Bears are the Thursday-night sleeper that stopped being a sleeper. Caleb Williams keeps uncorking long completions, Chicago's pass rush has quietly been one of the best in the league, and now they get a Patriots offensive line held together with painter's tape on a three-day week. The three-point line is the market saying "we buy it." ESPN's FPI number at 60% is the public saying the same.
The 49ers, somehow, are STILL rolling against Atlanta even with half the roster on the injury report. Kyle Shanahan plays a different sport than the rest of us.
Trending down
The Giants — a 9.5-point dog is a mercy spread. The market stopped pretending.
The Dolphins are road underdogs TO THE JETS. Nothing more needs to be said. In a league where betting lines double as a weekly obituary, this one is written.
The Falcons keep opening as home favorites against banged-up 49ers teams and keep not covering them. At some point the market will catch on. This is not that week.
The pick'em trap of the week: five games on the board have spreads inside a field goal — Patriots-Bears, Steelers-Saints, Browns-Titans, Bengals-Ravens, and Chiefs-Seahawks. If your pool allows a safety-first strategy, pick AROUND those games. The pool is going to split on all five, meaning they will not separate you from the pack even if you nail them. Your edge lives in the clear favorites and the one smart upset call.
What the experts are saying
Pete Prisco's CBS write-up this week leans heavily chalk, which is unusual for him — he's a man who normally has a sneaky dog or two circled on the sidebar. His five featured picks:
| Game | Prisco's pick | Score |
|---|---|---|
| Patriots at Bears | Bears -3 | 27-20 |
| Bengals at Ravens | Bengals +3 | 27-24 |
| Packers at Lions | Packers +3 | 31-24 |
| Chiefs at Seahawks | Chiefs +2.5 | 27-20 |
| Cowboys at Eagles | Cowboys -4.5 | 23-17 |
Prisco's thesis on Bengals-Ravens is worth stealing wholesale: in a battle of two quarterbacks who have dealt with injuries, take the one who is actually healthy right now. Translated for pick'em: Cincinnati is live even on the road because Baltimore's banged-up side is MORE banged-up. That's one of the few places this week where you can beat the market on injury information the public hasn't fully priced in.
The CBS Sports expert panel gets more interesting where it diverges from the betting market:
- Picks the Browns on the road at Tennessee, 20-13 — a straight-up pick on a 1.5-point dog. One of the few experts who thinks Cleveland is more alive than the number says.
- Picks the Jets at home against Miami, 23-16 — the only team in football the Jets are favored against, and the panel is still buying it.
- Picks the Broncos on the road at Arizona, 19-16 — a road underdog outright. Call it the week's most quietly confident upset pick.
- Picks Dallas on Monday night at Philadelphia, 23-17, as a +4.5 road dog. The expert consensus matches our model here, and that is not nothing.
ESPN's expert panel tracks tighter to the market across the board — the FPI has Chicago at 60%, Houston comfortably at the top of the weekly chart, and Philadelphia favored on Monday night. The one place ESPN's analytics quietly diverges from the public: it is NOT nearly as high on Kansas City as the pool betting data suggests. Which is the perfect bridge to…
Our model's corner
Our in-house AI model — the one that powers Pick Insights — graded all 14 games this week and came back with a confidence ladder that mostly agrees with the sharps and sometimes doesn't. Two picks the model is willing to put its name on in writing:
The lock: Texans −9.5 vs. Giants
Model confidence: 75.8%. The highest number on this week's board, and it's not close.
The read is clean. Houston is the home favorite, the line has steam WITH the public (meaning sharps and the crowd agree, which is rare), and the Giants have no realistic path to covering 9.5 against a healthy, trending-up roster. In a confidence pool, this is where your highest-point allocation goes. In a straight-up pool, this is your "close your eyes and press go" square.
The market priced the Giants out of the game. Our model priced them out of the YEAR.
The contrarian: Cowboys +4.5 at Philadelphia
Model confidence: 71.1%. Second-highest on the board, and the one that will get you funny looks in your group chat.
Dallas is a 4.5-point road dog on Monday night and the public is on Philadelphia. Our model does not see it that way. The injury report cuts cleanly in Dallas's direction, and the Eagles' recent run has looked more like good variance than good football. The model's note on this one is unusually blunt — essentially "the healthier side is being underpriced."
If your pool awards upset bonuses, THIS is where you spend yours. A straight-up Cowboys call pays almost as well as hitting the primetime game of the week, and the model is grading it as a near-lock. In survivor, you don't play it — survivor rewards boring. In a confidence or ATS pool, you very much do.
Everything else — what the model has on the Steelers-Saints Paris game, where it stands on Chiefs-Seahawks at Lumen Field, the two true 50/50 reads it won't touch with a 10-foot pole — is sitting on Pick Insights, including the per-game confidence number and the reason text behind every call. We give you the lock and the upset here for free; the full board is where the model earns its keep.
Setting your Week 7 slate
If you're entering a confidence pool: Houston gets your top number, period. Minnesota at home against Indianapolis gets your second-highest tier — the model has that one at 67.9%, the market agrees comfortably, and nobody in your pool is picking against it. Stack chalk home favorites you don't actively hate (Chicago, Carolina, Baltimore) in the middle, and spend your bottom-tier numbers on the five coin-flips.
If you're entering a survivor pool: Houston is also fine here, and this is one of those weeks where "fine" is enough. Do NOT burn a top team you're saving for later just because you love Dallas on Monday night. Save your contrarian swing for your bracket entries, not your survivor life.
If you're playing ATS: Cowboys +4.5 is where the sharp money is quietly going. If you want a second ATS swing, our model also likes the Broncos getting points at Arizona — same "healthier side is underpriced" thesis.
The slate kicks off Thursday night at 8:15 ET. If you want to beat the pool, pick before Thursday afternoon, while the public is still flipping a coin. Week 5's preview had us pounding one road dog on Monday night too. Week 4's had a sneaky Thursday-night upset flagged early. The pattern is simple: pools reward the person who did the homework by Wednesday, not the person who flipped a coin at kickoff. Our fantasy desk covers the lineup half of this; the pick'em half is right here, and it's on you to actually set the thing.
Lock your picks before Thursday night. Trust the lock. Consider the upset. And if you like calling road dogs on Monday night, that's literally the game.