Key takeaways
- Two 2-0 teams already look shaky. Being undefeated in September isn't the same as being good in September, and the board reflects that.
- Falcons at Packers is the Thursday trap. Green Bay is home-favored by roughly a touchdown and has won 13 straight home openers — but Atlanta's skill talent is real, and short-week road dogs cover more than casual pickers assume.
- Chiefs are road-favored by 7.5 in Miami. Big number, road team, September heat. Confidence-pool bracket killer if you get it wrong.
- The market and our model don't agree on the safest lock. That's the pick worth stealing, and it's on the Pick Insights board.
- Lock your picks before Thursday night. The Falcons–Packers kickoff is the deadline; anything after that costs you Week 3 leverage.
Two weeks in and half your league is already telling you their team is a fraud, and the other half is quietly hoping nobody notices they haven't hit a pick yet. Welcome to the part of the season where pick'em pools stop rewarding vibes and start rewarding whoever actually looked at the schedule on Wednesday.
Week 3 is the first real board of the year — the reactionary Week 1 mispricings are gone, the market has caught up, and every game on the slate has an argument on both sides. Here's the read on who to trust, who's a trap, and where the model is telling us the crowd is wrong.
The Week 2 hangover: hot, cold, and lying to you
The teams that look real after two weeks and the teams that look real after two weeks are not the same list. Sort them out before Wednesday night, not after your Sunday morning has already gotten away from you.
Actually hot
The Bills went into a real opponent's building and won a game they were supposed to win, which sounds boring until you remember how many teams (looking at you, Cincinnati and Detroit last year) fumble those. Josh Allen is playing like Josh Allen. The Chiefs opened the season by dismantling Denver in a way that looked more like 2023 Chiefs than 2025 Chiefs, and their defense graded near the top of the league through two weeks. The Bears — yes, the Bears — are running the ball with purpose and letting Caleb Williams make plays down the field. This is not a drill.
Already cooked (or close)
Denver was blown off the field in Kansas City and has looked flat on both sides. The Broncos aren't 0-2 by accident. The Chargers laid a genuinely disgraceful egg against Arizona in Week 1 — the defense couldn't stop Jacoby Brissett, and if you're paying that off in confidence points this week against Buffalo, you have questions to answer. And on the NFC side, the Packers and Patriots both got Week 2 gut-punches that dropped them multiple spots in every power rankings column that matters. Green Bay gets a chance to reset on Thursday. New England doesn't.
Lying to you
Two 2-0 records that the market is not buying: watch how the spreads move as the week goes on. When the book prices an unbeaten team as an underdog to a 1-1 opponent, that's the market telling you which record is real and which one was schedule luck. Don't fight the number just because the standings look pretty.
Undefeated in September is a stat. Beating a good team on the road is evidence. Don't confuse the two on your pick sheet.
What the market is saying: the Week 3 board at a glance
The lines are the starting point for a pick'em read, not the finish line. Here's where the biggest games opened, with the pick'em read on each:
| Game | Market lean | Pick'em read |
|---|---|---|
| Falcons at Packers (Thu) | GB ~ −7 | Popular pick, real risk. Short week, road dog with weapons, Packers coming off a Week 2 stumble. |
| Chargers at Bills | BUF ~ −3 | High total. Not a rock fight — if you're taking LAC, you need shootout math, not a defensive stand. |
| Chiefs at Dolphins | KC ~ −7.5 | Big road spread. Confidence-pool players — be careful stacking this at max. |
| Vikings at Buccaneers | Close, TB slight fav | The board's true coin flip. Where the model earns its money. |
| Rams at Broncos | Close, LAR slight edge | Two teams heading opposite directions. Momentum is real; the number isn't priced for it yet. |
Everything else on the Week 3 slate — Colts hosting Texans in the early-season AFC South knife fight, Ravens getting Cowboys at home, 49ers going into LA against the Rams division rival — is priced tight enough that the expert consensus and the model have room to disagree. Which brings us to the useful part.
What the experts are saying (and where they're stalling)
The ESPN expert panel files their picks through the week, and CBS's Pete Prisco drops his weekly column midweek — both were still filling in as of Wednesday morning. Historically, once they land, the consensus follows the market on the two big-spread games (Packers and Chiefs) and splits on the closer ones. Prisco has a well-earned reputation for calling one weekly upset that the crowd hasn't priced yet — those are the picks worth reading his column for, not the chalk locks.
The tell to watch for once the columns drop: where the panel diverges from Vegas. If four ESPN writers pick the Chargers to win outright at Buffalo while the number stays at Bills −3, that's an information asymmetry you can act on before your pool locks. If the panel is unanimous with the market, congrats, you learned nothing. The disagreements are the alpha.
Do this: Check the ESPN panel and Prisco's column between Wednesday afternoon and Thursday morning, then compare their picks against the model's board on Pick Insights. Any game where all three (crowd, expert, model) land on the same side is a legit confidence-pool anchor. Anywhere they split, you're getting paid to have an opinion.
Trap watch: Public money will pile on the Chiefs at −7.5 and the Packers at −7 because they're the two biggest brand names on the board. When everyone in your pool picks the same two games, those picks are worth zero leverage even if they hit. The wins that move you up the standings are the ones half your league got wrong.
Where our model is pounding the table
The DaqsPickEm model refreshes as injury reports finalize and lines settle midweek — the Wednesday morning board is still warming up as we publish this. Once it locks (typically by Wednesday evening, before the Thursday-night deadline), the full slate lands on the Pick Insights page with every game, every confidence tier, and the reasoning behind the calls the model is willing to plant a flag on.
What the model does that a straight power-rating doesn't: it prices ownership. In a big pick'em pool, picking the game right isn't enough — you have to pick it right in a way that gains ground on the field. That's why the model's highest-confidence board sometimes leans into a game the market has priced as a coin flip, and away from a chalky 7.5-point road favorite that half your league is already locking in. Beating the pool means beating the pool's picks, not beating Vegas.
If you only check one thing this week, check the model's upset alert tier — that's where the leverage lives. And if you only pick one game confidently, don't make it the one everyone else in your league is picking confidently. That's a truth about pool math, not a hot take.
How to actually play Week 3
The three-step playbook that beats vibes:
- Lock the two anchor games where the market, the expert consensus, and the model all agree. In a normal week, that's two or three games. Those are your foundation.
- Find the one contrarian call where the model diverges from the crowd and you agree with the reasoning. That's your leverage pick.
- Fade the biggest chalk spread in your confidence rankings, not on the pick itself. Picking the Chiefs is fine. Picking the Chiefs as your top-confidence game when 90% of your league is doing the same thing is throwing away points.
The Falcons kicking off in Green Bay closes the door on your Week 3 picks — if it's Friday morning and you haven't locked, you missed the game that mattered. Get your picks in, check the model's full board before Thursday night, and if you like calling upsets, that's literally the game we built. See you on the leaderboard.