Denver won a Thursday night game and spent the weekend arguing about its quarterback. Around the league, that's the shape of this day: a Broncos defense good enough to steal a game, a Broncos offense that leaves the building unconvinced, and a computer model in a casino's pocket insisting it knows Week 6 better than the price does.
Start with the defense, because it's the part that isn't in dispute. Vance Joseph said Denver's unit tricked Sam Darnold into two interceptions in the Broncos' Thursday Night Football win, and the box score backs the con. Darnold finished 27-of-51 for 278 yards, no touchdowns, two picks, five sacks taken and a 52.6 passer rating. Fifty-one attempts at 5.5 yards a throw is not a passing offense, it's a punt with extra steps. Jaxon Smith-Njigba caught 12 of 16 targets for 134 yards and never scored, which tells you where those throws lived: short, contested, nowhere near the end zone. My read is that Joseph's group is good enough to win games on its own, and that's exactly what makes the other conversation so loud.
Ryan Fitzpatrick questioned Bo Nix's competitiveness after the Seahawks game. Aqib Talib defended him. Competitiveness is the cheapest critique in football because nothing can disprove it. There's no metric attached, no down-and-distance, no throw to argue over. But the volume of the shouting is itself the data point: Denver won a game in prime time and the postgame topic was still the quarterback. The debate is a proxy for a real complaint, and I think that complaint is that the Broncos offense hasn't turned Joseph's takeaways into points. Nix kills this story by finishing drives, not by winning a broadcast argument. If Denver's scoring stays flat while the defense keeps handing it short fields, the criticism migrates from television into the building, and that's when it starts costing somebody a job.
SportsLine's model simulated every Week 6 game 10,000 times and came out backing the New England Patriots and the Arizona Cardinals. Ten thousand simulations is a confidence costume. It's one model run repeatedly with different random seeds. The number measures the model's certainty, not its accuracy, and a pick is only a pick against a price. I'll take Arizona of the two. A model that lands on the Cardinals has a chance of being early on a roster the market hasn't settled. The New England pick is the one I'd demand a number for, because a simulation's edge against an efficiently priced team is usually just noise wearing a lab coat. Until someone shows me the line, "backs the Patriots" is decoration.
Here's what ties the day together. All three clubs are being described by something other than their own production. Denver is described by a television argument about its quarterback instead of a defense that just produced five sacks and two interceptions. New England and Arizona are described by a simulation instead of by anything they did on a field. The Broncos are both the story of the day and the least settled team in it. Joseph's defense can carry them into January. Whether Nix can carry them through it is the question nobody on that broadcast answered, and the next four weeks of Denver's red-zone offense will answer it louder than any panel can.