If you've been reading me for the last week you know I've already gone deep on SportsLine's 2026 projections once or twice, but this version of the piece has something the earlier runs didn't: the Daniel Jones receipt. SportsLine simulated the upcoming NFL season 10,000 times and the sales pitch here isn't the simulation itself, it's the track record. The model flagged Jones in a year when basically nobody else did, and that kind of call buys you credibility for a few seasons.

I've been burned by models before. I've also been burned by my own eyeballs. Ray Delgado loves reminding me of the time I faded a receiver because his target share math looked ugly and he promptly went off for 150 and two scores. Fine, Ray. You get one. But the whole reason I keep coming back to simulation-based rankings is that they don't fall in love. They don't draft a guy because of one highlight tape or because a podcast host yelled his name in June. They just run the scenarios and spit out outcomes.

That matters most in August, because this is peak hype season. ADP is moving daily, sleeper lists are everywhere, and I genuinely believe fantasy football has a hype escalation problem where one outlier season gets a guy pushed up two rounds past anything his usage profile supports. I wrote that last week and the model's bust list backs it up. The names being drafted off career-best efficiency years are the ones the simulation is coldest on, and if one of them is on your cheat sheet in round three, you should at least know the model disagrees with you.

As for actually using this thing, I keep preaching the same point: it's a tiebreaker, not a substitute for understanding the game. When I'm stuck between two players at the same ADP, the 10,000-run simulation is a great tiebreaker. When it tells me a player's floor is higher than the market thinks, that's actionable, because early season fantasy is about floors, not ceilings. Same logic I use for streaming pitchers in March and streaming quarterbacks in September. The ceiling guys are how you lose weeks in the first month. The floor guys are how you survive them.

The sleeper side is where I'd lean hardest on the model right now. Sleepers are, by definition, players whose value the crowd hasn't caught up to yet, and the crowd moves slowest in August before preseason usage data fully circulates. If the simulation likes a guy whose ADP sits three or four rounds below his projected finish, that gap is free expected value, and closing gaps like that is the entire job. You're not trying to be smarter than your league, you're trying to notice gaps faster than your league.

The breakout calls are trickier. A breakout usually needs a role change, a new coaching scheme, a departure ahead of someone on the depth chart, something structural. The model can weight those variables but it can't watch practice. So when a breakout pick lines up with something I can verify, like first-team reps or route participation in preseason games (which, again, is where real preseason value lives, not garbage-time box scores), that's when I buy in fully. Model says X, my eyes confirm Y, roster the player.

The busts are the least fun and the most useful. Nobody wants to hear that a player they've circled is overvalued. But a bust call in August saves you a round-three pick, and a saved round-three pick is worth more than any sleeper you'll hit in round ten. Check the bust list against your draft plan before you walk into the room. If you're still going to take the guy, fine, at least it's an informed disagreement instead of a blind one.

I'm logging all of this. When I covered the earlier version of these projections I said I'd track the accuracy of the model-backed calls throughout the season, and that stands for this batch too. Every sleeper, breakout and bust from this list goes in the ledger, and I'll publish the receipts whether the model eats or gets humbled. Anybody can make predictions in August; the whole point is showing your work in December.

So go read the full list, argue with it, then tell me which model call you think is most wrong. I'll pick my favorite disagreement and revisit it midseason. My money's on at least one of you being very loud about the Daniel Jones guy and very quiet about it come November.