Week 2 runs Thursday, Sep 17 through Monday, Sep 21, 16 games, lineups locking Thursday, Sep 17 at 8:15 PM ET with the Detroit Lions visiting the Buffalo Bills. We have 337 players projected — per-player DraftKings-point projections built from opportunity volume, per-opportunity production, and opponent-defense context, walk-forward validated on the 2025 season week-by-week from prior data only, beating the season-mean baseline on both MAE and rank correlation. Draftables with no projection are players with zero or near-zero NFL game history in the model's play-by-play store — rookies and first-year players the model has nothing to project from; they are omitted rather than assigned invented numbers, which historically accounts for 12–15% of Week-1 usage players. One more thing worth saying plainly: this column is projections and lineup construction, full stop — every number here is a DFS tool, and the lineup is yours to build.

The top of the projection board is led by a pair of running backs who dominate their offenses, a quarterback in the week's best environment, and a wide receiver whose salary still hasn't caught up to his role. Here are the eight names that anchored our projections at the top of the slate.

Robinson and Henry are the two running backs you'd build around this week. Robinson projects at 24.5 DK points (floor 10.4, ceiling 40.2) at $8,200 — a salary that reflects his usage but doesn't price out the ceiling. Henry at $7,200 is $1,000 cheaper with a nearly identical ceiling of 40.8, and both live in top projected stacks.

Allen is the week's top quarterback projection at 24.2 points and a 42.2 ceiling — he's in the Thursday night game with the highest market total on the slate, and his floor of 8.1 is the honest reminder that even the top projections carry variance. Gibbs at $8,500 is the priciest name here, and his floor of 3.7 is the widest downside band on the list — the ceiling of 42.8 is real, but this is a player whose range of outcomes is genuinely broad.

Lawrence at $5,800 producing a 22.0 projection is the clearest efficiency story at the top of the board. Purdy and Williams both project in the 20s at salaries that leave cap room; Nacua at $7,900 leads all wide receivers at 21.8 projected with a floor that stays above 9.

Value per $1K measures projected DraftKings points per $1,000 of salary — it's how you find the players who free up cap room to pay for the cores above. A higher number here means more projected production relative to what you're spending. A note before the names: 435 rosterable players on this slate carry no projection because they have no NFL game history in our play-by-play store. A cheap draftable with no number attached is a guess. It is not a value play.

A.J. Brown leads the value tier at 4.43 per $1K from a $3,000 salary — that's minimum price on DraftKings, which means he gives you maximum salary flexibility at almost no cost. His floor of 0.0 is real: the model's p10 scenario is essentially nothing. His ceiling of 34.3 is also real, and at $3,000 the risk-reward is worth understanding. Lawrence doubles as the slate's best value quarterback at 3.79 per $1K — he appears here and in the core plays because a 22-point projection from a $5,800 salary is genuinely rare.

Mahomes at $5,500 projects 19.1 points (3.47 per $1K) — that salary feels understated for a player with his ceiling of 32.9 and a floor of 7.0, and he's in a game the market prices as a competitive environment. Winston at $4,000 has the widest ceiling-to-salary ratio of any quarterback in the value block at 36.1, with the honest caveat that his floor of 0.8 reflects how lumpy his outcomes can be. Shough and Young both project in the high teens from sub-$5,500 salaries, enough efficiency to pair with a high-priced core.

Game Environments

The market has posted totals on all 16 of 16 games this week. Implied team totals below are arithmetic on the market's own posted total and spread — they belong to the market, not to us, and they're presented as game-environment context for DFS stacks, not as predictions or advice.

DET @ BUF is the week's highest-total game by a significant margin — the market prices it at 53.5, four points clear of the next game on the list. That's the environment context for Thursday's lock window: two teams both with implied totals above 24. WSH @ DAL at 50.5 and MIN @ CHI at 49.5 round out the top three. Below those three, the next tier — IND @ KC and NYG @ LAR, both at 48.5 — still represents good passing-game environment.

Team Stacks

The five stacks below are the top-4 projected players per team by combined DraftKings points, with a Monte Carlo ceiling drawn from each player's own measured volatility. These are correlated lineups, not four independent projections added together.

The DET stack at 73.0 projected and $28,100 salary is the week's most expensive four-player construct, and it's sitting inside the slate's best scoring environment. The ceiling of 134.9 is slightly below Cincinnati's 135.3, but the Thursday lock means any DET stack requires a decision before you see anything else play out on Sunday. That's the tradeoff.

Baltimore's stack has the highest ceiling on the board at 143.7 and the lowest price of the top three at $25,200 — with Henry and Jackson as the engine, Flowers as the air complement, and Huntley providing an interesting secondary piece. The opponent implied total of 21.9 for New Orleans means Baltimore is expected to be comfortable here, which can be a positive for run-heavy stacks that don't depend on trailing-game volume.

Cincinnati's stack is notable for its opponent implied total: the market prices Houston at 38.0 in the CIN @ HOU matchup, the highest opposing team total for any of the five stacks. That's a game-script environment that could push Burrow to throw in meaningful volume regardless of game flow. Atlanta at $23,400 is the cheapest four-player stack here and projects 65.9 with a ceiling of 127.2 — room to add a higher-priced finisher from another game.

Two angles stand out as the week's clearest DFS starting points, and neither requires a leap of faith.

The first is the Thursday night game environment. DET @ BUF is the week's highest-scoring market total at 53.5, and both the Lions and Bills project stacks with ceilings above 130 DK points. The DET stack leads the board at 73.0 projected and the BAL stack is a close complement. The complication is real: lineups lock Thursday at 8:15 PM ET, and any player you roster from this game is committed before a single Sunday snap. That's not a reason to avoid the game — it's a reason to be deliberate about how much of your lineup you're anchoring in one Thursday contest.

The second angle is quarterback value. Lawrence at $5,800 and 22.0 projected (3.79 per $1K) is the cleanest efficiency gap at the position this week. Mahomes at $5,500 and 19.1 projected is a reasonable alternative if you want slightly more salary relief with a strong floor of 7.0. Both give you room to spend up at running back — and this is a week where spending up at running back, toward Robinson or Henry, is a defensible construction choice given where their projections sit relative to their salaries.

These are projections with measured floor-to-ceiling bands, not certainties. The model is walk-forward validated, it has a track record on the 2025 season, and it will still be wrong on individual players with regularity — that's what the bands are there to communicate. Where you set the lineup is your call, and these numbers are the starting point.