1:00 lock, 2:55 answer: the injury call a Sunday slate forces blind

A DraftKings NFL main slate locks once, at the kickoff of its earliest game. Every player in every later game — the 4:05s, the 4:25s, whatever the week's international window happens to be — is already locked into your lineup before his own game has started. That much is obvious to anyone who has ever built a Sunday slate. What's less obvious, because nobody usually does the arithmetic, is that the NFL's official inactives list for those later games doesn't arrive until after that lock too — sometimes by close to two hours. The information you'd want to make the call is generated on a clock that starts running after your decision is already final.

Week 3's research store did the arithmetic once, for one player, because the game he was in happened to make the gap too large to miss.

The math on one receiver

BAL@DAL kicked off from Maracanã Stadium in Rio de Janeiro at 4:25pm ET — the week's international game, with Dallas designated the home team. It carried this slate's highest total, 53.5, and its own point spread, BAL -3.5, both settled since midweek. The one thing that hadn't settled by kickoff was Ravens receiver Zay Flowers, who sat out practice Wednesday and Thursday on a hamstring, came back for limited work Friday, and was still Questionable — not ruled out, not cleared — heading into Sunday.

Here's what a research pass running the morning of kickoff, six hours before the earliest games, worked out about when that would actually be known:

The 1:00pm games' inactives post 90 minutes before their kickoff, comfortably ahead of the slate lock — that's the version of "know before you build" everyone assumes the format guarantees. BAL@DAL's own inactives, 90 minutes before its kickoff, land at 2:55pm ET: nearly two hours after the whole main slate has already locked. Whatever Flowers' status turned out to be, nobody building a same-slate lineup around the 1:00pm lock was going to have it in hand when they built.

That's the finding a same-day beat-writer scan wrote down plainly: "a reader focused on the slate's nominal 1:00pm kickoff might expect" Flowers' word to land with everyone else's. It doesn't. The scan's own game-script pass, run the same morning, named the consequence for a builder directly: with the pool showing him Questionable at a $6,900 salary, a 14.8-point blend projection and 6.5% house ownership, and the market split three ways behind him — one source bullish at 23.9 points with no tag at all, one projecting him at zero, one holding a stale Questionable read — "a Flowers/Bateman build decision has to be made blind or held to a late swap." Rashod Bateman was the name that absorbed Flowers' target share if he sat; which of them to roster, in the game carrying the slate's single highest total, was the actual decision this gap forced.

The research store never learned which way it broke. Its last scheduled check on this slate ran that Sunday morning; the next scheduled event was a new slate's first pass. That's not a gap in the research — it's an honest description of what a same-day scan, run on a normal schedule, cannot reach.

This isn't an international-game problem

Rio makes the story easy to tell because someone can point at a single, large, oddly-timed gap and say "look, almost two hours." But run the same 90-minutes-before-kickoff rule on an ordinary Sunday without a single international game on it. A main slate that includes the 4:05pm and 4:25pm windows — which almost every DraftKings NFL main slate does — locks at 1:00pm ET and gets its late-window inactives at roughly 2:35pm and 2:55pm ET. Same 1-to-2-hour gap. Same "already locked" problem. Every single week, for every player in an afternoon-window game, not just the one international matchup that happens to make the math legible.

What Rio actually changes is not the mechanism, it's the visibility of it. A 4:05pm kickoff still feels like it belongs to "today's slate" in a way that makes the gap easy to forget about. A 4:25pm kickoff from Brazil, on a team whose fans expect it to be a normal 1:00 game, makes the surprise the headline instead of the arithmetic — which is exactly why this pass wrote the arithmetic down instead of assuming it.

What GameScript's own lock model has to do with it

This repository's own DraftKings CSV handling parses one consistent kickoff per game and "derives the slate lock from the earliest kickoff" — the same rule DraftKings itself runs. That's correct and it's not the gap; the gap exists on DraftKings' platform before any tool touches it. What's specific to a tool is what happens after that lock. DraftKings' own late-swap feature would let an entrant replace Flowers, or any afternoon-window player, right up until his individual game starts — a manual move, on DraftKings, after the fact. GameScript does not support late swap. A build made here locks at 1:00pm and stays that lineup; reacting to a 2:55pm inactives list for a 4:25pm game is something a user would have to do by hand, directly on DraftKings, outside anything this product exports.

That's a real, stated limitation, not a hedge — worth knowing before you're the one holding a Questionable tag on the slate's highest-total game an hour after your export already locked.

The rule

A "Questionable" tag doesn't mean the same thing on a 1:00pm player and a 4:05/4:25pm player, even though it's the identical two words on the identical CSV column. For the early-window player, you'll have his real answer — active or inactive — before you have to decide anything. For the late-window player, you won't. You're not weighing a probability against a known outcome you'll react to; you're locking in a probability and finding out afterward, unless you're prepared to manage a manual swap directly on DraftKings in the ninety minutes between his inactives list and his kickoff.

That argues for treating exposure to an unresolved afternoon-window tag as a real, sized risk rather than a coin flip you'll get to correct — concentrate less of a build's roster there than the projection alone would justify, or accept the exposure deliberately, in multiple entries, rather than by default in one. It does not argue for avoiding late-window games generally; BAL@DAL was this slate's best environment by total, and ducking it isn't costless either. It argues for knowing, before you build, which of your questionable tags are going to answer themselves before your decision and which ones are going to answer themselves after.