Bye-week planning

The mechanic, first

There is no basketball analogue for this at all, so start from zero.

Each NFL team plays 17 games across an 18-week calendar. The extra week is a bye — one scheduled week where the team simply does not play. Every team gets exactly one, at a week the league assigns.

A player on bye scores zero. Not a low score, not a rest day, not a decision by his coach — there is no game. He is perfectly healthy, and he produces nothing, and you knew about it in August. Nine starting slots must be filled every week from your sixteen roster spots, and if you have not planned, you fill them with whoever is left.

That is the entire mechanic. Everything else here is arithmetic about bench space.

Why it is genuinely tight

You start 9 players: 1 QB, 2 RB, 2 WR, 1 TE, 1 FLEX, 1 D/ST, 1 K. You have 7 bench spots and 1 IR slot.

Those seven benches are not spare capacity — they are already doing three jobs. They hold your handcuffs (the backup to a bell cow, worthless until the moment he is not — see Player archetypes), your speculative adds off the wire, and your injury cover. Bye-week coverage is the fourth claim on the same seven spots, and it is the only one you can see coming from four months away.

If five of your starters share a bye, that week you are starting five bench players, and you will lose. One loss out of fourteen regular-season games, in a league where six of ten make the playoffs.

The wire is deeper in a 10-team league, so a one-week fill-in is usually easier to find than it was in a 12-team room. That makes isolated byes manageable; it does not make a concentrated bye cluster harmless. Use that depth to replace one missing starter, not to excuse drafting several players who disappear together.

The rules

State them as rules, apply them at the draft, and do not re-derive them in September.

1. No more than two high-value players sharing a bye. By high-value, mean roughly your first five picks. Two of them off in the same week is survivable — you fill two slots from the bench and take a below-average score. Three is a week you have essentially forfeited.

2. No more than about five common-bye players overall. Across the full 16-man roster. Past five, the bad week stops being a bad week and becomes an unfillable lineup, because your bench is simultaneously on bye too.

3. Early byes hurt more than late byes. This is the non-obvious one, and it is the one worth remembering. A bye in week 13 or 14 is cheap, because by then the waiver wire still holds real players — injuries and role changes have shaken loose starters all season and somebody useful is always available. A bye in week 5 or 6 is expensive, because in early October the wire is picked clean: nobody has been dropped yet, no roles have changed yet, and the free agent pool is exactly the 51 players nobody drafted. Same zero on the scoreboard, very different cost to replace. Combine this with Waivers: your access to the wire is worse the better your team is doing, so an early bye on a good team is the hardest version of this problem.

The 2026 bye weeks

Byes run from week 5 to week 11, then skip week 12, then finish in weeks 13 and 14. Weeks 1–4 and 15–18 have none — which incidentally means the fantasy playoffs, weeks 15–17, are bye-free.

Counted from the bye column of Data/players.csv and verified against the NFL's official schedule on 2026-08-02 (32 teams, one bye each):

Week 11 is the crunch: six teams off, nearly a fifth of the league at once. No other week comes close. If your roster is at all concentrated, week 11 is where it shows, and it is the one week worth checking explicitly before you leave the draft.

After week 11, five weeks are tied at four teams — 6, 7, 8, 10 and 13. By rule 3 they are not equally bad: week 6 is the worst of them, because it is the earliest, and week 13's four teams are the cheapest four on the board. Weeks 5, 9 and 14 are quiet at two teams each.

Note the shape of it. There is no week where a majority of the league is off, so no bye week is unmanageable in principle. The failure mode is not the calendar — it is a roster that accidentally concentrates in one column of that table.

What this means at the draft

Bye weeks are a tiebreaker, never a reason to pass on a clearly better player.

The order of operations is: build the board on value, then use bye weeks to break ties. If two players are genuinely close — same archetype, same tier, projections within a point or two — take the one who does not stack onto a bye you are already heavy in. That is a free improvement and you should take it every time.

But the moment the players are not close, stop. With the locked 1.01, the first comparison point is the back-to-back 20/21 turn, after an 18-pick wait—not the former 12-team 24/25 turn. Passing on a clearly better three-down back at 20 or 21 solely because he shares a bye with your first-rounder trades fourteen weeks of edge for one week of convenience. That is a bad trade and it is not close. You can solve a bye with a waiver claim, a streamed kicker, or simply eating one bad week. You cannot solve a worse roster.

Two practical additions. Keep an eye on week 11 specifically as you draft, since it is twice the size of a normal bye week. And do not bother planning bye coverage for kicker and defence at all — you stream both by Implied team total anyway, so their byes cost you nothing beyond a two-minute swap.

Open questions

• The counts above are per team, not per roster slot. What actually matters is how many of your starters share a week, which cannot be known until after the draft in late August 2026.
• Bye weeks in players.csv match the published NFL schedule. Flex scheduling can move individual games, but not these bye assignments.
• The "no more than about five common-bye players" threshold is a rule of thumb, not something derived from this league's roster limits. Four may be the safer number given only seven bench spots.