How to pick your captain

You get one captain a week. He scores double, and the bonus counts separately in the standings.

Captain a player who goes for 40 and you bank 40 raw plus a 40 point bonus, 80 points out of one roster slot. Captain a player who gets ruled out an hour before kickoff and you bank nothing at all. There is no vice captain, no auto promotion, no safety net.

Most people pick that slot on instinct. This article is three seasons of real NFL data run through the actual scoring engine, and by the end you will know which positions win the captain week, which players win it, and the one axis nobody publishes that decides whether your captaincy got wasted.

Article 2 told you one player on your roster doubles. This is the article about which one.

Your captain scores 2.0x. The captain bonus is the extra points, and under a flat multiplier that bonus is exactly the player's raw score. A 30 point captain is 30 raw plus a 30 point bonus, 60 from the slot.

Flat means flat. Same 2.0x for a quarterback, a running back, a receiver and a tight end. No position gets its own number.

If your captain does not play, the bonus is 0. Nobody gets promoted in his place.

One more thing worth knowing early, because it makes everything below obvious: the captain bonus is the first tiebreaker in the standings, ahead of fewest hits and ahead of account age. Two managers tied on points get split by who captained better.

And a consequence that saves you a lot of arithmetic. Under a flat multiplier, the best possible captain in any week is just the highest raw scorer that week. There is no positional math to do. That is the whole reason the rest of this article is about ceilings.

The design did not start here, and the way it changed is the most useful thing in this article.

The original spec was position aware: QB 1.7x, RB and WR 2.0x, TE 2.5x. The idea was that every position should win the captain slot about a quarter of the time, and the multipliers would be the dial that got it there.

Then we scored every quarterback, running back, receiver and tight end week from 2023 and 2024 through the production engine. 10,521 player weeks, 1,104 players. At 1.7x, quarterbacks won zero weeks out of 36. Not a few. Zero.

So we swept the quarterback multiplier looking for the number that would get QB to a 25% win rate. It was 2.4x. That is not a refinement of 1.7, it is a different design, and it is a number with no story a manager would ever accept. "Your quarterback counts 2.4 times" just makes no sense.

Here is the mechanism, and it is the part worth stealing even if you never play MTC. Quarterbacks have the highest mean of any position, around 14.5 points per game. They also have the most compressed ceiling, because a passing touchdown is worth 4 points and a rushing or receiving touchdown is worth 6. A 3 passing touchdown game is a common good day and lands near 30 raw. A receiver with 3 receiving touchdowns is at 45 and still climbing.

Captaincy pays for ceilings, not means. Tuning a captain multiplier by average is a trap, and we walked into it before the data walked us out.

There is a smaller confession attached. The pre pass that produced the original recommendation estimated Lamar Jackson at 29 points per game. He was really at 24.4. That single overestimate is most of why the first answer was wrong.

So the multiplier is flat 2.0x. Consistency beat a balance target that could only be hit with a number nobody could explain. What that gives up is real and worth saying out loud rather than dressing up: your quarterback is on your roster for his floor and his slot, and the captain race is mostly a receiver and running back race. We decided that was fine. It is a consequence of the design, not a bug in it.

Three seasons, MTC weeks 1 to 17, pooled 2023 through 2025. 51 weeks.

The pooled row hides the actual story, which is the trend. Running backs won 4 captain weeks in 2023, 5 in 2024 and 7 in 2025. 2025 is the first season running backs took more captain weeks than any other position. Across all three seasons receivers still lead 23 to 16, so this is a trend and not yet a changing of the guard.

One 2025 detail worth a line: Jonathan Taylor won 3 captain weeks by himself, more than any player has won in a single season in the three years we scored.

Now the tight end paragraph, which is the one everyone gets backwards.

MTC pays tight ends 1.5 points per reception against 1.0 for backs and receivers. That premium is Tier 1, it is locked, and it is applied in every single number on this page. It is also not what people think it is. An extra half point per catch lifts a tight end's floor, which is exactly why tight ends get to FLEX competitive scoring in this game. Captaincy is decided by ceilings. Tight ends won 5 weeks out of 51, the fewest of the four positions.

Both things are true at once. The premium is real and it is not going anywhere, and it does not make tight end a captain position. Do not read this as a walk back on the premium, because it isn't one.

One note on the denominator, in the body rather than a footnote, because anyone who recomputes this will check it. The NFL plays 18 weeks. MTC scores 17. One season is 17 weeks, two is 34, three is 51. Both 2024 and 2025 had a running back as the Week 18 optimal captain, so counting out of 54 would have quietly inflated running backs specifically. Every figure above is out of 51.

And the honest caveat about what this table is, these are the best captain in the entire league that week, not the best captain on a roster anybody actually had. Jauan Jennings in 2024 Week 3, Tre Tucker in 2025 Week 3 and Michael Wilson in 2025 Week 14 were not on your team and were not on mine. A version filtered to a realistic roster is deliberately not built yet, because the only price sheet that exists right now covers 65 players, of whom 3 are quarterbacks and 7 are tight ends. Filtering to that would suppress quarterback captaincy for reasons about the sheet being incomplete rather than anything about the game.

Two facts that look like they cannot both be true.

Quarterbacks take 14% of captain weeks, the second rarest of the four positions. And the highest outright captain win rate of any individual player across three seasons belongs to Josh Allen, a quarterback, at 4 wins in 48 played weeks.

Both are true, and the resolution is the reusable idea. There are 32 starting quarterbacks and their ceilings compress against each other, so the position rate is low. The ceiling that does exist is concentrated in about three of them. Allen's 90th percentile week is 38 points, the joint highest in the entire player pool, tied with Ja'Marr Chase.

Captaining a quarterback is usually wrong. Captaining Josh Allen is not.

Before anyone over rotates on that, the counterweight. Lamar Jackson has 1 outright win in 44 weeks. "My quarterback is hot" is not the argument and never was. A specific quarterback with a specific ceiling is.

The general version, which applies to everything else in this article: a positional base rate tells you where to look. It does not tell you who to pick.

Two captains with the same average are not the same bet, and almost nothing you can read anywhere will tell you the difference.

Bust rate is the share of played weeks under 10 points. Think of it as the captaincy wasted rate. Here are five players from the 50 highest equity in the pool:

Look at Keenan Allen and Patrick Mahomes. They are about 3 points apart on captain equity and 42 points apart on bust rate. You get one captain a week and there is no vice. A 49% chance of wasting the slot is a completely different proposition from a 7% chance, and no ranking built on averages will ever surface that.

There is a tidy callback hiding in that column. The lowest bust rates in the pool are almost entirely quarterbacks. Mahomes at 6.5%, Drake Maye at 7.1%, Josh Allen at 8.3%, Jalen Hurts at 8.5%. The quarterback floor advantage is real and it is measurable, and it does not convert into captain wins, because captaincy pays for ceilings. That is the whole paradox sitting in two columns.

This is not a recommendation to captain Mahomes. The point is that average and bust are separate axes, and if you leave this article checking both, you got the value out of it.

Negative results, published, because they stop you acting on folklore.

The December running back effect does not exist. Everyone says run games take over late in the season. We measured position share of top 5 captain slots, weeks 1 to 6 against weeks 13 to 17, across three seasons. Chi square 8.25, 6 degrees of freedom, p = 0.22. Nothing. Running back share actually drifts down, 30.9% early to 22.2% late, which is the opposite of the hypothesis, and that drift is inside the noise as well. The honest conclusion is not "running backs decline in December", it is that 51 weeks cannot detect an effect this size either way. Captain the same way in Week 15 as you do in Week 3.

The tight end line that looks real and isn't. Tight end share of top 5 slots goes 7.2% early to 16.8% mid season. Same underpowered table, same 273 slots. That is a hypothesis for a fourth season, not a finding, and we are not going to dress it up as one.

MTC prices come from ADP. ADP comes from redraft consensus. Redraft consensus discounts quarterbacks, because in a 1 QB league a quarterback is replaceable.

Captaincy does not pay for replaceability. It pays for weekly ceiling. The pricing mechanism and the captain mechanic are looking at two different things, and that gap is where the value sits.

On the 64 players who appear on both the provisional price sheet and the qualified pool, Lamar Jackson at 11.0 returns 20.2 equity and Josh Allen at 12.3 returns 21.9. On equity per dollar, both beat every receiver and every running back on the sheet except one.

That one is Keenan Allen, who tops the whole table at 1.92 equity per dollar and busts in 48.8% of his weeks. He is a cheap way to buy points. He is not a cheap way to buy a captain. Read that table with the bust column or do not read it at all.

Trey McBride belongs in this section too, as the purest version of the floor captain profile in the data: 10.3 on the sheet, 17.6 equity, top 5 in 14.9% of his weeks, and an outright captain win rate of exactly zero across 47 played weeks. He is genuinely good and he has never once been the right captain.

Now the caveat, in the body where you can't miss it. That price sheet covers ADP 1 through 65 and nothing below it, it is provisional, and prices lock on 2026-08-26. This table gets re run against the real sheet then. It is not a buy list, and if it gets quoted back at me in September as one, that is not what it said.

Three rules that survive the data, in the order you should apply them.

And the honest close. This is 51 weeks of hindsight. The numerators are small, a single week moves a player's win rate by about two points, and none of it is a projection. It tells you where to look. It does not tell you who is going to score 40 on Sunday.

One tease before you go: there is an audible that turns your captain into 3.0x for a single week, once a season. Picking the week for it is a harder version of everything you just read. That is next Tuesday.

Reply and tell me: who is your first choice captain in Week 1, and is it a ceiling pick or a safety pick? I read every reply and the good ones end up as material.

If you know someone who would argue with this table, forward it to them. Somebody has to finish last.

Rob, Move The Chains