What is a snake draft?
A snake draft reverses the pick order every round. Round 1 runs from slot 1 to slot 12; round 2 runs from slot 12 back to slot 1; round 3 runs forward again. The order zig-zags down the board, which is where the name comes from. The alternative — a straight or linear draft, where slot 1 picks first in every round — is rare in fantasy football because it hands one manager a permanent advantage that compounds every round.
The whole point of the reversal is compensation. If you pick last in round 1, you pick first in round 2. The snake exists to balance the board, and as the arithmetic below shows, over an even number of rounds it balances it exactly.
What does 1.01 mean?
1.01 means round 1, pick 1 — the first selection of the entire draft. The notation is round.pick, always with the pick zero-padded to two digits. So 1.12 is the twelfth pick of round 1, 2.01 is the first pick of round 2, and 4.07 is the seventh pick of round 4.
Two numbers describe the same selection and people mix them up constantly. The round.pick notation tells you where you are inside a round. The overall pick number counts from the top of the draft: in a 12-team league, 2.01 is overall pick 13, and 4.07 is overall pick 43. Draft software and mock-draft tools usually show overall numbers; league chat usually uses round.pick. When someone says "I have the 1.01," they mean they own the first slot in the order. When someone says "I'm on the turn," they mean something more specific — see below.
Where does every slot pick in a 12-team league?
Here is the complete board for a 12-team snake through six rounds, in overall pick numbers.
| Slot | R1 | R2 | R3 | R4 | R5 | R6 | Total |
|---|---|---|---|---|---|---|---|
| 1 | 1 | 24 | 25 | 48 | 49 | 72 | 219 |
| 2 | 2 | 23 | 26 | 47 | 50 | 71 | 219 |
| 3 | 3 | 22 | 27 | 46 | 51 | 70 | 219 |
| 4 | 4 | 21 | 28 | 45 | 52 | 69 | 219 |
| 5 | 5 | 20 | 29 | 44 | 53 | 68 | 219 |
| 6 | 6 | 19 | 30 | 43 | 54 | 67 | 219 |
| 7 | 7 | 18 | 31 | 42 | 55 | 66 | 219 |
| 8 | 8 | 17 | 32 | 41 | 56 | 65 | 219 |
| 9 | 9 | 16 | 33 | 40 | 57 | 64 | 219 |
| 10 | 10 | 15 | 34 | 39 | 58 | 63 | 219 |
| 11 | 11 | 14 | 35 | 38 | 59 | 62 | 219 |
| 12 | 12 | 13 | 36 | 37 | 60 | 61 | 219 |
The formula for any league size. In a snake draft with T teams, the manager in slot k picks at overall number (r-1)×T + k in odd rounds, and (r-1)×T + (T+1-k) in even rounds. That single line reproduces the whole board for 8, 10, 12 or 14 teams.
What is the wheel, or the turn?
The turn is the pair of back-to-back picks at each end of the board, where a round reverses. In a 12-team league, slot 12 owns picks 12 and 13 — the last pick of round 1 and the first pick of round 2, with nobody selecting in between. Slot 1 owns 24 and 25 at the round 2/3 turn. "The wheel" usually refers to slot 12 specifically, and "turning the wheel" means using those consecutive picks to take two players you want without risking one of them to another manager.
The turn is the real structural difference between slots, and it shows up in the gaps between your picks:
- Slot 1: gaps of 23, 1, 23, 1, 23 picks. Elite first pick, then a very long wait, then two selections in a row.
- Slot 12: gaps of 1, 23, 1, 23, 1. Two selections immediately, then the longest waits on the board.
- Slot 6: gaps of 13, 11, 13, 11, 13. No back-to-back picks and no long droughts.
The ends of the board are feast-or-famine. The middle is a metronome. That is a genuine difference in how a draft feels and how you have to plan for it, and it is not a difference in how much you get.
Does any slot get more picks than the others?
No. In a snake draft, after any even number of rounds, every slot holds exactly the same total pick capital. Look at the Total column above: every slot sums to 219 after six rounds. It is 25 after two rounds, 98 after four, 388 after eight, and 870 after all twelve.
The mechanism is simple. Rounds pair up. Slot k takes position k in an odd round and position T+1-k in the following even round, so each pair of rounds contributes a constant to every manager's total. The snake is not approximately fair over even rounds; it is exactly fair.
The one honest caveat: an odd number of rounds does leave a small spread, because the final unpaired round runs forward. Most leagues draft 15 or 16 rounds. In a 15-round, 12-team draft, slot 1's pick numbers total 1,352 and slot 12's total 1,363 — a spread of 11 across a draft of 180 selections, concentrated entirely in the last round where you are choosing between backup tight ends. It is real, and it is trivial.
What is third-round reversal, and what problem does it solve?
Third-round reversal (3RR) does not reverse at round 3 in the usual sense — it runs round 3 in the same direction as round 2, so slot 12 picks first twice in a row, then the draft resumes normal snake alternation. Ian Allan writes the sequence as 1–12, 12–1, 12–1, 1–12, 12–1.
| Slot | R1 | R2 | R3 | R4 | Total |
|---|---|---|---|---|---|
| 1 | 1 | 24 | 36 | 37 | 98 |
| 2 | 2 | 23 | 35 | 38 | 98 |
| 3 | 3 | 22 | 34 | 39 | 98 |
| 4 | 4 | 21 | 33 | 40 | 98 |
| 5 | 5 | 20 | 32 | 41 | 98 |
| 6 | 6 | 19 | 31 | 42 | 98 |
| 7 | 7 | 18 | 30 | 43 | 98 |
| 8 | 8 | 17 | 29 | 44 | 98 |
| 9 | 9 | 16 | 28 | 45 | 98 |
| 10 | 10 | 15 | 27 | 46 | 98 |
| 11 | 11 | 14 | 26 | 47 | 98 |
| 12 | 12 | 13 | 25 | 48 | 98 |
The problem 3RR is meant to solve is that the 1.01 is worth more than its pick number suggests. In a year with one clearly dominant player, slot 1 gets an asset nobody else can match, and no amount of pick-number arithmetic compensates for that. 3RR pays the back of the board with an earlier third-round pick.
Two things are true about it at once, and most explainers only give you one. First, over the first three rounds 3RR does not flatten the board — it mirrors it. Under a plain snake, first-three-round pick totals run from 50 at slot 1 up to 61 at slot 12. Under 3RR they run from 61 at slot 1 down to 50 at slot 12, the exact reverse. By round 4 both formats are level again at 98. Second, published projections still show 3RR narrowing the gap in real value: Ian Allan's July 2014 analysis for Fantasy Index, using scarcity-adjusted PPR projections over ten rounds, had slots 1–3 accumulating 514–525 points and slots 10–12 only 481–484 under a standard snake, while under 3RR slot 1 fell from 525 to 507, slot 12 rose from 484 to 502, slots 2–4 landed at 500, 501 and 485, and slots 5–11 all landed between 492 and 498.
Both statements hold because pick 1.01's marginal value exceeds what its position implies. That study is from 2014 and used one season's projections; treat the direction as informative and the exact point totals as historical. Sleeper supports 3RR natively, and FantasyPros' Draft Wizard offers it as a draft type.
Is there a best draft slot?
Not in any stable, provable way. The pick capital is identical, so "best slot" is entirely a question about the shape of a given season's player-value curve and your league's scoring format — which changes every year.
You will find confident numbers claiming the first three slots win 20–30% more often, or that a top-three pick carries a "114% VORP advantage." Those figures circulate widely without a named sample size, year range, league set, or method, and the sites they are attributed to do not publish them. Do not build a league rule on them. There is also a common search trap: analyses of whether NFL draft position predicts player production are a different subject entirely and say nothing about your fantasy slot.
The more telling evidence is that the outlets which do rank slots each year disagree with each other and with themselves — different slots come out on top in different scoring formats and different seasons. That instability is the answer. Some real data does exist: RotoWire publishes win rate by draft slot from best-ball contest results, which is worth reading directly rather than through a summary.
What you can say without inventing anything:
- Every slot holds the same pick capital over an even number of rounds.
- Slot 1 gets the single best available player; slot 12 gets two players in a row at every turn.
- Middle slots never wait more than about 13 picks, which makes tier-based drafting easier to execute.
- Which of those is most valuable depends on whether this particular season has a dominant top tier or a flat one.
How should your league decide who gets which slot?
Given that no slot is provably best, the thing that actually matters is that the assignment is fair and that everyone believes it was. Those are two separate problems. Reverse standings, weighted lotteries, in-person games and randomizers all distribute slots differently and fail in different ways — reverse standings, in particular, pays managers to lose late-season games. We compare the options honestly on the draft order methods page.
The belief problem is the harder one. Most randomizers, including the built-in tools in league platforms, are pressed once by one person in private, and the button that runs the draw is the same button that re-runs it. If your league has ever argued about this, the fix is a commitment published before the draw, which we walk through in how to prove your draft order wasn't rigged. Draft Knockout is one implementation: it derives the order from a seed and your roster, lets you post a SHA-256 hash of those inputs to league chat beforehand, and produces a share link that replays the identical result for everyone. FantasyPros' generator and other free randomizers will produce an order faster if verification is not something your league cares about.
Whichever method you use, write it into the league constitution before the season rather than the week of the draft, announce the time so nobody is absent, and reveal last pick first — 1.12 up to 1.01 — so the announcement people actually care about comes last. Then run the draw once, in front of everyone.