Snake Draft Board

Why players rank where they do

Every category counts, including the ones a player hurts. A star who helps you in five categories but costs you heavily in three can rank below a less famous player with no real weak spots. That's why Giannis, whose free throws, threes and turnovers cost a lot, sits well below his ADP, while Maxey, strong in five categories with only small weak spots, is a fairly clear top-five pick.

Percentages count by volume. A .640 free-throw shooter taking eight a game hurts you far more than one taking two.

Head-to-head is a fresh contest every week. Categories that swing a lot from week to week, like steals, are less reliable to win in any given week, so value built on them counts for less head-to-head than in roto. Steadier categories, like assists and blocks, hold up better.

Your team changes the order. As your roster fills, a player who adds what you're short of moves up and one who adds more of what you already have moves down, so the ranking reflects your team and the teams you're drafting against, not a fixed list. Tap Why next to any player for his own breakdown.

Scoring and roster

These settings in Setup change how players are valued. Active and bench slots, lineups and the games cap sit under More league settings. The defaults are Yahoo, roto, 9-cat, 10 active and 3 bench slots, daily lineups and no games cap.

• Platform. Sets the lineup slots and, when you pick it, that platform's defaults. Yahoo: PG, SG, G, SF, PF, F, C, C and Util, 9-cat. ESPN: PG, SG, SF, PF, C, G, F and Util, 8-cat (no turnovers). Both default to 10 active and 3 bench slots, daily lineups and no games cap; change any of them afterwards. Other (Fantrax, Sleeper and the rest) uses ESPN-style slots and leaves your settings alone. Projections and ADPs are Yahoo's on every platform, so a room drafting from another site's rankings may take players at different picks.
• Categories: 9-cat or 8-cat. 9-cat counts FG%, FT%, 3PM, PTS, REB, AST, ST, BLK and TO. 8-cat drops turnovers, so turnover-heavy playmakers stop paying for them. Gains, chances and standings then count 8 categories, with TO shown as —. Match it to your league: in 500 simulated snake drafts scored on real 2025-26 weeks of an 8-cat league, the 8-cat setting won +0.14 ± 0.03 more categories per weekly matchup and about 4% more matchups than leaving it on 9-cat.
• Scoring: Roto, H2H: each category or H2H: most categories. Roto scores season totals, so a small lead in a category is usually safe and extra strength in a category you already win is wasted. Head-to-head is a fresh contest every week, so the board adds each category's week-to-week swing (measured from three real seasons, 2023-24 to 2025-26: drafted-style teams playing every real week with daily lineups, injured players covered from free agents. Steals, points and FG% swing the most relative to how far teams differ; points, rebounds, assists and threes swing together because a team's weekly game count varies from 2 to 5). Close category edges then count for less, and strength in a category you lead keeps paying because you still lose some weeks. Gain now, the model pick order, the category chances and the projected standings all switch to weekly odds. Scored on real 2025-26 weeks (real box scores, schedule and absences, daily lineups) in 240 simulated drafts (10–14 teams, every slot), head-to-head mode won +0.08 ± 0.03 more categories per weekly matchup than roto mode, about 1.5 more category wins over a 20-week season, so match the setting to your league. H2H: most categories gives one win, loss or tie per week, to whoever takes more categories. The board then values your chance of winning the week instead of adding up categories: a category counts for more the likelier it is to decide your matchup, so one you win or lose almost every week counts for little and punting a category becomes a real plan. Gains stay in category-sized units (scaled by how often a single category decides a week); the forecast adds your weekly matchup win chance, and projected standings show expected wins in a week where you play every other team once. Tested on real 2025-26 weeks in 500 simulated snake drafts, it won exactly as many weekly matchups in most-categories leagues as each-category mode (0.000 ± 0.003), so either setting works there. In the same test, committing to a punt from the first pick lost matchups even in most-categories leagues (about 6% fewer when punting FT%, 1% fewer punting TO).
• More league settings: active slots, bench slots, lineups and games cap. Active plus bench slots is your roster size, so it sets how many players each team drafts or buys. If your league caps the games each active slot can count per week, set that cap too (leave it on No cap otherwise). With a cap, a missed game often costs less than a full game, because a bench player fills in or the cap would have cut a game anyway, and the split between active and bench slots decides how much. The board works it out by simulating weeks of NBA schedules (daily lineups, each player missing 15% of games): with 10 active slots and a cap of 4 per slot, a missed game costs about 0.55 of a game with 3 bench and 0.4 with 4, at 3 per slot about 0.1, and at 5 per slot about 0.9. Without a cap every game counts in full; daily lineup limits alone waste only about 1% of a 10 + 3 roster's games. A covered game is played by a bench-level player, so each player gets that bench player's per-game output for the covered share of the games he's projected to miss. Injury-prone players who are strong per game rise and durable role players fall. The typical player at each pick shifts the same way. Set it if your league has one: in 400 simulated snake drafts scored on real 2025-26 weeks of a league capped at 3 games per slot, the cap setting won +0.26 ± 0.02 more categories per weekly matchup and about 6% more matchups than leaving it on No cap. The 15% miss rate is an average rather than each player's own, and the player list holds 300 players, so very large leagues will log some late picks by name. Lineups: with weekly lineups the starters are set before the week, so a bench player can't stand in when a starter sits. The board simulates that too: a 10 + 3 roster loses about 12–13%% of its games that way (against under 1%% daily), and a missed game costs a full game unless a cap would have cut it anyway. The loss is similar for most rosters, so it shifts values less than it sounds.
• News adjustments. Unticked (the default), every player uses Yahoo's projections, which is what mock drafts are scored with, so leave it off for mocks. Ticked, a handful of players with dated injury or role news that Yahoo's numbers don't yet reflect use an adjusted line instead, marked "news" in the list (hover for the reason). The News panel in Setup lists every adjusted player and why. Shorter absences scale every counting stat by the games lost; percentages stay the same. Only dated injuries and role changes a coach has confirmed count; stated plans, such as cutting a player's minutes, don't. Use it for real leagues.
• Lineup positions. Ticked (the default), the board counts your platform's lineup slots (Yahoo with 10 active slots: PG, SG, G, SF, PF, F, C, C and two Util; ESPN: PG, SG, SF, PF, C, G, F and three Util). Each day only players who fit the slots count, so a roster short at a position wastes games. The board simulates days of NBA schedules, fills the lineup as well as the slots allow, and measures the share of games lost; picks or buys still to come are assumed to fill whatever slots you're short of, so a gap only counts once your remaining picks can't cover it, in practice the last few rounds. A player's value then includes how much he cuts that loss for your roster. Most rosters lose under 1% and barely notice. A roster with one C-eligible player loses about 2%, with none about 4%, and with no SF, PF, PG or SG about 2%, so the setting mostly steers you towards a second center late in the draft if you're still short (ESPN's single C slot makes centers matter less: no center loses about 1.5%). Projected standings include each team's loss too. In 300 simulated snake drafts per scoring mode (10–14 teams, opponents drafting by ADP), scored with each team's lineup loss, ticking it gained +0.29 ± 0.19 standings points in roto mode and +0.030 ± 0.008 categories per weekly matchup in head-to-head mode, and cut the lineup loss from 0.5–0.9% to about 0.3%. An earlier, stronger version that treated your remaining picks as random positions (and so pushed centers up from the first rounds) gained about twice as much in the same drafts; it was toned down so positions only decide picks near the end. Untick it for mock drafts, where positions don't count.