Head-to-head vs roto fantasy basketball: how the same 9-cat board plays in each format

Two category formats, one board

Most category leagues score the same nine stats: points, rebounds, assists, steals, blocks, threes, FG%, FT% and turnovers. What differs is how a season gets decided.

Head-to-head (H2H) categories splits the season into weekly matchups. In the "each category" style, every category is a win, loss or tie, so a week ends something like 5-4. In the "most categories" style, the team that wins more categories takes the week as one win. Either way, each week starts from zero.

Rotisserie (roto) never resets. Every team's season totals are ranked in each category. In a 12-team league, first place in a category earns 12 points and last earns 1, and the standings add the nine.

The 9-cat rankings are per game. That is the right yardstick for how good a player is on the nights he plays. The question this guide answers is how much the format should change what you do with that number.

Games played: the per-game board vs season totals

To see how much games matter, we rescored 2025-26 using each player's season totals instead of his per-game averages. Same nine categories, same z-score method, same top-156 pool. It describes last season as it happened, so read it as a measure of how much a player's value depended on being on the floor, not as a forecast.

Source: Parlae 9-cat board and season-total rescore, Oct 13 2026, from Sleeper's 2025-26 season totals. Games from Sleeper's season totals; status is Sleeper's designation on Oct 13 2026, blank when none is listed.

The pattern is blunt. Fifteen players in the per-game 9-cat top 60 played fewer than 55 games last season, and all 15 fall outside the top 60 on season totals. The other direction is gentler: the twelve biggest risers in the top 60 climb between 10 and 26 places, not 100. Barnes moves from 12th to 6th, Bane from 37th to 13th.

The top of the board barely moves. Jokić is first both ways, and the season-total top five are the per-game top five in a slightly different order: Jokić, Gilgeous-Alexander, Wembanyama, Maxey, Leonard. Dončić drops from 6th to 10th.

Why games hit roto harder

Both formats lose the games a player misses. The difference is where the loss lands.

In roto, a missed game is a permanent hole in a season total. The 40 games Markkanen did not play last season are counting stats your team never gets back in six categories; fewer turnovers is the only thing a missed game gives back. Your replacement fills some of it, at replacement-level production. The per-game board does not see any of this, which is why the season-total rank is the closer guide for roto.

In H2H, a missed game costs you in that week only. A player who misses two weeks in November loses you two matchups' worth of his stats, and every other week is untouched. If those weeks fall in your playoffs, that is the worst case; if they fall in November, you have the rest of the season to recover. The per-game rank is the better guide in H2H, with the games history as the risk you are pricing.

So the same table reads two ways. In roto, treat Davis, Embiid and Butler as the season-total ranks suggest unless you believe the games will come back, and use three seasons of games (Davis played 75 in 2023-24) to judge that. In H2H, you can draft closer to their per-game rank and plan to cover the weeks they miss from the waiver wire. The injury-risk guide has three-season games for the whole top 60.

Percentages: whose value lives in FG% and FT%

FG% and FT% are ratios, not totals. In roto they settle over a full season of attempts. In H2H they reset every week and rest on the handful of games each player gets that week, a small fraction of a season's attempts. A player whose value is mostly percentages delivers it over a season; in any single H2H week, a few cold nights can erase it.

Source: Parlae 9-cat board, Oct 13 2026. The last column re-ranks the full board on the other seven categories. Percentage scores are weighted by attempts.

Two kinds of players come out of this. Durant, Butler, Duren and Reaves rank in the 9-cat top 30 mainly because of percentages: without them, all four fall to 48th or worse. Adebayo, Ball, Mobley and White are the reverse, top 22 on the other seven categories and pulled down by shooting.

Gilgeous-Alexander is the exception at the top. His percentages are worth 4.3, but he is still 6th without them.

Punting: a tool for H2H, a cost in roto

In H2H, punting is a strategy. In the most-categories style you need 5 of 9 to take a week, and in the each-category style a punt costs one result of nine. Either way, conceding one frees every pick for the other eight. The punting guide ranks the board for each punt.

In roto, the arithmetic is harsher. In a 12-team league, the average finish in a category is worth 6.5 points (the middle of 12 and 1). Finishing last in a punted category earns 1, a loss of 5.5 points against average, and you have to win those back across the other eight categories, about 0.7 places each. It can work if the punt sharply improves the other eight, but it is a bet you settle in the standings, not a free category.

Volume and the calendar

Volume matters in both formats in different ways. In H2H, a team with more games in a given week piles up more counting stats that week; the opening week alone ranges from 2 games for some teams to 4 for Philadelphia, as the opening week guide lays out. In roto, what counts is total games across the season, and many roto leagues set a games limit per roster spot. Check your settings: under a cap, extra games past the limit earn nothing.

What to do on draft night

1. Roto: lean on games played. Weigh the season-total rank and three seasons of games for anyone in the per-game top 60 who played under 55 games last season. Full-season players like Barnes, Bane, Alexander-Walker and Bridges move up.
2. H2H: draft the per-game board, and plan to stream through the weeks your injury risks miss. Check your playoff weeks before you take on more than one.
3. H2H: know where your percentages come from. If your early picks lean on Durant, Butler or Duren for FG% and FT%, those categories rest on a few games each week.
4. Roto: build balance before punting. A punt costs about 5.5 standings points against an average finish, before it gives anything back.
5. Points league instead? None of this applies the same way; see points vs categories.