Your league's rankings are stuck in 2019.

While they print out someone's top-150, Ball Ranks puts the latest AI models to work on real NBA history — and replays every prediction against 15 completed seasons before it touches your board.

The typical rankings site

Last season's averages, reheated
One ranking for every league type
Gut-feel 'expert' tiers
No way to ask about your own roster

BallRanks

Latest AI models projecting 2026-27
Backtested against 15 completed NBA seasons
Punt-aware boards for your exact build
A draft assistant that reads your roster
Our own fantasy basketball GPT

Upgrade to AI rankings — free

Already ranked: Gilgeous-Alexander · Gilgeous-Alexander · Wembanyama · Wembanyama · Jokić — and every other projected NBA player

2026-27 fantasy basketball AI rankings: the projected top 15

These are Ball Ranks' AI projections for 2026-279-category fantasy basketball, not last season's finish. The full board covers 496 NBA players and every one of them has a page with the projected line and full season history.

Ball Ranks 2026-27 projected 9-category fantasy basketball top 15, per game.
Rank	Player	Team	PTS	REB	AST	STL	BLK	3PM	FG%	FT%	TO
#1	Shai Gilgeous-Alexander	OKC	31.5	4.7	6.5	1.6	0.9	1.8	53.8%	88.3%	2.2
#1	Shai Gilgeous-Alexander	OKC	31.5	4.7	6.5	1.6	0.9	1.8	53.8%	88.3%	2.2
#2	Victor Wembanyama	SAS	24.6	11.4	3.3	1.1	3.3	2.1	49.8%	82.5%	2.7
#2	Victor Wembanyama	SAS	24.6	11.4	3.3	1.1	3.3	2.1	49.8%	82.5%	2.7
#3	Nikola Jokić	DEN	25.2	11.5	9.2	1.4	0.7	1.6	57.1%	82.0%	3.1
#3	Nikola Jokić	DEN	25.2	11.5	9.2	1.4	0.7	1.6	57.1%	82.0%	3.1
#4	Luka Dončić	LAL	31.7	8.0	8.2	1.6	0.5	3.9	47.0%	78.3%	3.8
#4	Luka Dončić	LAL	31.7	8.0	8.2	1.6	0.5	3.9	47.1%	78.3%	3.8
#5	Tyrese Maxey	PHI	25.7	3.6	6.0	1.6	0.7	2.9	45.4%	87.9%	2.2
#5	Tyrese Maxey	PHI	25.7	3.6	6.0	1.6	0.7	2.9	45.4%	87.9%	2.2
#6	Joel Embiid	PHI	25.8	7.7	4.0	0.7	1.1	1.3	49.2%	86.2%	2.9
#6	Joel Embiid	PHI	25.8	7.7	4.0	0.7	1.1	1.3	49.2%	86.2%	2.9
#7	Kawhi Leonard	LAC	21.9	5.2	3.0	1.5	0.4	2.0	50.6%	87.7%	1.6
#7	Kawhi Leonard	LAC	21.9	5.2	3.0	1.5	0.4	2.0	50.6%	87.7%	1.6
#8	Chet Holmgren	OKC	16.9	8.7	1.9	0.6	2.1	1.4	54.0%	79.6%	1.7

See the full 2026-27 draft board

Who built this board, and how do we know it is any good?

Ball Ranks' projection model is built by AI: frontier AI models propose changes to how players are projected, each one is replayed against 15 completed NBA seasons, and only the changes that hold up are kept. Changes that failed are kept on the public record next to the ones that worked, so the board is auditable rather than a black box.

That AI-built board was replayed against 8 archived preseason ranking boards from ESPN and Yahoo (2021-22 through 2025-26), covering 1,438 matched player comparisons. Across all of them the average size of a prediction miss, weighted so misses on players who actually mattered count more, was reduced by 8.93% against those boards.

Against ESPN's archived boards specifically (784 player comparisons), Ball Ranks reduced that weighted miss by 13.58%. Against the Yahoo Sports boards (654 comparisons) it was 3.56% worse, which we publish too.

Every projection-model change is replayed against 15 completed NBA seasons (2011-12 through 2025-26) before it ships, with a few seasons held back during research and used only to double-check the final candidate.

Read the backtests