Simple rules, rich play
A leaderboard of 60 elegant games, ranked by how much play they get out of how few rules.
Some games are called beautiful because a handful of rules opens up a whole world. Go's core rules fit in two sentences, yet they have sustained 2,500 years of strategy. Tetris has seven falling shapes and is still being mastered 40 years on. Game designers call this quality elegance: much depth for few rules. They call the mechanism emergence: strategies arise from how the rules interact, not from content someone designed by hand.
This page ranks 60 such games: board and card games, video games, puzzles, and games of physical skill. Each game is scored 0–10 on six criteria: how simple the rules are, how deep the play is, how many people play it, its reputation, how players rate it, and the beauty of its central idea. Choose a different weighting below to see how much the ranking depends on what you value, and open any game to read the evidence and the judges' reasoning.
| # | Game | Simple rules | Depth | Reach | Repute | Player ratings | Beauty | Score |
|---|
Player ratings come from BoardGameGeek averages or Steam's share of positive reviews (fetched 25 October 2026), scaled to 0–10. An asterisk marks games with no comparable rating site, whose rating the judges estimated. BoardGameGeek users are hobbyists who rate mass-market classics such as Jenga, checkers and Mikado harshly.
How the ranking was made
The whole process was carried out by AI agents (Anthropic's Claude Opus 5.5 and Claude Fable 5.1) on 25–26 October 2026.
- Evidence. Researchers searched the web for each game: its rules, how many people play it, awards and hall-of-fame entries, player ratings, and signs of depth such as competitive scenes, unsolved status and mathematical results. A first list of 35 games was widened to 60 by four researchers who looked for missing candidates among traditional games, modern tabletop games, video games, and puzzles and physical games.
- Judging. Four independent judges scored every game against the same written rubric, each from a different perspective: a game designer, a game theorist, a historian and a lifelong player. The judges did not see each other's scores.
- Challenge and revision. An adversarial critic attacked the combined ranking, and each judge then revised the scores they were persuaded were wrong.
- Fact-checking. Fact-checkers confirmed 334 claims in the evidence and found 24 that were wrong, outdated or unverifiable. The evidence shown here includes their corrections.
A game's score in each criterion is the median of the four judges. The balanced weighting counts depth 25%, simple rules 20%, reach, reputation and beauty 15% each, and player ratings 10%. “Judges” under each game gives the range of the four judges' totals.
How far to trust it
- The top is solid; the middle is not. Go and Tetris lead under every weighting. Third and fourth place are close: the Rubik's Cube stays ahead of chess only through an estimated player rating. From about fifth place down, a point or two is within the judges' disagreement, so read the tiers rather than the exact positions.
- Fame counts twice. Reach and reputation largely measure the same thing, which lifts famous games such as Pac-Man, the crossword and Bejeweled above less-known games with more emergent depth. Try the “Rules, depth and beauty” weighting to remove that effect.
- Physical games are probably under-rated. The critic found that the judges equated beauty with mathematical purity and did not count measurable skill ladders, such as darts averages, as depth. The revision round only partly corrected this, so Mölkky, Mikado, Jenga and Crokinole likely deserve higher places.
- Agreement is not independence. All four judges are AI models from two model families reading the same evidence, so their agreement partly reflects shared assumptions.
Considered but left out
- Conway's Game of Life: after setup there are no player decisions.
- Coup: a role table makes the rules heavier, and its reputation is modest.
- 6 nimmt!: outcomes are too dependent on luck.
- Hnefatafl: many special-case rules.
- Nine Men's Morris: solved as a draw in 1993, with three phases of rules.
- Dominoes: outcomes are too dependent on luck.
- Team sports are out of scope.