Player Count Comes First

Before theme, before weight, before whatever your group is currently obsessed with, one question decides what can hit the table: how many people showed up. A tense two-player duel is dead weight when six friends arrive. A six-player negotiation game with three people is a sad little math exercise.

Board Game Menu applies that filter first, and it's a harder filter than it sounds. Weight just decides which of the three courses a game lands in; a game that's too heavy still gets considered, it just competes for the wildcard slot instead of the opener. Player count doesn't work like that. The site reads the official minimum and maximum for every owned game in your collection, and anything that doesn't stretch to your number is gone before weight ever gets a say. A game with no player count data on its BGG page at all is treated the same way: invisible, not defaulted to some guess. Get the count wrong and the whole menu comes from the wrong shelf, not just a slightly worse one.

The official range deserves some skepticism, though. Boxes are optimistic. 7 Wonders claims 2 to 7, and the 2-player variant is a different game wearing the same clothes, played with a robotic third hand instead of a third player. Catan says 3 to 4, but its own community "best at" votes on BGG lean toward 4. Agricola technically plays solo, but nobody buying it for family night cares. That "best at" data exists for every game on BGG: a running poll on each game's page where the community votes whether a given count is best, merely recommended, or not recommended at all. It's a genuinely different question than "does it technically work at this count," and it's worth a look whenever a suggestion feels off.

Odd numbers get the roughest treatment from game design generally, not just from this site. A lot of card and board games are built around passing turns to your left, dealing hands that divide evenly, or splitting a fixed set of resources — mechanisms that work cleanly at 2, 4, or 6 and get lopsided at 3 or 5. Negotiation and alliance games are the clearest case: with three players, any deal is a majority, so there's nothing to actually negotiate. Add a fourth and suddenly coalitions matter. None of that shows up in a box's printed range, which is exactly why the "best at" data is worth checking specifically when your table lands on an odd number.

Two-player tables have their own version of this problem. Plenty of games technically support two but were clearly designed for more, and what you get is a diminished version of the real thing rather than a variant built for the count. Others are the opposite: Star Realms and Jaipur are built as two-player games from the ground up, not stretched to fit, and the 2-player mode bolted onto a bigger box is sometimes its own separate ruleset rather than the normal game with fewer people. Best-at data tends to tell the two apart faster than reading the back of the box does.

Checking "prefer games rated best at this player count" on the generate form leans into this. With it on, the site still only pulls from games that support your count, but within that group it reaches for whatever's voted best at your exact number first, and only falls back to games that merely support it if nothing at that count has a strong vote. Leave it off and every game that supports the count is treated as equally eligible, best-at data or not. Locking a pick and rerolling the rest of a course doesn't turn any of this off, either. Whatever fills the empty seats still favors games voted best at your count first, the same as a fresh menu would.

A few things worth knowing:

• Count seats, not RSVPs. If two people are a maybe, generate a menu for both counts and see which lineup you like better — with best-at data in play, the two menus can differ by more than one swapped name, since a completely different game can be the top pick at five than at six.
• Treat the menu as a shortlist. If it suggests a game your group has learned plays poorly at that count regardless of what BGG says, reroll. You know your table better than the box does, and better than the crowd does too.
• Expansions are excluded automatically, so you won't get the Wingspan European expansion suggested as its own game.

Get the count right and the rest of the night mostly plans itself.