Board Game Recommendation Tool for Real Group Constraints
Find playable game options for tonight by combining player count, duration, complexity, and collection availability.
Recommendation quality depends on context, not popularity
A good board game recommendation tool starts with your real table constraints. Player count, remaining time, and complexity tolerance matter more than generic top-game lists.
Recommendations should be explainable so groups can trust why a game is on the shortlist and when AI re-ranking changes order.
Prioritize games attendees can actually play
If a game is not available in attendee collections, it can still be interesting but less practical for tonight. Your recommendation workflow should separate ideal fits from immediately playable fits.
What2Play uses collection awareness alongside constraints and optional AI re-ranking so your shortlist is useful in real event planning.
Filter by player count and time first
Apply complexity preferences for the current group
Highlight games attendees can bring now
Support fallback picks when turnout changes
Optionally apply AI re-ranking based on event context and recent feedback
Pair recommendations with voting and feedback
Recommendations improve when they are connected to actual outcomes. If a group consistently enjoys specific weight and duration ranges, that feedback loop should influence future shortlists.
This creates a learning system instead of a one-off recommendation result.
Speed up setup with shelf photo import
Recommendation quality depends on accurate inventory. Shelf photo parsing (Gemini + Claude + OCR supplemental matching) helps hosts capture missing titles quickly before event setup so candidate pools stay complete.
That means less manual library cleanup and better coverage when new games arrive between events.
Frequently asked questions
Can recommendations include games not owned by attendees?
Yes. Discovery-style recommendations can include non-owned games when hosts want broader exploration.
How many games should a shortlist include?
A shortlist of 5-10 options is usually enough to support quick voting without overwhelming attendees.
Do recommendations replace host judgment?
No. Recommendations narrow options, and hosts still decide how strict constraints should be.
Can hosts manually adjust recommendation lists?
Yes. Hosts can add or remove candidates before voting to reflect context that automation might miss.