Each questionnaire question can use a different scoring method depending on the type of groups you want to create.
These are optimization preferences for the algorithm — not strict guarantees. If it’s possible to satisfy them while balancing the rest of the dinner, the algorithm usually will.
Same
Group people who answered similarly.
Use this when you want people seated with others who selected the same answer.
Example:
“Looking to make friends” → seated with others who selected the same reason for joining.
Different
Create more diverse tables.
Use this when you want a mix of answers at each table.
Example:
Mixing introverts and extroverts, or different hobbies/interests.
Cohesion
Keep people near others who are similar on an ordered scale.
Use this when answer options follow a natural progression or hierarchy.
Example:
CEO
VP
Director
Senior
Junior
In this case, the algorithm will try to seat people close in seniority together.
At Least K Of
Ensure at least a certain number of people with a specific answer per table.
Use this when you want representation from a specific group at every table.
You’ll also need to select the answer as the “K-th option.”
Example:
“At least 1 person who selected ‘New to the city’ per table.”
