I would expect this sort of game to have difficult honesty issues even when it is a single gender. For example, if some individual has a fetish that is in some way connected to one of the individuals (say for example a celebrity that frequently wears some sort of clothing, or only one of the three falls into a racial group they have a fetish for) how likely is it that someone is going to be honest about that motivation.
That said, I agree that mixed groups will likely have more severe honesty issues.
The criteria are a little different, though.
I’ve played it in mixed groups, its generally about perceived personality features rather than subjective attractiveness.
I wouldn’t expect this to be a recipe for honesty.
I would expect this sort of game to have difficult honesty issues even when it is a single gender. For example, if some individual has a fetish that is in some way connected to one of the individuals (say for example a celebrity that frequently wears some sort of clothing, or only one of the three falls into a racial group they have a fetish for) how likely is it that someone is going to be honest about that motivation.
That said, I agree that mixed groups will likely have more severe honesty issues.
I’ve never treated the game as a data collection exercise. IT is more suited to social bonding and conversation stimulation.
For more statistically useful data okcupid has done studies, as have hotornot and its various imitators.