I've considered this in the past but the problem is that there is very little reliability in a survey with a sample of only, say, 4-6 people. The danger is that you may use the survey results in future situations with other players. It's probably ok if you only use the results for your own...