Assuming a population size of ~100,000 and your sample size of 85, your margin of error is about 10%. So actually, the results could be roughly split down the middle or could be overwhelmingly in favor of the (somewhat biased, see my comment above) "meh" option. I wouldn't quite go as far as...