I did not include people who voted for option 3 - they helped with deciding what to do with the knockback. So, 115 / 158 = 75%. We realistically cannot get higher than this, even with larger sample sizes. We understand that the sample size is low, is something we want to address at a later date...