Even without the contrast (impossible given this particular dataset since the survey wasn't given to a control), it's still interesting. Here's why -- when you

Even without the contrast (impossible given this particular dataset since the survey wasn't given to a control), it's still interesting. Here's why -- when you consider the broader context of the outlash against beer/alcohol at conferences, notice that only 5% of developers said they hated beer. That apparently means a tiny minority of people could have an outsized influence on conference behavior. Note that this survey wasn't given purely to conference-goers, so it's not selection bias.