This is generally what people say when they don’t understand statistics and the relevant statistics don’t support their viewpoint.
You do not need a huge sample size to draw conclusions in statistics. In a population of 100000 people you can actually get relatively significant data with less than 100 people.
There is also no evidence suggesting that the people who are sampled are in any way favorable to role queue (e.g. “but all the people who dislike it might have quit/not come to the forums anymore” is not a logically sound argument).
With these things in mind, once you also take into account the frequency of polls, the timings of them (before, shortly before, shortly after, and long after role queue was implemented), the variety of places sampled, and the consistency of the results, it’s pretty safe to conclude that role queue is overall more positively received than negative.
Statistics is often easy to disbelieve since the [proven] math behind it is actually extremely unintuitive to the human brain since we very rarely think in terms of probability. It generally doesn’t help that people in general are more willing to complain about what they dont like than praise what they do. It’s very common to see people make threads about the problems they have with role queue or how it’s bad. But people generally don’t make threads stating “all is good, I like it” because it’s kind of useless to spend the time stating you’re happy with the status quo.