^What you've described is EXACTLY what I'm trying to keep from happening. Since the minimum for inclusion in the results is "only" 6 people, it's very easy for a small number of people to make massive swings in ranking for coasters that not many people get to ride.
In all seriousness, I was pushing for 10-12 riders as a minimum, or 4-5% of the total respondants.
IIRC, we ended up at 6 for inclusion, but anything with 7-9 voters got a '*' in place of a numerical ranking to distinguish the low ridership. This is how the #2 (or was it #3?) coaster got a '*' last year. I think 7 people had it on a ballot in total.
I prefer the percentage system instead of a fixed number, as if the poll is popular one year, it takes more riders in agreement to affect the results, but if less people fill out ballots, coasters won't necessarily disappear from the results, either.
I said as much last year that I don't think last year's results were a deliberate orchestration, but at the same time, I'm not 100% confident that ERT conditions are the most accurate way to assess a coaster. (I know I enjoy myself more, and thus may skew my own ballot, so I'm sure it may have the same effect on others.)