I think a lovely addition to this site would be to give us more control over filtering the results of polls. Currently, the vast majority of polls are decided by partisan politics, not by people actually answering the question. Let me give you an example, below.

Ignore the number of 1 and 10 answers for a moment and this data suddenly becomes a bell curve. Sure, the 7 and 8 are swapped around, but that's fairly minor. This bell curve represents people who are legitimately answering the question. We can therefore hypothesise that the results of '1', legitimately got about 2% of the total vote, leaving slightly more than 15% of the vote representing people who are voting '1' because they prefer either Sony or Nintendo to Microsoft. Similarly, the answer of '10' should have got about 1% of the total vote, leaving ~2.5% of voters simply clicking the most pro-MS option there is.

There are many other examples, but I picked this one due to the fact that the nature of this one (10 answers, with a quantitaive grading of how positive they are) means that it will produce the best  bell curve.

I would like to see the site associate our data with a list of companies that we own currentish gen consoles for, and allow users to filter the results to only include certain console owners. This will allow us to determine, in this example, how many people genuienly thought that MS's conference deserved the lowest possible rating, and how many are simply trolls.

Thoughts? 

What about the people who own all 3? I feel like there would be far too much mislabeling and nonsense thrown back and forth if everyone is "tagged", more than now even.

Either it goes based on most games, or it simply puts you in with all 3, so an 'exclude' term would exclude you, and 'include' term would include you.

I think the downsides are outweighed by the positive. Even though my example demonstrated the effect well, it is probablt the case where it is least important. If there are the answers 1-10, you can pretty much ignore all trolls and not ignore any real answers by simply ignoring all 1's and 10s. but what about the case of 

Obviously, 41% of people don't actually think that there will be a pricecut this year. This poll is made completely meaningless by partisan voting.

While I agree that it is highly unlikely that Nintendo would do that, I don't think the results are so much fanboyism as general ignorance or overeaction to the PS Vita price news.  I think you overestimate people's breadth of knowledge on this subject :P.  People really think Nintendo should do that, sometimes people even vote based on what they would do lol.

I actually think your idea is pretty cool, I like that you're looking out for the site.  I just feel like the room for error is great enough where a lot of people would be forever alienated for being a fanboy or irrelevant in any discussion.  Forums are volatile enough, ya know?

P.S. Also, it's hard to mathematically adjust for "stupid" or for mistakes because of lack of facts.

P.P.S.  Maybe it can be changed or refined in some way?  But I've said enough, I'll check back to see what others think lol.