Man... Do you know how many interviews are needed to make a sociological study of 50 million people? Maybe 500...
Perhaps the spectrum of this poll is not the best, but with the amount of people who have voted, the margin of error may not be greater than 15%.
Using this on-line calculator
http://www.dimensionresearch.com/resources/calculators/conf_prop.html, a sample of 1000 showing a proportion of 50% has a 95% confidence interval of +/- 3.1
percentage points (not 3.1% of the 50%, a common error). The real value in the whole population is likely to lie between 46.9% and 53.1% ...
if the sample is representative.
The issue with this poll is not statistical significance, which deals with random error, or sampling error, ideally in the context of random sampling. The issue is more with who does this sample represent ... and is this a biased sample from the perspective of representing the entire set of purchasers of Civ 5? The answer has to be yes (non-forum readers have no way of being included in this poll).
But I don't think that this has to be representative of all Civ 5 purchasers to be meaningful. This poll is saying that nearly half of folks on major fansite who cared to vote, and who presumably both love and understand the franchise, think Civ 5 is a step backward, rather than forward.
Now that may not be representative of all CivFanatics members either, as participation is voluntary and those unhappy may be over-represented. And it may only be a step from A+ to A- (is the game that bad in an absolute sense, or just in a relative sense ... exclusive of performance glitches and focusing on gameplay vision?), but we were expecting A+++.
But rather than argue about whether 47% is a majority (not of the whole sample, but it is the majority of those with an opinion), or try to change people's minds, or insult them for their opinions, energy would be better spent understanding in more detail the concerns of those who think the game has regressed, and how those could be addressed (potentially in ways that could be options, so everyone can have the Civ 5 gameplay they want) to make the game better.
With that in mind, sometime this weekend I expect to post a poll about whether Civ 5 needs to be "smartened up", allowing multiple choices among various areas where improvement may be needed (and the choice of "no, leave it alone" as well).
dV