Well, unless you merge all servers into a single server, a cross-server tool will always be more efficient than a single server one.

That would be hard, since there are is valid reasoning to support what he said.

That has absolutely nothing to do with the subject being discussed.

Whether or not it will happen is still to be seen. The reason this post keeps showing up is because we're very far from a consensus when it comes to the subject. And you realize that it was a moderator that remade the thread when the last one got too big, right?

No, my point is that if we start limiting the functionality of one particular tool because it could, potentially, from time to time, be used by a limited subset of the players to harass other players, we might as well go after pretty much all of the tools in the game.

What about the inspect feature? Are you going to start crusading for it's removal next?

Of course I would. Because I believe that limiting the functionality of a tool is not the right solution in order to prevent potential misuse of said tool.

There are only two thing here that are beyond arguing. First, that the margin of error does not account for bias. That is a fact, and one you can easily verify by reading any introductory text on statistics. Second, the poll is biased. The participants are self-selected. That alone means that people with a strong opinion on the subject, towards one end or the other of the subject under polling, are much more likely to have answered. That alone is enough to prove the poll is biased.

Actually, the margin of error accounts for sampling error in a random poll. It has nothing to do with bias error. A 5% margin of error at 95% confidence level means that if you were to repeat the same poll 100 times with random samples, 95 times out of 100 the result would fall within a +/-5% interval centered around the true result. Selection bias, namely by the creators of a given poll, is not the only kind of bias. First of all, the only person that said that was you. Second, even if I was not pro-public logging, as you say it, I'd still be pointing out that we can not draw any conclusions from this poll.

No, that's not even remotely close to what we're arguing. What we're saying is that for the poll to be valid, for the margins of error to mean something, and for us to be able to draw any conclusions, we'd have to poll a random sample that is also representative of the distribution of the different play-styles in the population. Your failure to understand that just shows that it is in fact you that knows nothing about polling techniques.

I added the bolded part. If the sample is known to be biased or, as for this poll, we have no idea if the sample is biased or not, the results of the poll are meaningless.

No, since we have no idea on how biased the sample is we can make no assumptions on what the margin of error really is. Just to give you an extreme example, I'd bet you anything you want that if we were to poll the subject of stricter gun control laws next to a NRA rally we'd see results that are even more of a landslide win for one of the sides than the ones we have for this poll. It still wouldn't mean that the sample is unbiased and that the results reflect the opinion of the entire population.

The margin of error is a function of population size, sample size and confidence level you want from your poll results. All you said is correct, but you forgot to mention that your sample needs to be unbiased for the analysis to work. And we have no indication that the people who frequent the forums and chose to answer the poll are an unbiased sample of the game's population when it comes to the issue being discussed.

You implement something similar to what GW2 will be doing with their server "visit" system. We should be moving away from server barriers, not clinging to old limitations that only ever existed due to technology not being quite there yet.

It is if you want to have real-time analysis of the data.