I identify with being "mixed race" far more than any individual race

Not technically a race, but then again neither is "Hispanic", which keeps getting treated as if it was a race. Race is a social construct anyway, so might as well.

I'm a bit surprised "mixed race" didn't occur to me as an option to suggest. It is true that I don't emotionally identify with either of my races, but I don't emotionally identify with "mixed race" either, probably because I wasn't raised in a community of mixed-race individuals and don't know that many mixed-race people. I feel like there isn't really a unique shared culture to unite us. Upon reflection, I've decided that if "mixed race" became available as an option on a future LW survey, I would continue to pick "other", because I really do identify with the human race more than anything else. The word "identify" is key though. If it simply asked what race I am, I would defer to the general consensus for how people should be classified, because I'd assume that's how the survey-writers want us to answer.

I've been doing that too actually, although I am somewhat tempted to upvote some of the recent survey-takers just to make the playing field more equal for people whose other time commitments made them unable to take the survey very early.

I thought about suggesting to Yvain to edit his post by including a suggestion for people who have finished the survey to check back again later to upvote new survey-takers, but I get the impression he may prefer having this incentive against people procrastinating on taking the survey. It does at least mean that on average, the more heavily involved LWers are going to be awarded more karma since they're more likely to notice the survey as soon as it's posted.

This however has to be weighed against the disincentive for latecomers to take the survey if they didn't see or were otherwise unable to take the survey early. (Yvain has also on occasion made little changes to the survey after it's been posted, but I don't think that's enough to be a good incentive to take it later.)

This seems to be a much worse way of achieving what "Best" shoots for.

Not necessarily. Someone who has already seen the best comments and returns a while later to see what new but good comments have been posted may have a use for it.

That is interesting! I think some of Gwern's upvotes are coming from people who agree with his "Basilisk" comment / found it because of the discussion it generated

It didn't seem self-evident to me that his mention of the basilisk would help his comment's score overall. I don't personally believe in the basilisk and I do think it would make an interesting survey question, but I thought many LWers considered it a dangerous idea to discuss? They may think that even if they don't believe in it either. Or maybe Eliezer was just weird in his reaction to it. Judging based on Gwern's comment's 99% positive rating, that's certainly what it looks like.

I think there's also a factor of people not loading all the comments- otherwise we wouldn't expect the oldest comment to be lower than the early bulk of comments.

It's not so far off that I feel the difference can simply be attributed to people not loading all the comments. At the time of my writing this, the oldest comment has the same score as the third and fourth comments.

Gwern (79) and Vaniver (66) show significantly more upvotes than the next in line

Thanks. That's interesting. I hadn't noticed that. They even score higher than some people who posted earlier, and with similar quality posts.

If upvotes are handed out according to the rule and logically in order of occurrence the vots should roughly read n, n-1, n-2,

...At first I was going to say I think it would be more of an exponential decrease since most people take the survey in the first few days and I doubt many people diligently keep track of new comments, but then I remembered that the rate of new "I took the survey" comments themselves decrease exponentially, probably at a similar rate, which cancels out much of the effect. Oh well. This does make the situation less unfair.

At first I thought this person would only downvote short comments that have little content beyond saying that the user took the survey, but I've since noticed that even "I took the survey" comments with very detailed critiques are getting the single downvotes. My guess is this person doesn't like the idea of some people getting 100% positive ratings through posting only survey comments, as survey comments would be the easiest way to attain that otherwise, or thinks that the amount of karma awarded by other users for these comments (even the detailed ones) is too much, and that karma should mainly be reserved for quality discussions.

Personally I think the amount of karma awarded for the short and simple survey comments should be based on the difficulty, time commitment, and benefit from having people take these surveys, but I think the amount of karma being awarded already is in line with that. Sure, there might be a few people lying by saying they took the survey when they in fact didn't, but I suspect that's pretty rare. I would like it though if there were some users who prioritized quality in deciding whether to upvote comments, so that it would be easier for people to quickly locate the most useful comments when they choose the "Sort by: Best" option.

I'm not sure I understand. I wasn't able to find explanations by typing "upvote" into the search either. Can you please clarify?

FWIW, this line of reasoning comes up pretty regularly (especially in response to that survey question), so if the surveyors fail to realize the associated difficulties, it's not through failure to have it pointed out. I suspect they realize it just fine.

Continuing to complain about it may still have an effect though. I personally think they should post the definition they're using for "supernatural" in the description for the question, maybe right below their current description.