2% of gamers are hermaphrodite: official

[Gareth]

It's official: two per cent of gamers are neither male nor female, leading us to believe that this small yet significant section of the gaming demographic is either hermaphrodite, represents a species further down the evolutionary pyramid or is perhaps of silicon-based extraterrestrial origin.

That's according to the ESA (Entertainment Software Association) 2005 report into the state of the industry - enticingly entitled "ESA'S 2005 Essential Facts About the Computer and Video Game Industry"

source: http://www.theregister.co.uk/2005/10/17/gaming_shocker/

Looking at the 21 people in the NTHW Clan, I have my doubts about Ramrod personally!

[downquark1]

American polls usually have a "don't know" box.

Either that or a 2% margin of error.

[Anonymouse]

Even if that statement is true at all, 2% of how many people?

Whenever someone claims x% of people do/want/are such a thing, you should ask: x% of how many people?

Example, from Darrell Huff's delightful little book How To Lie With Statistics: not long after Johns Hopkins University first went co-ed, one of the local shock-horror rags printed a front-page story that fully one third of the co-ed students had married faculty members. Scandalous! The thing is, the story was entirely true and factual.

After all, 1 is a third of 3, isn't it?

All these surveys - all of them - are totally meaningless unless the sample size is statistically significant...which it never is because that would be horrendously expensive to do. Given that the Earth's population is now about 7,000,000,000 (the official figure puts it at less than that, but who trusts official figures anyway?), you'd have to survey millions of people to get a meaningful result.

Worse, you have to avoid sample bias to keep your results meaningful. You survey people during the day at home - missing people who are at work. You try to catch people at night - and miss the night workers and people who are out on the town. You can't win. Virtually every sample has some bias.

Besides, the one thing that no statistical analysis can take into account is the fact that people lie. That alone generally invalidates any survey you care to name; you can't use statistics - or, indeed, any mathematical analysis - on inaccurate data. I don't know why they bother, really.

For the record, I'm a bloke and have every intention of staying that way.

[zoombini]

I'd go for the "silicon-based extraterrestrial origin" bit, talking on XBL they all seem a bit alien to me. :-)