ego.istic.org

More from the Bureau of Information-free Numbers today. Language Log has covered the statistical white lies in rhetoric recently. FlipC often comments about that, and has posted recently about the latest attempt to blame child violence on computer games. The BBC has interviewed some “industry leaders” about this. Both Philip Oliver, Chief Executive of Blitz Games, and Paul Jackson, Chairman of ELSPA, have been quoted as saying that the average age of gamers today is 28.

Imagine the entire games market consists of 3000 eight-year-olds, and 1000 pensioners who've each counted 92 summers. The average gamer in this market is 28, but three-quarters of the market is too young to be exposed to sex, violence, and bad language (outside of a family setting, anyway). Now imagine the entire games market consists of 28-year-olds. The average gamer in this market is also 28, but 100% of the gamers will play whatever they feel like, and no amount of parental control of labelling will stop them.

The number 28, without any clue of what the distribution is, gives us no information at all. What's crucial for considering labelling is: what fraction of the target audience is over 18, or 15, or 12? When the industry decides what topics to base a game on, it will be more concerned with what fraction of the spending power is in the hands of over-18's, over-15's, and over-12's.

Messrs Philip and Oliver, or their press agents, or the BBC editors (there was no by-line on the article), presumably mentioned that number to make the rhetorical point that they make violent games for a large, grown-up audience. But it would be rhetorically much more effective to say that more than 70% of gamers are over 18. It's a simple number still readily comprehensible to the statistophobe, and it won't make statistics-aware people think you're trying to pull a fast one. (The 70% data are for the US, and interestingly enough, the linked page falls into the same trap of having a big caption over their pie chart saying, “The average gamer age is 33.” They also have the entertaining non-statement, “Seventy-nine percent of game players of all ages report exercising or playing sports an average of 20 hours a month.")

Just as I am wondering whether some kind of news story labelling would help to ensure that children are not exposed to statistical nonsense that might make them grow up stupid, I visited another BBC news story about women not feeling confident talking to staff in gadget shops. This is another piece of reporting-by-press-release, where Saatchi and Saatchi has conducted some poll, issued a release, and some reporter has retyped it into a “news” item. I've commented about such laxness before, as have those fine people at Language Log. This one has an extra sting to it, though. The conclusion universally drawn by reporters is that women are interested in technology, but are put off by retailers. This conclusion is supported by the datum that “a third of women internet users said they would spend more on gadgets if retailers learned how to approach them.” Note: conclusion does not actually follow from premise.

Let's play another game of imagining what other states-of-the-world could have led to the quoted observation. There are about 30 million women in the country. 29 million of them are utterly uninterested in technology, and would not describe themselves in a survey as “internet users”. However, they go into gadget shops regularly to buy presents for male family members or friends, and say that shop assistants are always friendly and give advice that's very helpful to customers who know nothing about what they're buying.

The remaining million women are very geeky. They would describe themselves as heavy internet users, and must have all the latest electronic crap on the market. But 300,000 of them find that whenever they go to shops, the pushy shop assistants keep trying to tell them what to buy, in the most patronising fashion, even though they already know exactly what they want. They end up buying everything on the internet. Everything's cheaper on the internet, so they spend less than they would if retailers learned how to approach them.

This state of the world is good. The high-street retailers can sell to more than 96% of women. The shop assistants are doing their job well. The female technophobes are happy that they can buy the right gifts for their spouses or sons. The female technophiles get good bargains via the internet.

The missing link in this case is with the “female internet users”. How many women is “a third of female internet users”? Is it a third of women as a whole? 1%? We just don't know, because they just don't feel like telling us. In any case, I'd be more interested to hear whether men in fact feel the same way about gadget shops, but we just don't like to admit it.

The same article said that “many said that they felt patronised by pink digital radios and diamante-encrusted mobile phones.” As I write this on my LED-backlit keyboard, I glance at my titanium-plated watch and notice it is time to turn on my football-shaped radio.