Without a control group, or at least a group doing something other than playing games during that time, it's hard to say the games caused the effect they observed.

Maybe just staying after school with highly interested researchers had a knock-on effect on intelligence?

IQ tests are to a certain extent a learned skill; you get accustomed to the sorts of questions which get asked. If the kids had never seen an IQ test before their first one then just the effect of testing 'em twice in a row could easily account for 13 points.

I bet I could do a study which showed you could get a similar gain by showing 'em a youtube clip of giraffes mating.

Right, because you know how to conduct a brain study better than a neuroscientist at UC Berkeley who's been working on this for years.

Well, if IQ is so easily changed that I would argue that it is a pretty useless metric for measuring innate capabilities.

I have an older (16 years older) sister who trained as an educational psychologist - I was given all kinds of tests from an early age and I certainly learned how to do them so that later in my childhood I would always get ridiculously high marks on those kinds of tests and it was always pretty clear to me that it was just because I knew the tricks of doing these tests, not that I was particularly bright or that doing the tests increases your actual level of intelligence.

The speed game group didn't see reasoning improve, and the reasoning game group didn't see speed improve. No control, but it seems pretty valid.

I'm sure a better trial could be done though.

I thought that lots of studies had been done like this, and these types of gains always turned out to be short term. The article doesn't say whether the gains stayed with them weeks or months after the games ended.