But they've missed the true comparison. The question isn't: "Would a London cabby using satnav software without having learned the city do better than a London

But they've missed the true comparison. The question isn't: "Would a London cabby using satnav software without having learned the city do better than a London cabby who learned the city?" The question is: "Would a London cabby using satnav software do better than one without - both having learned the city?" In other words: Assuming good satnav software that knew current traffic conditions and historic average speeds of the roads, with an easy interface to offer multiple routing options that are easily changeable (for instance) - wouldn't that be helpful to a cabby vs. a cabby that didn't have that extra tool? I don't think it's fair to compare human+tech vs. human-only performance with no tech. Better to compare them with tech, since the whole point is that technology is available and used.