When Hiroshi Kajiyama, program manager for the all-new Mazda 6, began work on the new car one of the first things he did was forbid his engineers from driving anything other than the BMW 3-series (old and new), Audi A4 and VW Passat. In different ways they represented the benchmarks he wanted his new car to hit, and he didn’t want his engineers’ senses dulled by driving anything else.

"If you evaluate too many cars, you aim at too many targets and get a dilution of what you are trying to achieve," says Kajiyama. "And I am looking to my test engineers to achieve the absolute best results."

Kajiyama, you see, is mildly obsessed by making sure his cars are the best in the areas he wants them to be, which in the case of the 6 most notably means leading the way on dynamics and efficiency. If that means sacrificing a smidgen of refinement, for instance, so be it. Good can be good enough, so long as there is excellence where he wants it.

What’s key, though, is that when Kajiyama says "best in class" he doesn’t mean looking at a spreadsheet and seeing what numbers come out of a computer, but by feeling and experiencing what the results are for himself.

"We want a linearity of steering weight," he says by way of example. "So as speed builds you want to feel that build through the steering in a consistent way. Now, a computer will draw you a graph with a straight line on showing, the weight building in exact reference to the speed. But that is not how a human experiences it. A human experience is more of a curve. I don’t know that all our competitors understand that."