> The approach we take is to track a lot, and figure out what works the best over time.
That's really, really bad statistics.
That's really, really bad statistics.
In general if you're going to provide a critical comment I think it would be better for the community here to expound on it a bit so everyone can understand your argument.