It's news to me. The only similar thing I ever heard was the mechanical system I proposed for the speed figure study that Sportstat was going to do(which they never did)- take the last 3 numbers, throw out the worst, average the other two. No idea how it would have worked out.
TGJB