Originally Posted by
Bill Miller
I'll buy that Andy, but if you're going to take that approach, I think you need to take it a step further. You need to document what the 'pure' process weight is, and what the adjustments are, and why they're being applied. In essence, you're doing a PCA at the same time you're classifying the car. Which is what you're already doing today, when you're deviating from process weight. It's just that it would be all out in the open. You guys have already put in place most (all?) of the tools that you need to do this. Use them. This goes to the whole transparency thing.
I'd like to make a comment on the subject of 'adders'. I think the ITAC did a great job in developing what's seemingly become a good, solid classification model that has gotten the cars pretty damn close. I think a way to refine that is to switch from a set weight amount for a given adder, to a percentage of car weight. To me, 50# on a 3000# car that makes 200 hp is a lot different than 50# on a 2000# car that makes 120 hp. I know this issue was discussed when we were working on the ITR car list. IIRC, it was felt that a 50# 'negative adder' for FWD may not be enough, given the makeup of ITR. I think that would be a step towards better refining the model.
Bottom line is, you'll never make everyone happy. However, if you develop something that addresses the three areas that Kirk lists, you'll have something that you can point to and say that it's as fair as possible. If people want closer parity, let them run spec cars. Eventually you get to a point of diminishing return, and you've gotten to the point in the S/N ratio that driver ability is the dominant factor.