Footie,
I don't know what the standard deviation that can be calculated from the regression analysis we did means when we start getting very specific about the exact chances of an event happening. As I said earlier, it shows that the chances are low, but I don't know that I fully agree with Swamp's interpretation of the computed deviation mainly because there are probably sampling issues that affect the outcome. Also, there might be issues of data integrity. I originally took a bunch of numbers from a webpage where times and power ratings were posted. Most were supposed to be SportAuto numbers. As I mentioned in that original post, I never checked the validity of any of those numbers. Regardless, I still stick to the claim that the chances of the GTR event taking place are pretty low.
The thing to do would be to use SportAuto times only. If the GTR is so magical, they should be able to replicate it as well--meaning it should remain an outlier in "their" dataset as well. If it doesn't stand out the way it currently stands out with respect to the other cars we plotted, then there was indeed something fishy about the power ratings of the car Nissan tested. We know that SportAuto does not always necessarily get the most THEY can get out of a car either, but the variation there is probably more like 5-6 seconds, and we can claim that to be random (to say otherwise would be mainly to say they favor certain manufacturers over others). So, it might be worth it just to run SportAuto times once they do their GTR Supertest if they haven't done so already (I haven't been following that).
Cheers.
