Our 2002 handicapping system was based on the PAX (Professional Auto-X) system used by SCCA. We realized from the outset that applying an SCCA system to our PCA classes would have some problems, but what the heck, it was just for bragging rights anyway and it might give us some insight into how to create a system better geared for Porsches. Now that we’ve had several season’s experience with the BRI, we have indeed been able to identify some significant flaws – in spite of which, the winners generally thought the system worked pretty well.

I suppose that if one were smart enough, it should be possible to create a sophisticated model of vehicle performance that accounts for power, weight, tire compound, suspension geometry, brakes, gearing, etc. However most of us are not that knowledgeable, so we adopt a simpler empirical approach to modeling performance based on past race results. The PAX system was built upon a large national database of SCCA autocross results. But since Porsches are not very well represented in the SCCA, we wanted to compile a P-car database using results from our own events. With help from Tom Tweed and Chris Benbow, we have collected results from PCA-SDR events for the past four years. In compiling this database, we focused on the winning times for each class, rather than using the average time of all the cars in the class. We reasoned that the winner’s time is more representative of the performance capabilities of a well-prepared and well-driven car of that class. With this approach, we make no distinction between the different Porsche models running within a particular class; so a 356, a 912, and a 914 running in AS/S will all have the same index this year.

Over the past several years, many of our best drivers have jumped from one class to another as they bought different cars or made modifications to their old cars. There have even been rumors of drivers who may have changed classes looking for a greater (or lesser) level of competition. Being naively optimistic, we had hoped that with so many drivers shifting classes, averaging 4 years worth of race results would tend to even out the drivers’ contributions and we’d be able to identify the performance characteristics of the cars in each class. But real world data is rarely as clean as one might wish and we’re still left with obvious inconsistencies, such as classes in which drivers on street tires were faster than drivers using sticky tires.

Since a handicapping system based strictly on past performance data can lead to unacceptable results; it is necessary to apply some subjective judgements when creating a model that fits the data in some approximate manner. So we attempted to follow a couple of some simple guidelines consistent with the Zone 8 rule permitting a competitor to run in a higher class:
(1) As cars move up in class, the index should reflect increasing levels of performance (FS/S should be faster than CS/S, which should be faster than AS/S, etc.),
(2) As cars within a given class become more modified, the index should reflect an increase in performance potential (prepared should be faster than stock, which should be faster than street stock).
In general, past race results are consistent with these trends, but there are some notable exceptions. We then attempted to fit a simple smoothed surface through the data consistent with these 2 rules. The resulting index is given in Table 1.

All of these numbers are considerably higher than those used in 2002; due simply to the manner in which the race results were normalized. The PAX system creates a ratio relative to the TTOD. We found that the data become more consistent if we use the average winning time for all the classes on that particular day to normalized the data. An index greater than 1 corresponds to a class that’s faster than average and an index that’s less than 1 corresponds to a slower than average class. As before, your corrected time will be simply your actual time multiplied by your class index.