BlkZoomZoom said:
For those that think this is happening in every other Protege, it's not. We have done 3 in 3 years. Wow.
There are two scenarios for these failures:
1. There is a small subset of vehicles which were not assembled
correctly and these will all fail within 50k miles (or whatever)
2. The design is defective and ALL vehicles may fail eventually but the
rate of failure is small, for instance as you say:
(3 failures)/(Number of 2003 P5s serviced by your dealership * 3 years)
For the sake of argument let's assume that there are 50 2003 P5s you might have seen at your dealership. That gives a failure rate of 2% per year in the second scenario. However, that rate has so far been measured on fairly new cars. Cars tend to shake things loose more and more as they get older - it's just the cumulative effect of all of the vibration. So in model 2 the failure rate should probably be increasing with time, whereas in model 1 it might have peaked and now be decreasing.
Mazda undoubtedly knows which of these scenarios holds true since when summed over the whole country the failure rate can be accurately determined over time. However none of us outside of Mazda can perform this analysis. All we have is hearsay and the observation that rather a lot of the people in this forum, which is probably not a very large subset of all Protege owners, have had this problem.
Hmm. I guess we could estimate the cumulative number of failures. Assume that everybody who is a member of this board and has had this problem with a P5 has reported it. So there's one number. Somebody count those up and we've got the number of observed failures = F. We also need to know the number of 2003 P5s belonging to forum members = P. F/P is the fraction of 2003 P5s affected (that forum members own). Perhaps whoever manages this site has some way of scanning the member list for car models to determine P. If we also knew the number of 2003 P5s sold in the USA and Canada (=S) we could estimate the total number of failures (=T) pretty accurately as
T=S*F/P
I'm betting that S is a matter of public record - who can find it? If the number is big enough then the class action guys may have a leg to stand on. Or at least enough of a message to interest the legal vultures.
Ok, it might be an overestimate as people who've thrown a screw and blown their motor are probably motivated to look around on the web for more info, and they might then join this forum.
Now I wonder what the values of F, P, S, and T are.