Ls Mot Failure Stats

[runbly]

Have a look here - the guy has converted the PDF to Excel:

http://www.bbc.co.uk/blogs/opensecrets/201...s_released.html

Shows MOT stats for the LS models (well, and everything else!).

[JellyBean]

Have a look here - the guy has converted the PDF to Excel:

http://www.bbc.co.uk/blogs/opensecrets/201...s_released.html

Shows MOT stats for the LS models (well, and everything else!).

Thanks - it's fascinating. But at 1,200 pages a bit daunting at first. This is what I came up with, for 1996 cars (same as mine):

Lexus LS400 = 26.4% MOT failure rate.

Merc S420 = 32.6%

Audi A8 = 36.5%

Jaguar XJ8 = 37.2%

BMW 740 = 39.4%

Hmmmmmmmmm indeed.

[Ninety-Eight]

I''ll say the same here as I've said elsewhere. These figures are wrong. Plain and simple. There are huge numbers of cars missing from the data. I have looked through the original PDF from VOSA and the data is wrong there too. So because we don't know which cars are missing we don't know how this effect the percentages.

If another 10,000 LS400 which passed first time were missed the % fail rate would be better even than is quoted there, but say they missed a significant number of S420s out which passed first time, maybe Merc would pip Lexus to the #1 spot there....

Its a sad example of incomplete data being released under FOI or press release and a bunch of journos and people on the internet making a series of assumptions based on that. I have already seen some interesting statements made and comments off people who clearly don't understand statistics for a start.

But this is irrelevant when the data is wrong.

[JellyBean]

I''ll say the same here as I've said elsewhere. These figures are wrong. Plain and simple. There are huge numbers of cars missing from the data. I have looked through the original PDF from VOSA and the data is wrong there too. So because we don't know which cars are missing we don't know how this effect the percentages.

If another 10,000 LS400 which passed first time were missed the % fail rate would be better even than is quoted there, but say they missed a significant number of S420s out which passed first time, maybe Merc would pip Lexus to the #1 spot there....

Its a sad example of incomplete data being released under FOI or press release and a bunch of journos and people on the internet making a series of assumptions based on that. I have already seen some interesting statements made and comments off people who clearly don't understand statistics for a start.

But this is irrelevant when the data is wrong.

You have a point. The resulting figures may be inaccurate, but as the same methods are used for all cars, they will still have some relevance as a comparison between different makes & models - providing the sample sizes are decent.

[Ninety-Eight]

While I agree to a point but since we don't know what data is missing (only that lots of data is missing) we can't say how accurate the data we do have available actually is, thus any conclusion drawn from it is unreliable at best.

If no Lexus data is missing and no Merc S420 data is missing the comparison is valid, however we cannot say for certain that all the data is there for these two models.

If you want to see how bad the data is try find a single Morris car in the whole data set! And there are thousands of them still on the road... There are other completely missing manufacturers and many models missing within the manufacturer. Not one single Mitsubishi FTO in the data for example.

[Ninety-Eight]

Oh, and, we know the LS400 is a better car than any of the others listed, we don't need VOSA stats to prove it!

[mikesands]

I too find some sympathy for the DVLA view. To me the data even if complete, is more a reflection of owner behaviour than model reliability. Most of the failures are down to brakes and tyres, and surely that will be down to whether the owner has replaced components when they should.

Mike

[Ninety-Eight]

^^^ this was also my first thought. The data proves which brand of cars are more likely to be correctly maintained.

Interestingly the BBC has lead with "Ford most likely to fail MOT" and the Register has lead with "Renault cars tops MOT failure charts".

Now what was that "lies, damn lies and statistics...." :)