After reading an article where Car & Driver compared the Z4 to the S2000 and showed the 0-60 time of the S2000 at over 6 seconds, I started thinking about what it really means. They had originally tested the S2000 back in 99 at 5.8 seconds 0-60 and retested it later, returning a higher figure. Interestingly, their methods assure that the throw out the old figure and use the new one.

Can anyone see anything wrong with that? Testing 1 car, and when you get a new data point, throwing out the old one? A sample size of 1 is just not representative of the population and any variation in the production process can throw those numbers up or down. The number I expect to see in magazines would be the average I should expect to get when I buy the car and I'd like to see a representation of variation in the results as well, or a standard deviation calculation.

Perhaps its not practical or possible for them to test 100 different cars and average their results, but publishing anything at all is anecdotal and misleading to the consumer. We all love to read the figures, but the bottom line is that testing a single car does not produce a representative figure, and you'd think that the policy would be that when they have the opportunity to test a new vehicle- that gets averaged into their sample. Having recently completed a statistical analysis certification, I find the method of sampling 0-60 from a population grossly inaccurate and flawed.

So don't believe the numbers you see because your results may vary, greatly.

Byron

 

Haha, I remember Russ's thread about this - it started a multi-page flamefest.

 

I like 60-100 more.

 

In the last test, I believe they explicitly stated that they were testing a car with something like 400 miles on the odometer, so it wasn't fully broken in yet. They test what the manufacturers' public relations department makes available to them, so blame Honda for not providing a fully broken-in unit.

Why wouldn't they list the number for the newest model year tested? The first data point was for a '00 model, the second for an '02 or '03, I forget which. The way cars change/evolve, what value would you attach to the average of a '00 and a '02 for any particular make? Should they average a 260 hp (or whatever, I don't really remember) M3 with the first 330 hp M3 they tested, rather than discarding the first data point and using the second?

Just take these numbers for what they are: the results that one driver got with one car.

 

What I wonder is why so many people care about tests like this? Does a bad test time or review in C&D change how you feel about your car? Do you have less fun driving it now? 0-60 is a nice number to compare cars with but anyone with half a brain knows that it can vary between different cars and drivers. It's not the definitive statistic that people should get worked up over. What is your suggestion that they do instead? Not publish any test results?

 

When you love something or are passionate about something, it's only human nature to take anything said about it personally or defensive. If someone wrote a horrible article about how the Oklahoma State Basketball team was absolutely horrible, you wouldn't die from it, but it would certainly bother you. Just like if I heard some anti-America comments now, it would bother the heck out of me. If someone posts an article about the S2000, which states the Z4 or the 350Z is better, I'll still love my S, but it would still bother me to read it...whether or not is was true.

 

[QUOTE]Originally posted by Red Blur
[B]When you love something or are passionate about something, it's only human nature to take anything said about it personally or defensive.

 

[QUOTE]Originally posted by Red Blur
[B]When you love something or are passionate about something, it's only human nature to take anything said about it personally or defensive.

 

Hey the Z4 is the NEW kid on the block ......... of course they're going to make it look good. Remember what they (C&D) did with the 350Z and S2000 comparo ?? C&D is only slightly better then Motor Trend.
At least in the past R&T will remind you of earlier test results.

 

[QUOTE]Originally posted by wickerbill
[B]

As for your example about my OSU Cowboy basketball team, I wouldn't get too worked up about an article that says our halfcourt offense is pathetic and we have trouble scoring.