If you are familiar with the system it is now rated O for average. I averaged out the scores for a 96 and it looks like it falls .2 short of "better than average" and goes .8 over "average". They are calling it a " C " , but it is more like a " B- ".

,

The last published info was in the April 98 issue of the magazine, I believe (actually on the newstands last March). What are you saying it was downgraded from? <p>Each year, all cars are judged on an "absolute" scale, no longer only relative to other cars of the same model year. Because of this, ALL cars show worse marks as they age, because more defects start showing up. When the cars are new, or less than a year old, they all get mostly "A" grades.<p>Based on the current scale you they are using, you would have to give the car a grade based on its relationship to the grade of other cars. The scores can't be looked at alone or converted to a letter grade. (The letter grade would be "F" for almost every 7-year old car made). They do publish, however, a couple of pages of charts that summarize all of the data relative to other models, but I don't think you were using that because you wouldn't have had to calculate anything.<p>That chart (page 61, April 98) shows the A4 to be 30% better than average, right above the Toyota Camry, surprisingly enough, and right below the Lexus ES300. This chart was based on 96 and 97 models, and is now, of course, one year old data. At the top of the chart for "medium cars" were the Maxima, Accord, Infiniti I30, Millenia, Legacy, BMW 3XX, Altima Volvo S70/V70, then the ES300, A4, Camry, Lumina, Breeze, Mercedes C. Those were the only above average models. At the bottom were Taurus, Cirrus, and Stratus (the worst, at >40% below average).

Your own experience with one car? Reading the forum and seeing people's complaints of defects? (Have you compared to the rate of complaints at other make's forums?) Maybe your own dealer's possible inability to fix problems?<p>The average car a couple of years old has a couple of problems per year. Sometimes we get lucky and own one that has nothing go wrong for years, and other times it averages out by having lots of trouble spots.<p>Do you have a source of data that says the A4 is less reliable than the average car?

My 98.5 2.8 FWD MAN has in the first 7 months been more reliable than my 88 Honda Accord during the same time period. Since I only had a 12 month warranty with that car, I won't be able to directly compare after 12 months, but if what happened to my Accord in the first 3 years happens to my Audi, I am going to be even happier I got the Audi because I will know that it was covered under warranty, not under secret warranties. And I thought the<br>Accord was reliable...<br>My 2 pfennig worth<br>Clair<br>

Bear in mind that CU reliability ratings are relative to the universe of cars they are testing. So the "decline" in reliability rating may not signify an absolute decline; the A4 may instead be holding steady while other makes are improving.

Dale, I thought the 98 ratings were pretty good, even better than I expected. Most of what you say about how I derived the B- rating is fairly true. I guess I am alittle surprised that the 96 model (in the 99 issue ) gets 7 "much better than average(5)" ratings and only 1 "less than average(2)" Okay, so it was a generalization on my part,after all, I'm the "Empirical Einstein"(see post below)<p>Bottom line 1998 predicted reliablity "better than average" pg 31<p> 1999 predicted reliablity " average" pg 135

The CU ratings do not reflect how an owner treats their car. For example, do they drive on rough roads, live in an extremely hot or cold environments, keep up with regular service, drive agressively, frequently drive in stop and go traffic, etc. There are so many other factors that could also contribute to poor reliability. Cars are products engineered and manufactured by humans - they will never be perfect.<p>For once I have a car that stirs my soul, and I don't care what it costs to maintain!<p>Dave M - 97 1.8tm, 12K miles, no problems

Yes, CU has a statistical sampling problem.

On what are you basing your statement that CR has a statistical analysis problem? Are you saying they don't receive enough data? It looks to me that when this happens, they say so and don't publish data for that year of car! They do have statisticians employed also, and I assume that they know what they are doing. <p>Also, if the Automakers could even make a reasonably credible case that CR was distorting data or presenting data that didn't reflect reality, they would sue CR in court and win. This hasn't happened. <p>Another also: If their sample size is too small, how come their data is consistent year after year?