Tires: Tire tests and changing compounds

Question
I'm wondering how common it is for manufactures to make significant tread compound changes to otherwise identical tires -- that is, same model and
what appears on pretty close inspection to be the same tread pattern (including siping amount, shape, etc).

I ask because I'm trying to understand the tire ranking changes in the new Consumer Reports of 6 AT tires (out of 15) that were previously tested 3
years ago and appear to be unchanged. I could understand the entire new list being graded on a curve to avoid the difficulties of trying to precisely
duplicate past test conditions, but that wouldn't explain the 6 tires as they have been realigned relative to each other. Among the 6, there were two
instances (separate tires) of a 2-notch ranking drop in objective traction categories, so imprecise data that drifted over the edges of ranking
assignments wouldn't seem to work as an explanation either. I looked for a trend based on  improved rolling resistance but, ironically, that column was
the least affected. Two attributes which saw a lot of change - all for the worse - were wet braking and snow traction. None of this seems to follow
any pattern including making much sense via the triangle of tire design (thank you). But focusing only on the 6 tire relative realignment, something that would explain it is finding out that tread compound changes are not uncommon in otherwise unchanged tires.

And since we're on the subject, how reliable do you personally consider the CR tire tests to be?  

Thanks

Answer
Jeff,

The first thing you have to understand is that tire testing is highly variable.  The test results can vary considerably, even though nothing else has changed.  That is why tire manufacturers tend to rely on laboratory testing, which is much more precise than on-vehicle testing.  Rolling resistance is one such lab test.

Second, for the same type of tire, most are grouped pretty closely for any particular parameter.  This means that the test variability is going to play a large role in how things get ranked.

Third, it is common for tire manufacturers to change things over time.  

I could go on, but the bottomline is that tire testing doesn't always yield good, consistent results.  This is why tire manufacturers spend a great deal of time and money in research.  

Over 20 years ago, Consumer Reports started doing tire testing - and their first results were awful.  They obviously had not consulted with the tire manufacturers as they mixed different types of tires, performed tests that were pretty much meaningless, and after they published the results and were called on them, were very arrogant about the validity of their results.  I gave them a couple more testing cycles to see if they would improve - and they did not.

Then they quit publishing their procedures.  So I do not have a lot of confidence that they can conduct valid, comparable tire testing.  Your analysis seems to bear this out.  I consider their testing to be just part of the picture people should look at when buying tires.