CainIV - 10:33 pm on Jan 6, 2010 (gmt 0)

Yeah, but if they changed the algo we should feel it

How do we know this conclusively?

How do we know that there will be results that are different in some way?

How do we know, after this long, that results that people see and claim are 'Caffeine' results are in fact those results.

Even when the original ip was given by Matt as a reference, the rankings there were in flux, as they should be on a dynamic system. It only stands to reason that this would still be the case, and if so, would make it very difficult for anyone to decisively say 'I seen Caffeine on DC x,y or z'