Re: compilation of pattern-matching?

When you tried switching Nil and Cons, did you try it on many examples? For a single example a 2-3% could be easily attributed to random effects like different instruction cache hit patterns. If you get it consistently over several programs then it seems likely to mean something, but I'm not sure what.

Agreed. There is a fairly consistent pattern over experiment/size variations (eg, different times for different orderings), but the time differences themselves are too small for useful interpretation. I would need a way to see what is really going on on the processor (I'm sure there are such things - are there any free ones?) to make sure there are no other effects interfering, or I might as well measure a random generator. Claus