Crit Crit Crit

Don't make ASSUMPTIONS about development. You can provide POSSIBILITIES, but don't present them as definitive explanations or justifications. Don't use invented development issues, which are only *possibilities*, as defense of bad design. Once it's a final product, problems are problems, even if they're understandable. Don't be so sympathetic.
In Crash 1, the physics *were bad*. They were not *necessarily* hard to program - I know nothing about the programming process of Crash's physics - but they were *definitely* poor. Analyze what's readily apparent. Don't make unsubstantiated claims, about development OR otherwise.

Give EVIDENCE for EVERY claim.

Don't make assumptions about others' experiences, use specific personal anecdotes.

Don't be nice to a game just because people like it. You won't be nice to Uncharted, so why were you to Crash?

What makes an element trivial or difficult? Don't dismiss stuff - go in-depth on everything or don't bring it up at all.

ALWAYS try to give constructive ideas as to how a bad aspect of a game could have been done better, preferably with contemporary examples that didn't have the benefit of hindsight. This requires you to ARGUE WITH YOURSELF.