Someone asked me at the end of the year if we'll have a performance review. My reaction was "why would I ever do that?". He said "for feedback and all..". I said "well, we've been doing that for years with our daily conversations" and cited a few instances of remarks and suggestions to refine our machine that he and I made. While making coffee and you tell me "do you think it'd be better if we did X" or "what do you think of Y" and then we try that for some time and then refine it? That's feedback and optimisation and tweaking. When I see your implementation of something and how it eliminated a whole family of problems and say it's cool and let's try that for other things and it works or vice versa? That's feedback and optimisation.

He said that his friends at other companies are going through them and he figured we'd be doing that as well. I explained to him why it doesn't make any sense in our situation.

Yes, for one specific person, I had to wait for about a year to assess because the period of the observed thing was that long and I had to take seasonality into account, but other than that, no thanks. When I talked with the person, the person said they meant to talk to me about it and ask for help but they couldn't muster the courage. But I had to wait and not react because the observed process was slow and I had to have a large time window to get enough data.

Then again, maybe my background in control theory and feedback systems prevents me from doing annual performance reviews because they don't make sense to me, and being human prevents me from engaging in a practice that maximizes passive aggressive resentful behavior.

In my opinion, doing the annual thing may be robbing everyone of a year of growth to the power of the number of entities involved. If you have a system made of subsystems who depend on each other's inputs and outputs, it would be insanity not to correct as early as possible, but not earlier to account for hysteresis and all.