One of the things we ought to know as technologists, is that measuring the easy thing instead of the right thing, can lead to entirely the wrong conclusions. Does it _mean_ they're the wrong conclusions? No. But it is _not_ evidence of them either.

I say "ought to know" but we still screw this up with monitoring and logging all the damn time.

Show thread

Gathering the wrong data, producing unactionable reporting, is often actively harmful. Worse than not doing it at all. We would be well served to remember this as we go to instrument our systems.

Show thread

If you say "I'll measure all the things, then look for patterns" I promise that you'll find them. But it doesn't mean they mean anything. In a big enough data set, some patterns will exist from random chance, no matter how unlikely.

@iarna I remember a website which goal was to find correlations in dataset that have nothing to do with each other. Things like sales of fridge in US create work in Germany or consuming more cheese in France meant dryer weather in Amazon

Sign in to participate in the conversation
Anarchism Space

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!