Hacker News new | past | comments | ask | show | jobs | submit login

I think the lesson he’s teaching is that examination should be hollistic, not just focus on specific tests. Different tests are appropriate in different situations and they often give conflicting results.

I’m no vet but if you have to dig hard for signs of illness, the patient likely isn’t sick. Our bird had increased uric acid. Could be anything. It was a little high.

If he was actua sick we wouldn’t be guessing whether the number is high. It’s be 10x. Waaay out of bounds. A clear signal.

It’s like in product design. Your conversion rate inproved 0.05% after you ran an experiment for 2 days. Is that a signal or noise? Eh probably just noise. Observe longer.

The problem with comfortably saying the client is wrong is tht they’re the authority figure. They have years of data, you have 30 minutes with the patient.

And yes I have had vets casually chat with me for 2 hours when they did spot something suspicious in the bird’s behavior and wanted to observe longer to see if it’s a pattern.




Consider applying for YC's Summer 2025 batch! Applications are open till May 13

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: