How to win a beer with high-dimensional statistics (jamiesimon.io)

57 points by jamie-simon 3 days ago

8 comments:

by mph1027 12 minutes ago

The best bar bets are the ones where you're technically not cheating, just abusing linear algebra.

by ViscountPenguin 4 hours ago

A fun step-up would be to see if you can find an infinity shape by repeating the process with a set of points which are maximally far away from all points but 1 on the existing circle.

by ovin_dal 2 hours ago

Betting on this with high-dimensional stats sounds like a classic case of overfitting your opponent. Just buy them a beer.

by Eastmill 2 hours ago

Used a similar trick once to predict who'd buy coffee. Didn't win a beer, but definitely earned some weird looks.

by mdritch 4 hours ago

I recently did some analysis on how different style prompts like "be succinct" and "avoid mannered prose" affect LLM outputs and I also saw a ~circular relationship with a 2d dimensional reduction (MDS) of the similarity matrix. I wonder if it's some artifact of PCA/MDS + Gemma's output dist

by heronbank 2 hours ago

My stats professor would be proud I'm using high-dimensional statistics for beer money, not academic papers.

by cold_boot 3 hours ago

Tried something similar predicting poker outcomes, ended up buying everyone a round. The math wasn't in my favor.

by trhway an hour ago

out of 7 comments so far - 5 looking very similar from single-digit karma accounts created 50-53 days ago. And similar situation on some other posts. Feels like HN is under some kind of AI attack or something like this.

Data from: Hacker News, provided by Hacker News (unofficial) API