Stvari koje zaista ponovo čitam

Cormac Kelleher

@statlinesunday

4Items

0Likes

These are the papers I keep coming back to when the match’s over, the pint’s half gone, and someone says “yeah but is xG really any good?” It’s mostly sport, lots of crowds, and a couple of proper nerd detours into how we mash opinions together. Nothing too abstract — every one of these has at least one number or idea you can drag straight into an argument about why your manager’s substitutions are chaos.

See all 4 items Follow Cormac KelleherOpen in app

The list

Showing 1–4 of 4

Stvari koje zaista ponovo čitam — 4 items
Name	Year	Sport / Domain	One big idea	Good for	My take
1The Hot Hand in Basketball: On the Misperception of Random Sequences	1985	Basketball / cognitive bias	What looks like a shooting streak can just be random noise dressed up as momentum.	Killing pub myths about “form” over a small run of games or shots.	Read this once and you’ll never trust a commentator who says “he just wanted it more” after three shots in a row.
2The Price of Anarchy in Basketball	2010	Basketball tactics / network models	If every possession chases the best shot for itself, the whole offence can get worse overall.	Thinking about why hero‑ball attacks and overused stars can tank team efficiency.	Explains in equations what every annoyed fan yells during a 5‑out ISO: the good shot isn’t always the selfish obvious one.
3Goal‑Line Oracles: Exploring Accuracy of Wisdom of the Crowd for Football Predictions	2024	Premier League / xG forecasting	Aggregated fan predictions track xG and results pretty well, but bias creeps in for the flashy big‑name teams.	Checking if your fan poll or Discord prediction channel is more than just vibes.	Basically shows the crowd’s decent over 38 rounds, but still has a soft spot for the usual suspects — just like every pub in Dublin on a Champions League night.
4Wisdom of the Crowds Forecasting the 2018 FIFA Men’s World Cup	2020	World Cup / probability forecasts	Aggregating thousands of fan probability picks can rival proper statistical models over a full tournament.	Building office‑sweep models or online leagues where everyone submits scorelines, not just banter.	If you’ve ever stared at a predictions leaderboard wondering if it’s just luck, this one gives you the math behind the madness.