Called It!

Using AI Classifiers to Score a Person's Annual Predictions

Every holiday season, people make predictions about what will happen in the next year. The problem? Few people go back and evaluate how they did with objectivity. The "Called It!" skill uses best-in-class AI evaluation models to make this trivial.

Read the article on rickmanelius.com

How it works

How Called It! works: seven stages from discovering the predictions to a replayable scorecard
Seven stages. The three bars are written and reviewed before any evidence is fetched.

Already Evaluated Annual Predictions

company · 2026

Every

4%
strict
32%
standard
55%
lenient
50
gap, pts
22 claims · 16 open at strict · as of 2026-10-24
person · 2026

Scott Galloway

0%
strict
27%
standard
47%
lenient
47
gap, pts
15 claims · 8 open at strict · as of 2026-10-24
person · 2026

Mike Ippolito

10%
strict
35%
standard
45%
lenient
35
gap, pts
20 claims · 7 open at strict · as of 2026-10-24
person · 2026

Bob Doll

21%
strict
29%
standard
36%
lenient
14
gap, pts
14 claims · 5 open at strict · as of 2026-10-24
Hit rate at each bar, by subject
Hit rate at each bar. Unresolved claims count against the rate.
The generosity gap by subject
The generosity gap: lenient minus strict, in points.
All subjects, one workbook (.xlsx)The skill and every run folder on GitHub

Grade someone

Install the skill, point it at a prediction post, and a run folder comes out. Open a pull request with the folder and it joins this gallery.

# Claude Code
/plugin marketplace add rickmanelius/called-it
/plugin install called-it@called-it
/called-it "Name" 2026 https://link-to-their-predictions

# Codex and other Agent Skills hosts
npx skills add rickmanelius/called-it -g -y
/called-it "Name" 2026 https://link-to-their-predictions