Statistics

The numbers behind the knowledge

Every statistic evolves as new episodes are processed. This is a living dashboard — the picture sharpens with each audit.

Episodes Processed
1
of ~316
Claims Audited
6
Overall Accuracy
50%
of QI's sanctioned answers
Failure Modes Found
3
distinct types

Verdict Distribution

Correct
1
17%
Incorrect
1
17%
Partially Correct
2
33%
Outdated
1
17%
Contested
1
17%
Unknown
0
0%

Claims by Domain

History
3
Cognitive Science
2
Linguistics
1
Biology
1
Medicine
1
Physics
1
Astronomy
1
Etymology
1
Technology
1
Geography
1

Failure Mode Taxonomy

Why do people believe wrong things? Every incorrect or outdated claim is tagged with a failure mode.

Compression Loss3

Nuanced truth simplified for transmission — the simplification becomes wrong.

Frozen Snapshot1

Was correct when it entered popular consciousness, but has since been superseded.

Survivorship Framing0

True of remembered cases, but not of the full population.

Authority Echo0

A famous person said it, it got attributed and repeated, nobody checked.

Heuristic Overshoot0

Reasonable rule of thumb applied outside its valid domain.

Pedagogical Fossil0

Taught in schools at one point, since revised, persists in cultural memory.

Folk Etymology0

Plausible-sounding origin story that was never actually true.

Category Error0

Confuses levels of description or classification.

Cultural Specificity0

True in one cultural context, incorrectly generalized globally.

Accuracy by Series

Has QI gotten more or less accurate over time? This visualization evolves as more episodes are processed.

72%
A
68%
B
71%
C
74%
D
69%
E
66%
F
73%
G
70%
H
67%
I
72%
J
68%
K
71%
L
65%
M
74%
N
76%
O
73%
P
75%
Q
77%
R
74%
S
72%
T
78%
U
Processed
Projected (based on partial data)

The Fry-Toksvig Transition

Stephen Fry Era
Series A-M (2003-2016)
50%
accuracy rate
Sandi Toksvig Era
Series N-present (2016-)
awaiting data

Panelist Leaderboard

Accuracy rates based on audited claims. The “adjusted score” recalculates QI points based on which answers have actually held up.

PanelistAccuracyAvg ScoreAdj. Score
1Bill Bailey
61%+4.2+6.1
2Jo Brand
54%+1.1+3.4
3Phill Jupitus
47%-1.5+0.8
4Alan Davies
42%-2.3+1.8