Evidence → judgment → production choices

Research comparison

Check the scope before drawing a lesson. Put two published samples side by side, examine their evidence and limits, then follow the full analysis into production choices.

Different measures do not establish a winner

Each sample retains its own platform, denominator and time window. Equal sample sizes do not make a controlled comparison. No combined ranking or cross-sample growth rate is calculated.

Sample 1Sample 2

4 published samples · Bookmark this page to keep your selection

Sample 1

Weekly chart: 40-title sample

Study coverage: English-language sample
Source: GASP
Verified: 13 Oct 2026, 15:26 UTC+8

Weekly chart: 40-title sample

2026-10-04 – 2026-10-11 · Sample denominator: 40 · titles

Bar scale: 0–40 titles

Chart data
Item	titles
Wolf-related words	9
Divorce / CEO / wealth union	4

40 titles; keyword coding. Categories may overlap; observed overlap is zero.

A weekly chart sample, not revenue or audience-gender shares.

Data source 1

The production question

Matched samples of 40 titles yield wolf signals of 9 versus 2 and divorce/CEO/wealth signals of 4 versus 8. Classification counterexamples and three catalog leads show what to investigate before production.

Read analysis & production choices →Check the full method & limits →

Sample 2

Cumulative chart: matching 40-title denominator

Study coverage: English-language sample
Source: GASP
Verified: 13 Oct 2026, 15:26 UTC+8

Cumulative chart: matching 40-title denominator

2026-10-12 – 2026-10-12 · Sample denominator: 40 · titles

Bar scale: 0–40 titles

Chart data
Item	titles
Wolf-related words	2
Divorce / CEO / wealth union	8

Same coding, 40 titles. Dates indicate the source check label; lifetime totals have no shared start date.

Differences from the weekly chart are not growth rates or controlled genre comparisons.

Data source 1

The production question

Read analysis & production choices →Check the full method & limits →

ShorStory Current editorial team · Samples come from published research. Verification dates do not imply continuous data updates. Editorial standards