Research surface
Country × Country censorship correlation
Pearson correlation across 365-day daily incident vectors for the top 50 most-censored countries (by Atlas Score v2). Pairs that fire together — regional shutdown waves, regime- similar blocking patterns — light up here the day they emerge.
Computed at request time over 2,029 incidents. 1-hour ISR. Source ranking →
Want pattern similarity that allows time-shifted phases (lead / lag)? See the DTW cohort clusters →
Want directional time-shifted pairs (“A leads B by N days”) with FDR-corrected p-values? See the lead/lag pairs →
Pearson r:
−1
−0.5
0
+0.5
+1
Hover any cell for the exact correlation. Diagonal masked (always r = 1). Countries with fewer than two incidents in the window appear as flat zeros — read the cell-count caveat below before interpreting weak signal.
Top 20 most-correlated pairs
- 01AE × THr = 0.889strong co-move between AE and Thailand.
- 02SG × AEr = 0.884strong co-move between SG and AE.
- 03QA × SGr = 0.861strong co-move between QA and SG.
- 04SA × THr = 0.844strong co-move between Saudi Arabia and Thailand.
- 05JO × BYr = 0.841strong co-move between JO and Belarus.
- 06SG × THr = 0.840strong co-move between SG and Thailand.
- 07BY × THr = 0.828strong co-move between Belarus and Thailand.
- 08AE × BYr = 0.826strong co-move between AE and Belarus.
- 09MY × KHr = 0.823strong co-move between Malaysia and KH.
- 10AE × SAr = 0.816strong co-move between AE and Saudi Arabia.
- 11QA × THr = 0.814strong co-move between QA and Thailand.
- 12QA × AEr = 0.814strong co-move between QA and AE.
- 13AE × VNr = 0.810strong co-move between AE and Vietnam.
- 14JO × KHr = 0.808strong co-move between JO and KH.
- 15BY × SAr = 0.806strong co-move between Belarus and Saudi Arabia.
- 16JO × SGr = 0.801strong co-move between JO and SG.
- 17SG × SAr = 0.795strong co-move between SG and Saudi Arabia.
- 18BY × KHr = 0.792strong co-move between Belarus and KH.
- 19QA × MYr = 0.789strong co-move between QA and Malaysia.
- 20SG × BYr = 0.781strong co-move between SG and Belarus.
Method & honest caveats
- Window: 365 days ending today (2026-07-06). Vectors are daily incident counts per country.
- Countries: top 50 by Atlas Score v2 (50% structural baseline + 50% recent activity).
- Statistic: Pearson correlation. No significance test surfaced here — small N-days mean even high r can be noise. Treat this as an exploratory surface, not a confirmatory one.
- The interpretation column above uses a hard-coded regional peer map. A "regional cluster" tag is a hint, not a causal claim.
- Diagonal is masked. Negative r is rendered blue; positive red; near-zero gray.
Cite as: Voidly Atlas Correlation Matrix, 2026-07-06, voidly.ai/atlas/correlation-matrix. CC BY 4.0.