voidly
Forecast

Country shutdown risk

Where to
look closer.

Seven-day model estimates.
Compare countries. Inspect their evidence.

Country risk ranking

Source JSON

Compare countries. Exact shutdown dates remain unreliable.

Source snapshot

shutdown_risk_v9

60 country records have old or unconfirmed data. See the flags beside each date.

CountrySeven-day model estimate
  1. 01
    MyanmarAs of Older source data
    38.6%Elevated
  2. 02
    PakistanAs of Older source data
    17.3%Elevated
  3. 03
    JordanAs of Older source data
    6.7%Low
  4. 04
    IraqAs of Older source data
    5.4%Low
  5. 05
    SyriaAs of Older source data
    5.4%Low
  6. 06
    TanzaniaAs of Older source data
    5.4%Low
  7. 07
    VenezuelaAs of Older source data
    5.4%Low
  8. 08
    EthiopiaAs of Older source data
    4.8%Low
Highest 8 of 60 returned country estimates. Full 0–100% scale.Source-recommended elevated threshold: 14.0%.
All 60 country records
Same leaderboard, sorted by the model estimate. Component estimates are separate model inputs, not additional observations. Missing values remain unavailable.
CountryModel estimateSource bandAs of / freshnessStructural inputOONI anomaly input
MyanmarMM38.6%ElevatedAs of Older source data36.0%71.9%
PakistanPK17.3%ElevatedAs of Older source data14.8%78.1%
JordanJO6.7%LowAs of Older source data4.4%81.8%
IraqIQ5.4%LowAs of Older source data5.2%41.1%
SyriaSY5.4%LowAs of Older source data1.7%81.2%
TanzaniaTZ5.4%LowAs of Older source data5.8%54.5%
VenezuelaVE5.4%LowAs of Older source data1.7%77.2%
EthiopiaET4.8%LowAs of Older source data1.0%86.2%
BelarusBY3.5%LowAs of Older source data1.0%77.6%
NigeriaNG3.5%LowAs of Older source data1.0%82.6%
OmanOM3.5%LowAs of Older source data1.0%84.1%
RussiaRU3.5%LowAs of Older source data14.8%6.5%
GNGN2.6%LowAs of Older source data1.0%74.0%
IndiaIN2.4%LowAs of Older source data36.0%0.4%
KyrgyzstanKG2.4%LowAs of Older source data1.0%66.9%
IranIR1.2%LowAs of Older source data5.2%3.4%
BFBF0.8%LowAs of Older source data0.1%89.8%
UgandaUG0.8%LowAs of Older source data0.1%86.0%
UzbekistanUZ0.8%LowAs of Older source data0.1%87.0%
United Arab EmiratesAE0.5%LowAs of Older source data0.1%81.3%
AzerbaijanAZ0.5%LowAs of Older source data0.1%71.9%
EgyptEG0.5%LowAs of Older source data0.1%71.1%
FranceFR0.5%LowAs of Older source data0.1%77.7%
KazakhstanKZ0.5%LowAs of Older source data0.1%72.6%
Saudi ArabiaSA0.5%LowAs of Older source data0.1%60.6%
SudanSD0.5%LowAs of Older source data1.7%11.1%
TurkeyTR0.5%LowAs of Older source data4.4%0.6%
VietnamVN0.5%LowAs of Older source data1.0%7.9%
CubaCU0.4%LowAs of Older source data0.1%31.9%
BahrainBH0.2%LowAs of Older source data0.1%12.4%
SwitzerlandCH0.2%LowAs of Older source data0.1%9.7%
DenmarkDK0.2%LowAs of Older source data0.1%14.7%
SpainES0.2%LowAs of Older source data0.1%7.5%
FinlandFI0.2%LowAs of Older source data0.1%7.6%
IrelandIE0.2%LowAs of Older source data0.1%7.4%
KenyaKE0.2%LowAs of Older source data1.7%0.9%
NorwayNO0.2%LowAs of Older source data0.1%9.8%
New ZealandNZ0.2%LowAs of Older source data0.1%11.5%
TGTG0.2%LowAs of Older source data1.0%0.5%
AustriaAT0.1%LowAs of Older source data0.1%2.5%
BangladeshBD0.1%LowAs of Older source data0.1%1.0%
CanadaCA0.1%LowAs of Older source data0.1%3.3%
ChinaCN0.1%LowAs of Older source data1.0%0.3%
GermanyDE0.1%LowAs of Older source data0.1%6.4%
United KingdomGB0.1%LowAs of Older source data0.1%5.0%
IndonesiaID0.1%LowAs of Older source data1.0%<0.1%
JapanJP0.1%LowAs of Older source data0.1%4.9%
Sri LankaLK0.1%LowAs of Older source data0.1%2.7%
MLML0.1%LowAs of Older source data0.1%0.5%
NetherlandsNL0.1%LowAs of Older source data0.1%2.7%
PortugalPT0.1%LowAs of Older source data0.1%3.7%
SwedenSE0.1%LowAs of Older source data0.1%1.8%
TurkmenistanTM0.1%LowAs of Older source data1.0%<0.1%
United StatesUS0.1%LowAs of Older source data1.0%0.2%
ZimbabweZW0.1%LowAs of Older source data0.1%2.0%
ThailandTH0.1%LowAs of Older source data0.1%0.3%
ItalyIT0.1%LowAs of Older source data0.1%0.2%
TajikistanTJ0.1%LowAs of Older source data0.1%<0.1%
AustraliaAU0.0%LowAs of Older source data0.1%<0.1%
BelgiumBE0.0%LowAs of Older source data0.1%<0.1%

Per-country pages retain shutdown history, component details and source JSON. Open the complete response.

Page checked the data service at . Country “as of” dates refer to the source data, not this page visit.

The question
matters.

Comparing countries and calling the day of a shutdown are different jobs.

Inspect the model card

Forward validation · AUC

Country ranking

0.89

Within-country day timing

0.45

0.5 is the random-ranking reference. AUC measures ranking skill, not the percentage of correct predictions. Values come from the leaderboard’s evaluation fields.

Use the country comparison. Within-country day timing has not been established as reliable.

Follow the changes.

Choose a country and alert settings on the existing subscription page.

Set up risk alerts

Keep the score honest.

Inspect archived predictions, outcomes and the model’s own accountability record.

Inspect accountability

Follow it to the source.

Sources, scope and freshness

The shutdown-risk model uses Access Now KeepItOn STOP shutdown labels and OONI measurement signals. Country-level estimates are less informative about subnational events; sparse records also limit what can be learned.

“Source snapshot” is the response’s generation timestamp. Country dates, reported data age, and the source’s stale flags are separate. A fresh page request does not make old measurements fresh. Missing flags remain unconfirmed; dates more than three days old receive an additional warning.

Source score formula: isotonic(sigmoid(0.6*logit(shutdown_risk_v4_prob) + 0.40*logit(honest_forecast_v1_prob)))

Forward cross-country AUC
0.87
Forward cross-country precision–recall lift
9.42×

Country bands are copied from the leaderboard, not recalculated from an invented threshold. The table preserves zero estimates and distinguishes missing inputs from zero.

Related Atlas forecasts and case studies

Atlas Sentinel answers a different question with a different model. Its heatmap, validation and case studies are separate from the shutdown-risk ranking above. A worked example explains the alert mechanism; it does not establish this model’s day-timing skill.