{"count":30,"generated_at":"2026-10-04T16:12:11.715059","methodology":"Composite formula: 40% forecast + 25% incident density + 20% trend + 10% calibration + 5% anomaly. Grade A (0-20) = open internet, Grade F (80-100) = censorship-heavy.","results":[{"components":{"anomaly":2.5,"calibration":10.0,"forecast":38.0,"incident_density":0.0,"trend":10.0},"country":"PK","country_name":"Pakistan","grade":"C","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.95,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":6,"interval_90":[0.6,1.0],"key_drivers":[],"trend_7d_delta":0.0},"score":60.5},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":25.88,"incident_density":0.0,"trend":10.45},"country":"UZ","country_name":"Uzbekistan","grade":"B","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.647,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":3,"interval_90":[0.29700000000000004,0.997],"key_drivers":[],"trend_7d_delta":0.0223},"score":48.8},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":18.6,"incident_density":4.85,"trend":10.0},"country":"RU","country_name":"Russia","grade":"B","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.465,"historical_outcomes":164,"incidents_24h":1,"incidents_7d":14,"interval_90":[0.11500000000000005,0.815],"key_drivers":[],"trend_7d_delta":0.0},"score":46.0},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":22.04,"incident_density":0.0,"trend":10.0},"country":"EG","country_name":"Egypt","grade":"B","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.551,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":7,"interval_90":[0.20100000000000007,0.901],"key_drivers":[],"trend_7d_delta":0.0},"score":44.5},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":10.88,"incident_density":4.85,"trend":10.0},"country":"IN","country_name":"India","grade":"B","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.272,"historical_outcomes":164,"incidents_24h":1,"incidents_7d":6,"interval_90":[0.0,0.622],"key_drivers":[],"trend_7d_delta":0.0},"score":38.2},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":10.8,"incident_density":4.85,"trend":10.0},"country":"VE","country_name":"Venezuela","grade":"B","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.27,"historical_outcomes":164,"incidents_24h":1,"incidents_7d":7,"interval_90":[0.0,0.62],"key_drivers":[],"trend_7d_delta":0.0},"score":38.1},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":15.32,"incident_density":0.0,"trend":10.0},"country":"BR","country_name":"Brazil","grade":"B","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.383,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":1,"interval_90":[0.03300000000000003,0.733],"key_drivers":["Brazilian General Election 2026"],"trend_7d_delta":0.0},"score":37.8},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":11.12,"incident_density":0.0,"trend":10.0},"country":"CN","country_name":"China","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.278,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":8,"interval_90":[0.0,0.628],"key_drivers":[],"trend_7d_delta":0.0},"score":33.6},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":10.96,"incident_density":0.0,"trend":10.0},"country":"MY","country_name":"Malaysia","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.274,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":2,"interval_90":[0.0,0.624],"key_drivers":[],"trend_7d_delta":0.0},"score":33.5},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":10.72,"incident_density":0.0,"trend":10.0},"country":"IR","country_name":"Iran","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.268,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":12,"interval_90":[0.0,0.618],"key_drivers":[],"trend_7d_delta":0.0},"score":33.2},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":10.56,"incident_density":0.0,"trend":10.0},"country":"MM","country_name":"Myanmar","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.264,"historical_outcomes":163,"incidents_24h":0,"incidents_7d":9,"interval_90":[0.0,0.614],"key_drivers":[],"trend_7d_delta":0.0},"score":33.1},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":10.28,"incident_density":0.0,"trend":10.0},"country":"TR","country_name":"Turkey","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.257,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":3,"interval_90":[0.0,0.607],"key_drivers":[],"trend_7d_delta":0.0},"score":32.8},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":10.08,"incident_density":0.0,"trend":10.0},"country":"BD","country_name":"Bangladesh","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.252,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":6,"interval_90":[0.0,0.602],"key_drivers":[],"trend_7d_delta":0.0},"score":32.6},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.8,"incident_density":4.85,"trend":10.0},"country":"ID","country_name":"Indonesia","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.045,"historical_outcomes":164,"incidents_24h":1,"incidents_7d":4,"interval_90":[0.0,0.39499999999999996],"key_drivers":[],"trend_7d_delta":0.0},"score":29.1},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.56,"incident_density":4.85,"trend":10.0},"country":"NG","country_name":"Nigeria","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.039,"historical_outcomes":164,"incidents_24h":1,"incidents_7d":10,"interval_90":[0.0,0.38899999999999996],"key_drivers":[],"trend_7d_delta":0.0},"score":28.9},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.4,"incident_density":4.85,"trend":10.0},"country":"LB","country_name":"Lebanon","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.035,"historical_outcomes":164,"incidents_24h":1,"incidents_7d":9,"interval_90":[0.0,0.385],"key_drivers":[],"trend_7d_delta":0.0},"score":28.8},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.28,"incident_density":4.85,"trend":10.0},"country":"SY","country_name":"Syria","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.032,"historical_outcomes":164,"incidents_24h":1,"incidents_7d":6,"interval_90":[0.0,0.382],"key_drivers":[],"trend_7d_delta":0.0},"score":28.6},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.84,"incident_density":0.0,"trend":12.63},"country":"SD","country_name":"Sudan","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.046,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":0,"interval_90":[0.0,0.39599999999999996],"key_drivers":[],"trend_7d_delta":0.1315},"score":27.0},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":2.12,"incident_density":0.0,"trend":10.0},"country":"SA","country_name":"Saudi Arabia","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.053,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":2,"interval_90":[0.0,0.40299999999999997],"key_drivers":[],"trend_7d_delta":0.0},"score":24.6},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.96,"incident_density":0.0,"trend":10.0},"country":"CU","country_name":"Cuba","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.049,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":4,"interval_90":[0.0,0.39899999999999997],"key_drivers":[],"trend_7d_delta":0.0},"score":24.5},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":2.04,"incident_density":0.0,"trend":10.0},"country":"PH","country_name":"Philippines","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.051,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":6,"interval_90":[0.0,0.40099999999999997],"key_drivers":[],"trend_7d_delta":0.0},"score":24.5},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.8,"incident_density":0.0,"trend":10.0},"country":"KP","country_name":"North Korea","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.045,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":0,"interval_90":[0.0,0.39499999999999996],"key_drivers":[],"trend_7d_delta":0.0},"score":24.3},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.76,"incident_density":0.0,"trend":10.0},"country":"TH","country_name":"Thailand","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.044,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":1,"interval_90":[0.0,0.39399999999999996],"key_drivers":[],"trend_7d_delta":0.0},"score":24.3},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.6,"incident_density":0.0,"trend":10.0},"country":"KZ","country_name":"Kazakhstan","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.04,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":2,"interval_90":[0.0,0.38999999999999996],"key_drivers":[],"trend_7d_delta":0.0},"score":24.1},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.56,"incident_density":0.0,"trend":10.0},"country":"TM","country_name":"Turkmenistan","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.039,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":0,"interval_90":[0.0,0.38899999999999996],"key_drivers":[],"trend_7d_delta":0.0},"score":24.1},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.68,"incident_density":0.0,"trend":9.92},"country":"NI","country_name":"Nicaragua","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.042,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":1,"interval_90":[0.0,0.39199999999999996],"key_drivers":[],"trend_7d_delta":-0.004},"score":24.1},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.44,"incident_density":0.0,"trend":10.0},"country":"BY","country_name":"Belarus","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.036,"historical_outcomes":163,"incidents_24h":0,"incidents_7d":5,"interval_90":[0.0,0.38599999999999995],"key_drivers":[],"trend_7d_delta":0.0},"score":23.9},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.36,"incident_density":0.0,"trend":10.0},"country":"VN","country_name":"Vietnam","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.034,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":4,"interval_90":[0.0,0.384],"key_drivers":[],"trend_7d_delta":0.0},"score":23.9},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.08,"incident_density":0.0,"trend":10.0},"country":"ER","country_name":"Eritrea","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.027,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":0,"interval_90":[0.0,0.377],"key_drivers":[],"trend_7d_delta":0.0},"score":23.6},{"components":{"anomaly":2.5,"calibration":10.0,"forecast":1.92,"incident_density":0.0,"trend":7.06},"country":"ET","country_name":"Ethiopia","grade":"B-","interpretation":"Higher score = more censorship risk. Score is a composite of (40%) calibrated 7-day forecast + (25%) log-scaled recent incident density + (20%) 7-day forecast trend delta + (10%) calibration-quality discount + (5%) anomaly disagreement. Grade A = 0-20 (open), B = 20-50, C = 50-65, D = 65-80, F = 80-100 (censorship-heavy).","raw_signals":{"anomaly_note":"neutral score; unsupervised model is experimental","forecast_max_risk":0.048,"historical_outcomes":164,"incidents_24h":0,"incidents_7d":8,"interval_90":[0.0,0.39799999999999996],"key_drivers":["Tigray War Anniversary 2026 in 25d","Tigray War Anniversary 2026 in 27d","Tigray War Anniversary 2026 in 28d"],"trend_7d_delta":-0.1472},"score":21.5}],"schema":"voidly-atlas-score-leaderboard/v1"}