voidly

PakistanPK

climbing (+53.2pp)

30-day day-by-day shutdown-risk trajectory. Each point is a per-horizon XGBoost prediction (h=1..30) with a 90% conformal band derived from absolute LOCO residuals. Peak risk highlighted.

Peak day
87.7%
d23 · 2026-07-01
Floor day
4.7%
d3 · 2026-06-11
30-day delta
+53.2pp
d30 vs d1 probability

30-day trajectory with 90% conformal band

model trajectory_v1
0%25%50%75%100%d1d7d14d21d30peak d23 88%
mean P(shutdown by day)90% conformal bandpeak day

Drivers (SHAP top 3 from the 7-day model — same features feed all horizons)

day 124.9%
2026-06-09
  • block_rate_roll7_std-0.043
  • recent_shutdown+0.031
  • incident_count_7d+0.023
day 738.6%
2026-06-15
  • block_rate_roll7_std-0.043
  • recent_shutdown+0.031
  • incident_count_7d+0.023
day 1472.1%
2026-06-22
  • block_rate_roll7_std-0.043
  • recent_shutdown+0.031
  • incident_count_7d+0.023
day 3078.1%
2026-07-08
  • block_rate_roll7_std-0.043
  • recent_shutdown+0.031
  • incident_count_7d+0.023

Honest caveats

  • Per-horizon XGBoost stacked into a trajectory — not a true joint distribution.
  • Conformal bands are absolute-residual q90 from LOCO predictions on 18 eval countries.
  • Long horizons (h>=14d) have wider effective uncertainty than the conformal q90 suggests; the same lag features feed all horizons.
  • Bands floor at [0,1]; probabilities below 0.05 are at the model's noise floor.
  • SHAP attribution above is from the 7-day point model. The trajectory uses 30 separate per-horizon XGBoost models that share the same feature set; we have not yet computed per-horizon SHAP.

Sample horizons

DayDateP(shutdown)90% interval
d12026-06-0924.9%[0%, 69%]
d72026-06-1538.6%[0%, 100%]
d142026-06-2272.1%[3%, 100%]
d302026-07-0878.1%[0%, 100%]
Spotlight countries