KP · per-country backtest
North Korea forecast vs reality
Compare published forecasts with observed outcomes for this country.
Updated every 30 min · CC BY 4.0 · Raw JSON · Current forecast →
Forecast time series
Blue line: forecast probability. Green ✓ markers: forecast was right. Red ✗ markers: forecast was wrong. Dashed line at 0.5 is the binary decision threshold.
Y axis: forecast probability · Faint tick: distance to observed outcome (0 or 1)
Forecasts evaluated
163
since Apr 17
Accuracy @ 0.5
84.0%
137/163 correct
Brier score
0.002
lower is better
Observed positive rate
0%
mean predicted 4%
All 163 predictions (newest first)
| Eval date | Forecast | Pred ≥ 0.5? | Observed? | Correct? |
|---|---|---|---|---|
| Sep 26 | 4.7% | — | no event | ✓ |
| Sep 25 | 5.4% | ↑ | no event | ✗ |
| Sep 24 | 4.8% | — | no event | ✓ |
| Sep 23 | 4.4% | — | no event | ✓ |
| Sep 22 | 5.3% | ↑ | no event | ✗ |
| Sep 21 | 4.4% | — | no event | ✓ |
| Sep 20 | 3.7% | — | no event | ✓ |
| Sep 19 | 3.8% | — | no event | ✓ |
| Sep 18 | 4.1% | — | no event | ✓ |
| Sep 17 | 5.2% | ↑ | no event | ✗ |
| Sep 16 | 4.1% | — | no event | ✓ |
| Sep 15 | 4.7% | — | no event | ✓ |
| Sep 14 | 4.4% | — | no event | ✓ |
| Sep 13 | 4.7% | — | no event | ✓ |
| Sep 12 | 5.9% | ↑ | no event | ✗ |
| Sep 11 | 3.9% | — | no event | ✓ |
| Sep 10 | 4.9% | — | no event | ✓ |
| Sep 9 | 4.7% | — | no event | ✓ |
| Sep 8 | 4.5% | — | no event | ✓ |
| Sep 7 | 4.0% | — | no event | ✓ |
| Sep 6 | 4.5% | — | no event | ✓ |
| Sep 5 | 4.0% | — | no event | ✓ |
| Sep 4 | 4.8% | — | no event | ✓ |
| Sep 3 | 4.8% | — | no event | ✓ |
| Sep 2 | 4.5% | — | no event | ✓ |
| Sep 1 | 5.0% | ↑ | no event | ✗ |
| Aug 31 | 3.3% | — | no event | ✓ |
| Aug 30 | 3.9% | — | no event | ✓ |
| Aug 29 | 4.1% | — | no event | ✓ |
| Aug 28 | 4.7% | — | no event | ✓ |
| Aug 27 | 4.7% | — | no event | ✓ |
| Aug 26 | 4.2% | — | no event | ✓ |
| Aug 25 | 4.1% | — | no event | ✓ |
| Aug 24 | 3.9% | — | no event | ✓ |
| Aug 23 | 4.5% | — | no event | ✓ |
| Aug 22 | 3.8% | — | no event | ✓ |
| Aug 21 | 5.2% | ↑ | no event | ✗ |
| Aug 20 | 3.5% | — | no event | ✓ |
| Aug 19 | 5.3% | ↑ | no event | ✗ |
| Aug 18 | 4.2% | — | no event | ✓ |
| Aug 17 | 2.9% | — | no event | ✓ |
| Aug 16 | 4.5% | — | no event | ✓ |
| Aug 15 | 5.3% | ↑ | no event | ✗ |
| Aug 14 | 4.9% | — | no event | ✓ |
| Aug 13 | 5.4% | ↑ | no event | ✗ |
| Aug 12 | 4.0% | — | no event | ✓ |
| Aug 11 | 4.6% | — | no event | ✓ |
| Aug 10 | 4.8% | — | no event | ✓ |
| Aug 9 | 3.5% | — | no event | ✓ |
| Aug 8 | 3.7% | — | no event | ✓ |
| Aug 7 | 4.7% | — | no event | ✓ |
| Aug 6 | 4.4% | — | no event | ✓ |
| Aug 5 | 4.1% | — | no event | ✓ |
| Aug 4 | 3.5% | — | no event | ✓ |
| Aug 3 | 4.3% | — | no event | ✓ |
| Aug 2 | 4.0% | — | no event | ✓ |
| Aug 1 | 4.4% | — | no event | ✓ |
| Jul 31 | 4.0% | — | no event | ✓ |
| Jul 30 | 4.5% | — | no event | ✓ |
| Jul 29 | 3.2% | — | no event | ✓ |
How to read this
- Each row is one historical forecast. We made the prediction at eval_date, then waited the 7-day horizon, then graded against the observed outcome.
- Forecast % is the calibrated probability we published that day. Post-recalibration (2026-05-20) these now match actual observed rates much better — see the refit finding.
- Pred ≥ 0.5 shows the binary alert decision. Note we usually fire alerts at a lower threshold (see /v1/sentinel/global_heatmap for the live cutoff) — the 0.5 column here is for backtest scoring consistency with the global confusion matrix.
- Correct = both (pred ≥ 0.5) and (observed) agree.
Related
- /atlas/forecast/kp — current 7-day calibrated forecast for North Korea with SHAP drivers
- /kp — North Korea country page (current state, recent incidents)
- /sentinel/backtest — global reliability diagram + per-country comparison table
- Calibration refit writeup
About this view
Every Voidly Sentinel shutdown-risk forecast for North Korea, plotted against what actually happened. 163 (predicted, observed) pairs from the rolling 30-day evaluation window.