KZ · per-country backtest
Kazakhstan 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
41.1%
67/163 correct
Brier score
0.299
lower is better
Observed positive rate
24%
mean predicted 40%
All 163 predictions (newest first)
| Eval date | Forecast | Pred ≥ 0.5? | Observed? | Correct? |
|---|---|---|---|---|
| Sep 26 | 64.8% | ↑ | no event | ✗ |
| Sep 25 | 67.6% | ↑ | no event | ✗ |
| Sep 24 | 4.1% | — | no event | ✓ |
| Sep 23 | 72.6% | ↑ | no event | ✗ |
| Sep 22 | 70.1% | ↑ | shutdown | ✓ |
| Sep 21 | 57.8% | ↑ | shutdown | ✓ |
| Sep 20 | 57.8% | ↑ | shutdown | ✓ |
| Sep 19 | 51.4% | ↑ | shutdown | ✓ |
| Sep 18 | 51.0% | ↑ | shutdown | ✓ |
| Sep 17 | 60.3% | ↑ | no event | ✗ |
| Sep 16 | 50.4% | — | no event | ✓ |
| Sep 15 | 46.1% | — | no event | ✓ |
| Sep 14 | 58.5% | ↑ | no event | ✗ |
| Sep 13 | 52.9% | — | no event | ✓ |
| Sep 12 | 52.8% | ↑ | no event | ✗ |
| Sep 11 | 56.2% | ↑ | no event | ✗ |
| Sep 10 | 60.3% | ↑ | no event | ✗ |
| Sep 9 | 66.9% | ↑ | no event | ✗ |
| Sep 8 | 73.9% | ↑ | no event | ✗ |
| Sep 7 | 68.4% | ↑ | no event | ✗ |
| Sep 6 | 72.4% | ↑ | no event | ✗ |
| Sep 5 | 83.0% | ↑ | no event | ✗ |
| Sep 4 | 80.6% | ↑ | no event | ✗ |
| Sep 3 | 61.7% | ↑ | no event | ✗ |
| Sep 2 | 47.2% | — | no event | ✓ |
| Sep 1 | 64.2% | ↑ | no event | ✗ |
| Aug 31 | 63.4% | ↑ | no event | ✗ |
| Aug 30 | 48.8% | ↑ | no event | ✗ |
| Aug 29 | 50.6% | ↑ | no event | ✗ |
| Aug 28 | 58.1% | ↑ | no event | ✗ |
| Aug 27 | 55.4% | ↑ | no event | ✗ |
| Aug 26 | 73.5% | ↑ | no event | ✗ |
| Aug 25 | 57.5% | ↑ | shutdown | ✓ |
| Aug 24 | 49.5% | ↑ | shutdown | ✓ |
| Aug 23 | 37.3% | — | shutdown | ✗ |
| Aug 22 | 35.3% | — | shutdown | ✗ |
| Aug 21 | 39.2% | — | shutdown | ✗ |
| Aug 20 | 40.6% | — | no event | ✓ |
| Aug 19 | 40.1% | — | no event | ✓ |
| Aug 18 | 48.3% | — | no event | ✓ |
| Aug 17 | 55.3% | — | shutdown | ✗ |
| Aug 16 | 57.9% | ↑ | shutdown | ✓ |
| Aug 15 | 61.6% | ↑ | shutdown | ✓ |
| Aug 14 | 65.6% | ↑ | shutdown | ✓ |
| Aug 13 | 71.1% | ↑ | no event | ✗ |
| Aug 12 | 74.0% | ↑ | no event | ✗ |
| Aug 11 | 68.2% | ↑ | shutdown | ✓ |
| Aug 10 | 69.7% | ↑ | shutdown | ✓ |
| Aug 9 | 71.7% | ↑ | shutdown | ✓ |
| Aug 8 | 80.4% | ↑ | shutdown | ✓ |
| Aug 7 | 76.1% | ↑ | shutdown | ✓ |
| Aug 6 | 76.6% | ↑ | no event | ✗ |
| Aug 5 | 68.2% | ↑ | no event | ✗ |
| Aug 4 | 76.8% | ↑ | shutdown | ✓ |
| Aug 3 | 75.4% | ↑ | shutdown | ✓ |
| Aug 2 | 67.9% | ↑ | shutdown | ✓ |
| Aug 1 | 60.4% | ↑ | shutdown | ✓ |
| Jul 31 | 58.8% | ↑ | shutdown | ✓ |
| Jul 30 | 56.4% | ↑ | no event | ✗ |
| Jul 29 | 37.0% | ↑ | 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/kz — current 7-day calibrated forecast for Kazakhstan with SHAP drivers
- /kz — Kazakhstan 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 Kazakhstan, plotted against what actually happened. 163 (predicted, observed) pairs from the rolling 30-day evaluation window.