Pre-Registered Predictions — November 2026 Releases
Written: 2026-10-02 (frozen before every release it targets; the commit is the tamper-evidence). Context: research/20. Rows #29–39 (frozen 2026-09-19) are unchanged and still pending.
Ranges come from calibration.json floors: local price ±$65 (and ≥21.7% of point), state price ±$30 (and ≥12.5%). Each numeric row records its mechanical anchor and the judgment adjustment, so the judgment layer is graded separately.
NASS Agricultural Prices, Mon Nov 30, 2026 (OCTOBER prices) — family state_price_usd_ton
| # | Cluster | Item | Anchor | Adj | Point | Range | Reasoning |
|---|---|---|---|---|---|---|---|
| 40 | A | CO alfalfa, October (state avg) | $235 (Aug persistence) | +10 | $245 | $214–276 | state grid Oct; rising-analog base rate (discretionary); cash complex $260–330 |
| 41 | C | CO all-hay, October | $231 (Aug persistence) | +9 | $240 | $210–270 | tracks #40 minus ~$5 |
| 42 | B | KS alfalfa, October | $127 (Aug persistence) | 0 | $127 | $97–157 | last cycle's +$30 judgment on Kansas (#19) missed by $28; the KS average does not chase cash |
AMS Colorado Direct Hay Report — Oct 22, Nov 5, Nov 19
| # | Cluster | Item | Anchor / base rate | Adj | Point | Range | Reasoning |
|---|---|---|---|---|---|---|---|
| 43 | A | A SE CO FOB-farm Good–Supreme large-square alfalfa trade ≥100 t prints in Oct 22, Nov 5 or Nov 19 | 6 of 243 reports (2.5% each, ~7% over three) | — | NO | categorical | base rate; the 2022 analog's 1,000-t lot printed Nov 17, so this is not zero. LOW-INFORMATION: ~93% likely by base rate alone; reported with and without |
| 44 | A | IF any CO region (non-organic) prints a FOB-farm/ranch per-ton alfalfa trade ≥500 t in Nov 5/Nov 19: tonnage-weighted $/ton | $260 (Sep 10 NE 2,000 t Good — the season's only such print) | +15 | $275 | $210–340 | comp-set anchors $262–308 (median $273.5); NO-PRINT if none |
US Drought Monitor, Thu Nov 5, 2026 (map valid Nov 3)
| # | Cluster | Item | Anchor | Adj | Point | Range | Reasoning |
|---|---|---|---|---|---|---|---|
| 45 | C | Bent County (08011) D1–D4 area ≤ 10% | Sep 29: D1 48%, D2+ 0% | — | YES | categorical (NEAR if 10–25%) | 3.1–4.5" fell Sep 30–Oct 1 after the Sep 29 map; CPC drought outlook = removal at McClave; 6–14 day outlook leans dry, the risk to this call. LOW-INFORMATION: near-certain on the evidence; reported with and without |
Scoring protocol
Unchanged (frozen 2026-08-08, predictions-2026-08-13.md §protocol): HIT / NEAR (numeric: within 5% of nearest edge; categorical: adjacent label) / MISS / NO-PRINT. Conditional row #44 grades NO-PRINT if the condition never occurs. Clusters: A: #40, #43, #44 · B: #42 · C: #41, #45. Source specs for every row are in data/current/prediction_sources.json at this commit (#40–42 auto-grade via grade_predictions.py, month pinned to OCTOBER; the rest grade by hand and are verified by --check). Rows are never edited after release.
Pre-publication review (2026-10-02), before any target release: a draft row #46 (weekly Niño-3.4 +3.2, range +1.9 to +4.5 for the Nov 4 week) was withdrawn as padding: a 1.3°C move in five weeks is very rare, so the row was a near-free numeric HIT that would inflate interval coverage. #43 and #45 are kept as categorical rows (they do not enter the X-of-Y coverage count) and are flagged low-information. The state-price floor (±$30) was challenged by the reviewer (2× the worst state-row error is now ~$56); it is kept under its written trigger and the disagreement is recorded in calibration_floors.json.