The questions — Era 3

Every question here was registered before its answer was knowable, names the registry that settles it, and is graded by code against a field — never by a model reading prose. frozen spec era3-v1 · hash 5f736e8cd424 · frozen 2026-08-02.

Era 3 · 6 of 100 storylines resolved. May not claim edge before the gate. Macro storylines co-move — n is a storyline count, not a claim of independence (decision 0014). Era 2 closed with gate unreached (42 of 100) — no edge claim.

Null-copy rate: 32% of 725 forecast states within ±1pp of the null (echoing the null scores zero edge by construction).

Power check: at n=100 the gate can detect a true edge of ≥0.032 Brier. Era 2's only real evidence — 42 resolved storylines, a different method — showed +0.0166 Brier (σ=0.124), indistinguishable from zero. The gate can find an edge if era 3 has one; it cannot manufacture one from a null.

The open book

A forecast is a living probability, not a one-shot call. Each question is scored at daily checkpoints against the null beside it, weighted so early conviction counts for more. Copying the null scores exactly zero edge.

Question	Workshop	The null to beat	Resolves
PAYEMSWill FRED PAYEMS (nonfarm payrolls) for 2026-09 print above its prior reading of 158984.0 (as of 2026-06-01)? fred · 50 states	95%+20 vs null	75% computed base rate, K=24	2026-10-08
UNRATEWill FRED UNRATE (unemployment rate) for 2026-09 print above its prior reading of 4.2 (as of 2026-06-01)? fred · 50 states	35%-3 vs null	38% computed base rate, K=24	2026-10-08
PPIACOWill FRED PPIACO (PPI (all commodities)) for 2026-09 print above its prior reading of 286.827 (as of 2026-06-01)? fred · 50 states	78%+16 vs null	62% computed base rate, K=24	2026-10-17
RSAFSWill FRED RSAFS (retail sales) for 2026-09 print above its prior reading of 768553.0 (as of 2026-06-01)? fred · 50 states	87%+16 vs null	71% computed base rate, K=24	2026-10-18
INDPROWill FRED INDPRO (industrial production) for 2026-09 print above its prior reading of 102.6395 (as of 2026-06-01)? fred · 50 states	62%+12 vs null	50% computed base rate, K=24	2026-10-19
HOUSTWill FRED HOUST (housing starts) for 2026-09 print above its prior reading of 1427.0 (as of 2026-06-01)? fred · 50 states	50%-4 vs null	54% computed base rate, K=24	2026-10-20
PAYEMSWill FRED PAYEMS (nonfarm payrolls) for 2026-10 print above its prior reading of 158858.0 (as of 2026-07-01)? fred · 22 states	73%+2 vs null	71% computed base rate, K=24	2026-11-08
UNRATEWill FRED UNRATE (unemployment rate) for 2026-10 print above its prior reading of 4.1 (as of 2026-07-01)? fred · 22 states	35%+2 vs null	33% computed base rate, K=24	2026-11-08
PPIACOWill FRED PPIACO (PPI (all commodities)) for 2026-10 print above its prior reading of 284.057 (as of 2026-07-01)? fred · 22 states	60%+2 vs null	58% computed base rate, K=24	2026-11-17
RSAFSWill FRED RSAFS (retail sales) for 2026-10 print above its prior reading of 763602.0 (as of 2026-07-01)? fred · 22 states	85%+18 vs null	67% computed base rate, K=24	2026-11-18
INDPROWill FRED INDPRO (industrial production) for 2026-10 print above its prior reading of 102.9939 (as of 2026-07-01)? fred · 22 states	57%+7 vs null	50% computed base rate, K=24	2026-11-19
HOUSTWill FRED HOUST (housing starts) for 2026-10 print above its prior reading of 1239.0 (as of 2026-07-01)? fred · 22 states	52%-2 vs null	54% computed base rate, K=24	2026-11-20

Settled

Question	Outcome	Null	Settled
Will FDX report quarterly EPS below the analyst consensus (4.0486) for its report dated 2026-09-16?	UNRESOLVABLE	14%	2026-09-22
Will FRED HOUST (housing starts) for 2026-08 print above its prior reading of 1427.0 (as of 2026-06-01)?	NO	54%	2026-09-20
Will FRED INDPRO (industrial production) for 2026-08 print above its prior reading of 102.6395 (as of 2026-06-01)?	YES	50%	2026-09-19
Will FRED RSAFS (retail sales) for 2026-08 print above its prior reading of 768553.0 (as of 2026-06-01)?	YES	71%	2026-09-18
Will FRED PPIACO (PPI (all commodities)) for 2026-08 print above its prior reading of 286.827 (as of 2026-06-01)?	YES	62%	2026-09-17
Will FRED UNRATE (unemployment rate) for 2026-08 print above its prior reading of 4.2 (as of 2026-06-01)?	NO	38%	2026-09-08
Will FRED PAYEMS (nonfarm payrolls) for 2026-08 print above its prior reading of 158984.0 (as of 2026-06-01)?	YES	75%	2026-09-08

What could not be settled

1 of 7 closed questions could not be settled honestly — 14.3%. Pre-registered tripwire: above 10% flags the era, because a question that can only resolve YES-or-unresolvable is a free option on being right. TRIPPED.

What was not asked

The committed class registers every eligible member as soon as its calendar date is known — so the exam cannot be chosen after the fact. Members that did not become questions are listed here with the reason, because hiding the cull would recreate the selection bias the commitment exists to prevent.

Why it was not asked	Members
no report scheduled in the window	59
already registered (idempotent skip)	55
mechanical base rate 0.96 outside [0.1, 0.9]	2
mechanical base rate 0.00 outside [0.1, 0.9]	1

CPIAUCSL, PAYEMS, PAYEMS, UNRATE, UNRATE, PPIACO, PPIACO, RSAFS, RSAFS, HOUST, HOUST, INDPRO, INDPRO, PCEPI, AAPL, ABBV, ABT, ACN, ADBE, AIG, AMD, AMGN, AMT, AMZN, AVGO, AXP, BA, BAC, BK, BKNG, BLK, BMY, BRK.B, C, CAT, CHTR, CL, CMCSA, COF, COP, COST, CRM, CSCO, CVS, CVX, DE, DHR, DIS, DOW, DUK, EMR, ETN, F, FDX, GD, GE, GILD, GM, GOOGL, GS