Skip to report

ARCANA PROTOCOL · PREDICTION LEDGER

Election eve: the midterms headline the board.

November 2, 2026· Batch juicy-2026-11-02· 9 forecasts · election check-in Nov 4 · external resolution sources

Reconciled — the Fed-and-Series week

The supplied update reports four settled calls: three hits and one miss. The Fed held as the market said, Big Tech delivered the volatility, and the Brewers closed it out faster than our drama call wanted.

0.88The Fed holds rates steady on Oct 28
✓ YES · hit (10–2 hold) · Brier 0.0144
0.72A Big-5 megacap swings 5%+ after earnings
✓ YES · hit (Alphabet, Amazon) · Brier 0.0784
0.55The Brewers win the World Series
✓ YES · hit (4–1 in 5) · Brier 0.2025
0.56The World Series reaches at least a Game 6
✗ NO · miss (ended in 5) · Brier 0.3136
This pass: 3 / 4 correct · mean Brier 0.1522 Reported running totals — pending reconciliation: ~19 scored · Brier ~0.16

Scores use the outcomes reported in this supplied update and the unchanged October 27 probabilities; this is not a fresh independent adjudication. Brier = (probability − outcome)², with YES = 1 and NO = 0. Lower is better; the mean uses unrounded scores. Running totals remain provisional, including the earlier Dodgers direction-label correction. Four results do not establish forecasting skill.

Apple's print, the bitcoin-after-Fed call, and the Nov 3 geopolitics lines are still being trued up for the running ledger.

The big one — U.S. midterms (Nov 3)

All 435 House seats and 35 Senate seats are on the ballot tomorrow. These are forecasts, not endorsements or picks — each is anchored to where live prediction markets (Kalshi, Polymarket) and final polling sat on Nov 2, so they bet with the crowd where it's confident and flag the genuine coin-flips. A November 4 follow-up can record available AP calls; uncalled races and unsettled chamber control remain pending. Certified results are recorded separately when available.

Democrats win control of the U.S. House
0.92

The most lopsided line on the board: markets price the House at 89–94% Democratic and the generic ballot and forecasters agree. High-conviction, anchored just below the top of the market range.

Expected check-in ~Nov 4, 2026 · AP race calls; official certified results when available

Democrats win control of the U.S. Senate
0.74

The real suspense. Markets moved to ~74–78% Democratic after final NYT/Siena polls showed leads in six of seven battlegrounds, but a four-seat net gain across high-variance states keeps this well short of a lock.

Expected check-in ~Nov 4, 2026 · AP race calls; official certified results when available

Democrats sweep both chambers
0.72

Essentially the Senate call conditioned on the near-certain House — markets put the double at ~74%. The main alternative the board respects is a split: Democratic House, Republican Senate (~22%).

Expected check-in ~Nov 4, 2026 · AP race calls; official certified results when available

Republicans hold the Florida Senate seat
0.85

A deliberate GOP-favored line for balance: Florida sits ~88% Republican in the markets. If the night is a Democratic wave, this is one of the seats that would fall and signal it.

Expected check-in ~Nov 4, 2026 · AP race calls; official certified results when available

The Democrat wins the Ohio Senate seat (Brown vs. Husted)
0.70

A marquee comeback story priced ~71% for the Democrat. Ohio has drifted red in recent cycles, so a win here is both a strong signal for Senate control and a real test of the polling.

Expected check-in ~Nov 4, 2026 · AP race calls; official certified results when available

At least one Senate race is still uncalled by midday Nov 4
0.72

The election-night meta-call. With Iowa a coin-flip and several states inside a point, at least one Senate race dragging past Wednesday morning is the likely outcome — and the reason your follow-up may not score everything at once.

Resolves Nov 4, 2026 (midday ET) · major-outlet calls

Markets

A post-Fed, post-Big-Tech tape heads into the election with records in reach.

The S&P 500 sets a new all-time closing high at some point this week
0.45

The index closed near 7,744 and has been grinding up for days, but an election week with a contested Senate is exactly the kind of event risk that stalls a melt-up. A slightly-under coin-flip.

Resolves Nov 6, 2026 · market close

World & power

The slow crises hold while Washington votes.

The Strait of Hormuz stays effectively closed through Nov 9
0.85

Months of suppressed transit with no diplomatic break in sight; another week of blockade is the base case and keeps oil and inflation pressure live.

Resolves Nov 9, 2026 · UKMTO / IEA transit data

No internationally recognized government is seated in Venezuela by Nov 9
0.55

A recognized successor to Maduro remains more likely weeks out than days; the contested-vacuum state persists for now.

Resolves Nov 9, 2026 · official recognition / consensus reporting

A note on the election calls. Arcana forecasts outcomes; it does not take sides. The six midterm lines are probability estimates anchored to live betting markets and final public polling as of Nov 2 — they describe what the odds implied, not who should win, and carry no view on any candidate, party, or the value of any vote. The next follow-up records available AP race calls, hits and misses alike. AP reports and calls races; election officials certify official results. No verdict is forced by a calendar deadline. The separate “uncalled by midday” forecast retains its November 4, noon ET deadline. How AP calls races · How results become official.
How to read this. Every call is listed with a probability, a resolution date or expected check-in, and a named external source — AP race calls, official certified results, major-outlet race calls, market closes and UKMTO. Probabilities span 0.45 to 0.92 and track live market odds where those exist. Political lines are neutral forecasts, not endorsements. Hits and misses will both remain visible when graded. The House, Senate and both-chambers forecasts are correlated, not three independent tests. Any unspecified benchmark or resolution criterion must be disclosed when adjudicating and must not be selected to fit the observed outcome.