Sonoma County · D2 Supervisor · November Runoff

Vote Universe Model

Paun vs. Lemus — modeling the November electorate from June primary results
Last updated: Tuesday, June 16, 2026, 1:49:03 PM
← Back to precinct dashboard
Two layers below. Banked & contested are real counts from the June primary — solid numbers, reflecting the ~57% who had voted as of the June 16, 2026 count (the primary count is essentially complete). More ballots are still being processed (the county counts through certification), so these figures will rise. The November projection is a scenario model: it applies assumptions you control to the larger general electorate. Projected figures are not forecasts — they show what follows from the assumptions you set. November will be a different, larger electorate than June, and the King/Moreda voters' split is genuinely unknown.
01 — Banked & contested (real primary counts)
Strong Yes
14,106
41.4% of primary vote
Voted Paun in June. A floor among primary voters only — most November voters haven't voted yet.
Hard No
8,563
25.1% of primary vote
Voted Lemus in June. Same caveat: a primary-electorate floor, not her November ceiling.
Consolidation Opportunity
11,442
33.5% of primary vote
Voted King or Moreda — now up for grabs. Larger than Lemus's entire total. Their split is the ballgame.
02 — November projection (interactive — set your assumptions)
45%78%
The June primary finished near ~57%. November generals run higher still — a 62–70% range is a reasonable planning anchor.
20% (Lemus-favorable)65% (Paun-favorable)
The ~6,283 who voted King or Moreda in June. Below 50% reflects the centrist/conservative tilt of that field.
20% (Lemus-favorable)65% (Paun-favorable)
The softest assumption in this model. These are people who didn't vote in June but turn out in November — their partisan makeup isn't anchored to the primary. Note a competitive Democrat-vs-Republican governor's race (Becerra vs. Hilton) at the top of the November ballot may energize right-leaning voters who skipped the Democratic primary, and they could tilt toward the centrist (Lemus). Treat any single value as a guess; slide it to see the range.
Paun (projected)
—
Lemus (projected)
—
Total Nov. ballots (modeled)
—
03 — Scenario table (labeled assumptions)
ScenarioTurnoutK+M→PaunNew→PaunPaunLemusMargin
HOW THE PROJECTION WORKS — Total November ballots = turnout × 67,759 registered voters, split into three tiers: (1) Paun's and Lemus's banked June voters are assumed to turn out and hold; (2) the King + Moreda bloc (~6,283 June voters) turns out and splits by the "K+M→Paun" rate you set; (3) net-new voters — everyone beyond all June voters, implied by the turnout assumption — splits by the "New→Paun" rate. The new-voter split is the least certain input, since those voters aren't anchored to the primary at all. This ignores per-precinct detail and treats drop-off implicitly. It is a planning tool, not a prediction.
Primary source: Sonoma County official results, 6/16/2026 snapshot (D2 Supervisor) · Last updated: Tuesday, June 16, 2026, 1:49:03 PM · Banked/contested figures are real counts; November figures are modeled from user-set assumptions and are not forecasts · This model uses explicit assumptions rather than a historical turnout index because past precinct codes don't map cleanly to the current precincts. November 2026 features a competitive Democrat-vs-Republican governor's race (Becerra vs. Hilton), so gubernatorial-general turnout (Sonoma ~66% in 2022) is a reasonable anchor · For exact voter-level propensity, the county voter file (general-election vote history) is the clean source.