Where a dollar moves a West Virginia race.

Where

1. Kanawha County +2,723
   Starting line Capito +11. 2,005 persuadable voters leaning Capito; 27,849 more ballots in 2024 than in 2022.
2. Cabell County +1,272
   Starting line Capito +18. 956 persuadable voters leaning Capito; 13,264 more ballots in 2024 than in 2022.
3. Putnam County +943
   Starting line Capito +40. 883 persuadable voters leaning Capito; 10,846 more ballots in 2024 than in 2022.
4. Raleigh County +839
   Starting line Capito +48. 737 persuadable voters leaning Capito; 11,779 more ballots in 2024 than in 2022.
5. Monongalia County +834
   Starting line Capito +3. 335 persuadable voters leaning Capito; 15,131 more ballots in 2024 than in 2022.

The top 12 counties carry 56% of the plan.

How

• Persuade +15,670
  15,670 voters who split their ticket in recent elections and lean Capito today.
• Mobilize +1,145
  At least 13,008 people who voted in 2024 but not in 2022, in precincts where Fetty Anderson's side wins at least 55 percent.
• Consolidate +1,948
  Voters who went third party in recent races, in the counties that lean Fetty Anderson's way.

Plus a 34.1-point national shift (167,278 votes) that no campaign controls.

What has to be true

1. Fetty Anderson wins 50 percent of the ticket splitters now leaning Capito.
2. In the precincts Fetty Anderson's side carries in Kanawha County and Monongalia County, 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. It would take a national shift of about 34.1 points toward Fetty Anderson's side, larger than any midterm swing in recent memory.
4. No 2026 public poll exists for this race, so the starting line is a weighted blend of official recent results; the first real poll will move it.

The numbers behind it

• Crossover regression
  1,645 precincts, slope 0.923, R² 0.965, 265 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -223,779 to -109,968
• Cash on hand, June 30
  Anderson $22,129 · Capito $4,377,201
• Call program
  34,440 households, 103,319 dials, about 5,414 live conversations and 1,846 voter IDs, at $0.3219 per household (published HyperPhonebank rate)

Where

1. Precinct 14 (Lincoln County) +73
   Starting line Miller +40. 14.5% split their ticket.
2. Precinct 403 (Kanawha County) +66
   Starting line George +37. 4.7% split their ticket. About 397 more ballots in 2024 than in 2022 (estimate).
3. Precinct 22 (Cabell County) +65
   Starting line Miller +2. 13.9% split their ticket.
4. Precinct 406 (Kanawha County) +63
   Starting line George +41. 7.8% split their ticket. About 431 more ballots in 2024 than in 2022 (estimate).
5. Precinct 174 (Kanawha County) +63
   Starting line George +46. 11.8% split their ticket. About 111 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 5% of the plan.

How

• Persuade +10,959
  10,959 voters who split their ticket in recent elections and lean Miller today.
• Mobilize +488
  At least 5,350 people who voted in 2024 but not in 2022, in precincts where George's side wins at least 55 percent.
• Consolidate +1,409
  Voters who went third party in recent races, in the precincts that lean George's way.

Plus a 32.5-point national shift (77,085 votes) that no campaign controls.

What has to be true

1. George wins 50 percent of the ticket splitters now leaning Miller.
2. In Precinct 406 (Kanawha County) and Precinct 403 (Kanawha County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. It would take a national shift of about 32.5 points toward George's side, larger than any midterm swing in recent memory.
4. No 2026 public poll exists for this race, so the starting line is a weighted blend of official recent results; the first real poll will move it.

The numbers behind it

• Crossover regression
  893 precincts, slope 0.822, R² 0.930, 114 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -104,761 to -48,892
• Cash on hand, June 30
  George $258,330 · Miller $604,379
• Call program
  21,592 households, 64,775 dials, about 3,394 live conversations and 1,157 voter IDs, at $0.3468 per household (published HyperPhonebank rate)

Where

1. Precinct 8 (Berkeley County) +115
   Starting line Moore +10. 30.0% split their ticket.
2. Precinct 32 (Monongalia County) +85
   Starting line Parsi +19. 1.1% split their ticket. About 1,185 more ballots in 2024 than in 2022 (estimate).
3. Precinct 2A (Monongalia County) +82
   Starting line Parsi +42. 1.9% split their ticket. About 443 more ballots in 2024 than in 2022 (estimate).
4. Precinct 2B (Monongalia County) +73
   Starting line Parsi +46. 1.8% split their ticket. About 381 more ballots in 2024 than in 2022 (estimate).
5. Precinct 10 (Monongalia County) +61
   Starting line Parsi +37. 3.0% split their ticket. About 395 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 10% of the plan.

How

• Persuade +6,053
  6,053 voters who split their ticket in recent elections and lean Moore today.
• Mobilize +505
  At least 5,638 people who voted in 2024 but not in 2022, in precincts where Parsi's side wins at least 55 percent.
• Consolidate +374
  Voters who went third party in recent races, in the precincts that lean Parsi's way.

Plus a 34.4-point national shift (87,273 votes) that no campaign controls.

What has to be true

1. Parsi wins 50 percent of the ticket splitters now leaning Moore.
2. In Precinct 32 (Monongalia County) and Precinct 2A (Monongalia County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. It would take a national shift of about 34.4 points toward Parsi's side, larger than any midterm swing in recent memory.
4. No 2026 public poll exists for this race, so the starting line is a weighted blend of official recent results; the first real poll will move it.

The numbers behind it

• Crossover regression
  753 precincts, slope 0.844, R² 0.942, 76 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -116,596 to -57,441
• Cash on hand, June 30
  Parsi $43,669 · Moore $751,072
• Call program
  12,981 households, 38,944 dials, about 2,041 live conversations and 696 voter IDs, at $0.4214 per household (published HyperPhonebank rate)