Where a dollar moves a Georgia race.

Where

1. Fulton County +28,914
   Starting line Ossoff +54. 28,130 persuadable voters leaning Ossoff; 120,121 more ballots in 2024 than in 2022.
2. Cobb County +22,907
   Starting line Ossoff +23. 21,690 persuadable voters leaning Ossoff; 92,928 more ballots in 2024 than in 2022.
3. Gwinnett County +20,392
   Starting line Ossoff +27. 19,513 persuadable voters leaning Ossoff; 124,927 more ballots in 2024 than in 2022.
4. DeKalb County +17,677
   Starting line Ossoff +76. 17,492 persuadable voters leaning Ossoff; 71,028 more ballots in 2024 than in 2022.
5. Cherokee County +12,726
   Starting line Collins +31. 8,161 persuadable voters leaning Ossoff; 42,821 more ballots in 2024 than in 2022.

The top 12 counties carry 46% of the plan.

How

• Persuade +242,998
  Ticket splitters and the 8 percent still undecided in the latest poll: 242,998 leaning Ossoff today.
• Mobilize +78,382
  At least 588,952 people who voted in 2024 but not in 2022, in counties where Collins's side wins at least 55 percent.
• Consolidate +7,976
  Voters who went third party in recent races, in the counties that lean Collins's way.

Plus a 0.1-point national shift (2,157 votes) that no campaign controls.

What has to be true

1. Collins wins 50 percent of the ticket splitters and undecided voters now leaning Ossoff.
2. In the precincts Collins's side carries in Cherokee County and Hall County, 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 0.1 points toward Collins's side between now and November 3.
4. The starting line comes from published 2026 polls, including campaign or party sponsored polls where those are the only ones; if the true gap is different, so is the plan.

The numbers behind it

• Crossover regression
  2,718 precincts, slope 0.983, R² 0.994, 366 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 5-point baseline error. 90% range of the final margin with the plan: -361,989 to 364,220
• Cash on hand, June 30
  Collins $2,079,981 · Ossoff $42,587,451
• Call program
  757,438 households, 2,272,313 dials, about 119,069 live conversations and 40,603 voter IDs, at $0.2749 per household (published HyperPhonebank rate)

Where

1. Windermere (Forsyth County) +447
   Starting line Kozycki +2. 8.0% split their ticket.
2. Nichols (Forsyth County) +421
   Starting line Kozycki +2. 8.3% split their ticket.
3. Johns Creek (Forsyth County) +389
   Starting line Kozycki +8. 8.1% split their ticket.
4. Rw01 (Fulton County) +341
   Starting line Kozycki +10. 8.5% split their ticket.
5. Ap01a (Fulton County) +315
   Starting line Kozycki +24. 9.0% split their ticket.

The top 12 precincts carry 33% of the plan.

How

• Persuade +10,307
  Ticket splitters and the 11 percent still undecided in the latest poll: 29,367 leaning McCormick today.
• Mobilize +1,344
  At least 33,471 people who voted in 2024 but not in 2022, in precincts where Kozycki's side wins at least 55 percent.
• Consolidate +168
  Voters who went third party in recent races, in the precincts that lean Kozycki's way.

What has to be true

1. Kozycki wins 18 percent of the ticket splitters and undecided voters now leaning McCormick.
2. No national wave is required. The plan closes at 35 percent of what recent Georgia results show is possible.
3. The starting line comes from published 2026 polls, including campaign or party sponsored polls where those are the only ones; if the true gap is different, so is the plan.

The numbers behind it

• Crossover regression
  174 precincts, slope 0.944, R² 0.991, 39 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 5-point baseline error. 90% range of the final margin with the plan: -27,178 to 28,842
• Cash on hand, June 30
  Kozycki $277,427 · McCormick $1,007,984
• Call program
  19,111 households, 57,334 dials, about 3,004 live conversations and 1,024 voter IDs, at $0.3717 per household (published HyperPhonebank rate)

Where

1. Redbone (Lee County) +233
   Starting line Day +52. 5.2% split their ticket. About 860 more ballots in 2024 than in 2022 (estimate).
2. Colquitt-Miller (Miller County) +167
   Starting line Day +46. 6.8% split their ticket. About 578 more ballots in 2024 than in 2022 (estimate).
3. Precinct 1 (Schley County) +153
   Starting line Day +58. 4.3% split their ticket. About 550 more ballots in 2024 than in 2022 (estimate).
4. Chattahoochee (Muscogee County) +149
   Starting line Day +12. 4.6% split their ticket. About 1,644 more ballots in 2024 than in 2022 (estimate).
5. Byron 2 (Peach County) +143
   Starting line Day +20. 4.3% split their ticket. About 1,107 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 20% of the plan.

How

• Persuade +4,905
  4,905 voters who split their ticket in recent elections and lean Bishop today.
• Mobilize +3,407
  At least 26,413 people who voted in 2024 but not in 2022, in precincts where Day's side wins at least 55 percent.
• Consolidate +151
  Voters who went third party in recent races, in the precincts that lean Day's way.

Plus a 6-point national shift (14,593 votes) that no campaign controls.

What has to be true

1. Day wins 50 percent of the ticket splitters now leaning Bishop.
2. In Redbone (Lee County) and Precinct 1 (Schley County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 6.0 points toward Day's side between now and November 3.
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
  229 precincts, slope 0.935, R² 0.994, 53 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -42,898 to 14,252
• Cash on hand, June 30
  Day $1,768 · Bishop $447,808
• Call program
  25,291 households, 75,873 dials, about 3,976 live conversations and 1,356 voter IDs, at $0.3344 per household (published HyperPhonebank rate)

Where

1. Precinct 5-10C (Chatham County) +199
   Starting line Hollowell +80. 8.2% split their ticket. About 495 more ballots in 2024 than in 2022 (estimate).
2. Lyman Hall School (Liberty County) +190
   Starting line Hollowell +39. 6.8% split their ticket. About 955 more ballots in 2024 than in 2022 (estimate).
3. Victory Baptist (Liberty County) +176
   Starting line Hollowell +40. 7.1% split their ticket. About 847 more ballots in 2024 than in 2022 (estimate).
4. Upp (Glynn County) +170
   Starting line Hollowell +71. 6.4% split their ticket. About 540 more ballots in 2024 than in 2022 (estimate).
5. Precinct 8-08C (Chatham County) +170
   Starting line Hollowell +84. 6.0% split their ticket. About 508 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 19% of the plan.

How

• Persuade +6,354
  6,354 voters who split their ticket in recent elections and lean Kingston today.
• Mobilize +4,081
  At least 29,513 people who voted in 2024 but not in 2022, in precincts where Hollowell's side wins at least 55 percent.
• Consolidate +195
  Voters who went third party in recent races, in the precincts that lean Hollowell's way.

Plus a 16.2-point national shift (43,077 votes) that no campaign controls.

What has to be true

1. Hollowell wins 50 percent of the ticket splitters now leaning Kingston.
2. In Precinct 8-08C (Chatham County) and Precinct 5-03C (Chatham County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 16.2 points toward Hollowell's side between now and November 3.
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
  216 precincts, slope 0.939, R² 0.996, 32 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -73,917 to -11,929
• Cash on hand, June 30
  Hollowell $7,389 · Kingston $294,093
• Call program
  29,518 households, 88,553 dials, about 4,640 live conversations and 1,582 voter IDs, at $0.3344 per household (published HyperPhonebank rate)

Where

1. Precinct 310 (Richmond County) +302
   Starting line Smith +53. 2.3% split their ticket. About 1,429 more ballots in 2024 than in 2022 (estimate).
2. Precinct 405 (Richmond County) +299
   Starting line Smith +80. 3.2% split their ticket. About 886 more ballots in 2024 than in 2022 (estimate).
3. Precinct 607 (Richmond County) +197
   Starting line Smith +82. 3.9% split their ticket. About 540 more ballots in 2024 than in 2022 (estimate).
4. Precinct 506 (Richmond County) +179
   Starting line Smith +69. 4.0% split their ticket. About 578 more ballots in 2024 than in 2022 (estimate).
5. Calhoun Park (Laurens County) +166
   Starting line Smith +42. 5.4% split their ticket. About 683 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 25% of the plan.

How

• Persuade +5,233
  5,233 voters who split their ticket in recent elections and lean Allen today.
• Mobilize +3,185
  At least 19,371 people who voted in 2024 but not in 2022, in precincts where Smith's side wins at least 55 percent.
• Consolidate +167
  Voters who went third party in recent races, in the precincts that lean Smith's way.

Plus a 14.5-point national shift (38,942 votes) that no campaign controls.

What has to be true

1. Smith wins 50 percent of the ticket splitters now leaning Allen.
2. In Precinct 310 (Richmond County) and Precinct 405 (Richmond County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 14.5 points toward Smith's side between now and November 3.
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
  257 precincts, slope 0.964, R² 0.996, 39 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -70,707 to -7,502
• Cash on hand, June 30
  Smith $3,502 · Allen $1,320,958
• Call program
  20,917 households, 62,752 dials, about 3,288 live conversations and 1,121 voter IDs, at $0.3468 per household (published HyperPhonebank rate)

Where

1. Newnan Centre (Coweta County) +322
   Starting line Jack +14. 5.3% split their ticket.
2. North Hampton (Henry County) +194
   Starting line Keller +29. 2.7% split their ticket. About 1,399 more ballots in 2024 than in 2022 (estimate).
3. Madras (Coweta County) +171
   Starting line Jack +39. 4.8% split their ticket.
4. Canongate (Coweta County) +155
   Starting line Jack +48. 5.7% split their ticket.
5. Mirror Lake Elementary (Douglas County) +150
   Starting line Keller +17. 3.6% split their ticket. About 1,415 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 22% of the plan.

How

• Persuade +6,701
  6,701 voters who split their ticket in recent elections and lean Jack today.
• Mobilize +1,507
  At least 16,892 people who voted in 2024 but not in 2022, in precincts where Keller's side wins at least 55 percent.
• Consolidate +229
  Voters who went third party in recent races, in the precincts that lean Keller's way.

Plus a 29.3-point national shift (92,279 votes) that no campaign controls.

What has to be true

1. Keller wins 50 percent of the ticket splitters now leaning Jack.
2. In North Hampton (Henry County) and Central (Coweta County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 29.3 points toward Keller's side between now and November 3.
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
  191 precincts, slope 0.989, R² 0.996, 22 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -128,985 to -55,845
• Cash on hand, June 30
  Keller $87,576 · Jack $1,142,625
• Call program
  21,355 households, 64,066 dials, about 3,357 live conversations and 1,145 voter IDs, at $0.3468 per household (published HyperPhonebank rate)

Where

1. Duluth I (Gwinnett County) +123
   Starting line Duffie +7. 9.0% split their ticket.
2. Pinckneyville M (Gwinnett County) +118
   Starting line Duffie +23. 8.0% split their ticket. About 456 more ballots in 2024 than in 2022 (estimate).
3. Ashford Dunwoody Rd (DeKalb County) +107
   Starting line Johnson +12. 9.8% split their ticket.
4. Duluth A (Gwinnett County) +102
   Starting line Johnson +16. 6.6% split their ticket.
5. Ashford Parkside (DeKalb County) +101
   Starting line Johnson +19. 8.8% split their ticket.

The top 12 precincts carry 22% of the plan.

How

• Persuade +5,086
  5,086 voters who split their ticket in recent elections and lean Johnson today.
• Mobilize +76
  At least 1,277 people who voted in 2024 but not in 2022, in precincts where Duffie's side wins at least 55 percent.
• Consolidate +154
  Voters who went third party in recent races, in the precincts that lean Duffie's way.

Plus a 51-point national shift (122,668 votes) that no campaign controls.

What has to be true

1. Duffie wins 50 percent of the ticket splitters now leaning Johnson.
2. In Pinckneyville M (Gwinnett County) and Pinckneyville A1 (Gwinnett County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 51.0 points toward Duffie's side between now and November 3.
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
  168 precincts, slope 1.082, R² 0.992, 29 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -150,811 to -94,761
• Cash on hand, June 30
  Duffie n/a · Johnson $117,906
• Call program
  8,205 households, 24,615 dials, about 1,290 live conversations and 440 voter IDs, at $0.4711 per household (published HyperPhonebank rate)

Where

1. Silver Lake (DeKalb County) +113
   Starting line Williams +16. 7.6% split their ticket.
2. Precinct 08L (Fulton County) +109
   Starting line Williams +43. 8.3% split their ticket.
3. 06d1 (Fulton County) +106
   Starting line Williams +59. 8.1% split their ticket.
4. Ashford Park Elem (DeKalb County) +102
   Starting line Williams +5. 11.3% split their ticket.
5. Brookhaven (DeKalb County) +92
   Starting line Williams +18. 8.2% split their ticket.

The top 12 precincts carry 18% of the plan.

How

• Persuade +5,800
  5,800 voters who split their ticket in recent elections and lean Williams today.
• Mobilize +0
  There are almost no precincts Salvesen's side carries, so turnout alone cannot help here.
• Consolidate +88
  Voters who went third party in recent races, in the precincts that lean Salvesen's way.

Plus a 69.9-point national shift (196,851 votes) that no campaign controls.

What has to be true

1. Salvesen wins 50 percent of the ticket splitters now leaning Williams.
2. The national environment moves about 69.9 points toward Salvesen's side between now and November 3.
3. 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
  262 precincts, slope 1.062, R² 0.980, 46 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -229,892 to -163,955
• Cash on hand, June 30
  Salvesen n/a · Williams $76,092
• Call program
  8,178 households, 24,535 dials, about 1,286 live conversations and 439 voter IDs, at $0.4711 per household (published HyperPhonebank rate)

Where

1. Smyrna 6A (Cobb County) +152
   Starting line McBath +20. 7.8% split their ticket.
2. Precinct 08J (Fulton County) +145
   Starting line McBath +15. 13.6% split their ticket.
3. Smyrna 4A (Cobb County) +143
   Starting line McBath +22. 7.0% split their ticket.
4. Precinct 07B (Fulton County) +138
   Starting line Martin +12. 11.2% split their ticket. About 344 more ballots in 2024 than in 2022 (estimate).
5. Smyrna 7A (Cobb County) +138
   Starting line McBath +49. 6.0% split their ticket.

The top 12 precincts carry 21% of the plan.

How

• Persuade +7,125
  7,125 voters who split their ticket in recent elections and lean McBath today.
• Mobilize +83
  At least 1,280 people who voted in 2024 but not in 2022, in precincts where Martin's side wins at least 55 percent.
• Consolidate +165
  Voters who went third party in recent races, in the precincts that lean Martin's way.

Plus a 46.5-point national shift (135,594 votes) that no campaign controls.

What has to be true

1. Martin wins 50 percent of the ticket splitters now leaning McBath.
2. In Precinct 07B (Fulton County) and 08f1 (Fulton County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 46.5 points toward Martin's side between now and November 3.
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
  240 precincts, slope 1.062, R² 0.993, 31 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -169,091 to -102,205
• Cash on hand, March 31
  Martin $100 · McBath $676,349
• Call program
  11,042 households, 33,125 dials, about 1,736 live conversations and 592 voter IDs, at $0.4462 per household (published HyperPhonebank rate)

Where

1. Lowndes Civic (Lowndes County) +206
   Starting line Esti +31. 2.8% split their ticket. About 1,427 more ballots in 2024 than in 2022 (estimate).
2. Rozr (Houston County) +171
   Starting line Scott +39. 4.0% split their ticket.
3. Rainwater (Lowndes County) +154
   Starting line Scott +28. 3.6% split their ticket.
4. Bms (Houston County) +147
   Starting line Scott +35. 4.5% split their ticket.
5. Mott-Litman Gym (Tift County) +141
   Starting line Esti +46. 15.1% split their ticket. About 366 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 25% of the plan.

How

• Persuade +5,350
  5,350 voters who split their ticket in recent elections and lean Scott today.
• Mobilize +893
  At least 7,694 people who voted in 2024 but not in 2022, in precincts where Esti's side wins at least 55 percent.
• Consolidate +135
  Voters who went third party in recent races, in the precincts that lean Esti's way.

Plus a 32.5-point national shift (87,258 votes) that no campaign controls.

What has to be true

1. Esti wins 50 percent of the ticket splitters now leaning Scott.
2. In Lowndes Civic (Lowndes County) and Mt. Calvary Church (Lowndes County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 32.5 points toward Esti's side between now and November 3.
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
  225 precincts, slope 0.945, R² 0.988, 27 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -118,735 to -55,937
• Cash on hand, June 30
  Esti $1,688 · Scott $1,349,452
• Call program
  12,957 households, 38,872 dials, about 2,037 live conversations and 695 voter IDs, at $0.4214 per household (published HyperPhonebank rate)

Where

1. Lawrenceville M (Gwinnett County) +341
   Starting line Gegen +47. 3.6% split their ticket. About 1,732 more ballots in 2024 than in 2022 (estimate).
2. West Jackson (Jackson County) +252
   Starting line Clyde +57. 4.7% split their ticket.
3. Rabun County (Rabun County) +241
   Starting line Clyde +59. 5.7% split their ticket.
4. Friendship I (Hall County) +223
   Starting line Clyde +50. 7.1% split their ticket.
5. Central Jackson (Jackson County) +219
   Starting line Clyde +59. 4.4% split their ticket.

The top 12 precincts carry 28% of the plan.

How

• Persuade +7,197
  7,197 voters who split their ticket in recent elections and lean Clyde today.
• Mobilize +1,140
  At least 10,268 people who voted in 2024 but not in 2022, in precincts where Gegen's side wins at least 55 percent.
• Consolidate +235
  Voters who went third party in recent races, in the precincts that lean Gegen's way.

Plus a 34.6-point national shift (99,718 votes) that no campaign controls.

What has to be true

1. Gegen wins 50 percent of the ticket splitters now leaning Clyde.
2. In Lawrenceville M (Gwinnett County) and Lawrenceville F (Gwinnett County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 34.6 points toward Gegen's side between now and November 3.
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
  145 precincts, slope 0.995, R² 0.994, 9 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -132,858 to -66,198
• Cash on hand, June 30
  Gegen $4,670 · Clyde $176,253
• Call program
  17,479 households, 52,438 dials, about 2,748 live conversations and 937 voter IDs, at $0.3965 per household (published HyperPhonebank rate)

Where

1. Lowes (Henry County) +470
   Starting line DeLancy +45. 2.8% split their ticket. About 2,500 more ballots in 2024 than in 2022 (estimate).
2. Wesley Lakes (Henry County) +320
   Starting line DeLancy +58. 2.5% split their ticket. About 1,355 more ballots in 2024 than in 2022 (estimate).
3. Westside (Henry County) +313
   Starting line DeLancy +64. 2.5% split their ticket. About 1,242 more ballots in 2024 than in 2022 (estimate).
4. Locust Grove (Henry County) +281
   Starting line DeLancy +43. 1.8% split their ticket. About 1,624 more ballots in 2024 than in 2022 (estimate).
5. Precinct 5D (Clarke County) +241
   Starting line DeLancy +45. 3.6% split their ticket. About 1,364 more ballots in 2024 than in 2022 (estimate).

The top 12 precincts carry 29% of the plan.

How

• Persuade +6,496
  6,496 voters who split their ticket in recent elections and lean Gaines today.
• Mobilize +3,466
  At least 26,472 people who voted in 2024 but not in 2022, in precincts where DeLancy's side wins at least 55 percent.
• Consolidate +237
  Voters who went third party in recent races, in the precincts that lean DeLancy's way.

Plus a 20.8-point national shift (64,142 votes) that no campaign controls.

What has to be true

1. DeLancy wins 50 percent of the ticket splitters now leaning Gaines.
2. In Lowes (Henry County) and Westside (Henry County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 20.8 points toward DeLancy's side between now and November 3.
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
  165 precincts, slope 0.985, R² 0.993, 21 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -99,736 to -28,780
• Cash on hand, June 30
  DeLancy $8,705 · Gaines $703,420
• Call program
  27,698 households, 83,093 dials, about 4,354 live conversations and 1,485 voter IDs, at $0.3344 per household (published HyperPhonebank rate)

Where

1. Kennesaw 3A (Cobb County) +189
   Starting line Harden +21. 4.6% split their ticket. About 1,619 more ballots in 2024 than in 2022 (estimate).
2. Oregon 03 (Cobb County) +180
   Starting line Harden +39. 3.6% split their ticket. About 1,004 more ballots in 2024 than in 2022 (estimate).
3. Little River (Cherokee County) +179
   Starting line Cowan +30. 6.9% split their ticket.
4. Terrell Mill 01 (Cobb County) +176
   Starting line Harden +33. 4.9% split their ticket. About 971 more ballots in 2024 than in 2022 (estimate).
5. Roswell 01 (Cobb County) +175
   Starting line Cowan +13. 8.9% split their ticket.

The top 12 precincts carry 18% of the plan.

How

• Persuade +9,880
  9,880 voters who split their ticket in recent elections and lean Cowan today.
• Mobilize +1,184
  At least 12,675 people who voted in 2024 but not in 2022, in precincts where Harden's side wins at least 55 percent.
• Consolidate +337
  Voters who went third party in recent races, in the precincts that lean Harden's way.

Plus a 25.4-point national shift (80,205 votes) that no campaign controls.

What has to be true

1. Harden wins 50 percent of the ticket splitters now leaning Cowan.
2. In Oregon 03 (Cobb County) and Chalker 01 (Cobb County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 25.4 points toward Harden's side between now and November 3.
4. Published 2026 polls do not set the starting line here (one has a documented methodological flaw), so it is a weighted blend of official recent results; an independent head-to-head poll would move it.

The numbers behind it

• Crossover regression
  152 precincts, slope 0.929, R² 0.978, 26 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -116,764 to -43,675
• Cash on hand, June 30
  Harden $2,237 · Cowan $17,630
• Call program
  23,043 households, 69,130 dials, about 3,622 live conversations and 1,235 voter IDs, at $0.3468 per household (published HyperPhonebank rate)

Where

1. Garners A (Gwinnett County) +90
   Starting line Clark +4. 7.9% split their ticket.
2. Cates M (Gwinnett County) +86
   Starting line Clark +5. 6.8% split their ticket.
3. Baycreek H (Gwinnett County) +82
   Starting line Clark +26. 3.7% split their ticket.
4. Baycreek C (Gwinnett County) +80
   Starting line Clark +31. 4.2% split their ticket.
5. 17-Oxford (Newton County) +74
   Starting line Clark +3. 23.3% split their ticket.

The top 12 precincts carry 18% of the plan.

How

• Persuade +4,764
  4,764 voters who split their ticket in recent elections and lean Clark today.
• Mobilize +0
  There are almost no precincts Chavez's side carries, so turnout alone cannot help here.
• Consolidate +178
  Voters who went third party in recent races, in the precincts that lean Chavez's way.

Plus a 41.6-point national shift (112,891 votes) that no campaign controls.

What has to be true

1. Chavez wins 50 percent of the ticket splitters now leaning Clark.
2. The national environment moves about 41.6 points toward Chavez's side between now and November 3.
3. 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
  149 precincts, slope 0.981, R² 0.970, 8 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -144,425 to -81,513
• Cash on hand, June 30
  Chavez n/a · Clark $229,145
• Call program
  6,926 households, 20,777 dials, about 1,089 live conversations and 371 voter IDs, at $0.5208 per household (published HyperPhonebank rate)

Where

1. Russom Elementary (Paulding County) +184
   Starting line Fuller +25. 5.1% split their ticket.
2. Acworth 1C (Cobb County) +168
   Starting line Harris +30. 4.1% split their ticket. About 1,147 more ballots in 2024 than in 2022 (estimate).
3. Kennesaw 5A (Cobb County) +156
   Starting line Harris +14. 5.2% split their ticket. About 1,260 more ballots in 2024 than in 2022 (estimate).
4. Lindale (Floyd County) +132
   Starting line Fuller +49. 5.1% split their ticket.
5. Ringgold (Catoosa County) +123
   Starting line Fuller +46. 5.8% split their ticket.

The top 12 precincts carry 20% of the plan.

How

• Persuade +7,037
  7,037 voters who split their ticket in recent elections and lean Fuller today.
• Mobilize +472
  At least 5,868 people who voted in 2024 but not in 2022, in precincts where Harris's side wins at least 55 percent.
• Consolidate +218
  Voters who went third party in recent races, in the precincts that lean Harris's way.

Plus a 31.8-point national shift (88,504 votes) that no campaign controls.

What has to be true

1. Harris wins 50 percent of the ticket splitters now leaning Fuller.
2. In Acworth 1C (Cobb County) and Kennesaw 5A (Cobb County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
3. The national environment moves about 31.8 points toward Harris's side between now and November 3.
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
  143 precincts, slope 0.954, R² 0.988, 21 crossover precincts
• Monte Carlo
  20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -120,417 to -56,098
• Cash on hand, June 30
  Harris $18,230 · Fuller $70,518
• Call program
  14,189 households, 42,568 dials, about 2,231 live conversations and 761 voter IDs, at $0.4214 per household (published HyperPhonebank rate)