The Political Group · Michigan 2026 · as of September 27, 2026

Where a dollar moves a Michigan race.

14 federal races, seen from the underdog's side. Ranked by how far a disciplined voter contact plan moves the underdog's odds, from 20,000 seeded simulations per race built on official precinct returns.

RaceOdds: today → with the planLiftNet votesCall program
Michigan U.S. Senate
Mike Rogers vs. Abdul El-Sayed
30% → 51%+21 pts116,706$59,812
217,578 households
Brief ›
MI-04
Sean McCann vs. Bill Huizenga
50% → 66%+16 pts6,921$5,185
11,620 households
Brief ›
MI-08
Thomas J. Smith vs. Kristen McDonald Rivet
19% → 33%+14 pts21,130$9,463
28,297 households
Brief ›
MI-07
William Lawrence vs. Tom Barrett
50% → 62%+12 pts7,577$7,150
20,616 households
Brief ›
MI-10
Christina Hines vs. Michael Bouchard
43% → 50%+7 pts4,254$3,850
7,392 households
Brief ›
MI-01
Callie Barr vs. Jack Bergman
<1% → 2%+2 pts75,130$8,594
25,699 households
Brief ›
MI-03
Terri DeBoer vs. Hillary Scholten
7% → 9%+2 pts36,411$6,032
15,213 households
Brief ›
MI-02
Ben Ambrose vs. John Moolenaar
0% → <1%+0 pts100,270$6,870
17,326 households
Brief ›
MI-05
Christian Vukasovich vs. Tim Walberg
<1% → <1%+0 pts87,818$7,002
18,837 households
Brief ›
MI-06
Heather Smiley vs. Debbie Dingell
0% → <1%+0 pts108,687$4,671
10,469 households
Brief ›
MI-09
Ray Pooley vs. Lisa McClain
0% → <1%+0 pts118,333$7,305
21,065 households
Brief ›
MI-11
Ethan Baker vs. Jeremy Moss
<1% → <1%+0 pts79,890$5,575
13,229 households
Brief ›
MI-12
James D. Hooper vs. Rashida Tlaib
0% → 0%+0 pts125,325$6,155
14,605 households
Brief ›
MI-13
Taras "T.P." Nykoriak vs. Donavan McKinney
0% → 0%+0 pts111,252$4,209
8,935 households
Brief ›

Lift is the gain in win probability when the plan is executed. Call program: every household on the plan's target list, three attempts each, at published HyperPhonebank rates. Calls are the voter ID and ballot chase engine of each plan: they find the persuadable voters and the supporters so every other dollar lands on the right door.

Upset Brief · Michigan U.S. Senate · Mike Rogers (R) vs. Abdul El-Sayed (D)

Rogers can win. The plan uses 35 percent of what history allows.

Odds today30%
With the plan51%
Net votes to find116,706
Call program$59,812

Where

  1. Macomb County +21,816
    Starting line Rogers +6. 27,421 persuadable voters leaning El-Sayed; 124,516 more ballots in 2024 than in 2022.
  2. Wayne County +20,107
    Starting line El-Sayed +33. 42,716 persuadable voters leaning El-Sayed; 213,253 more ballots in 2024 than in 2022.
  3. Oakland County +10,438
    Starting line El-Sayed +15. 41,634 persuadable voters leaning El-Sayed; 142,610 more ballots in 2024 than in 2022.
  4. Livingston County +6,751
    Starting line Rogers +20. 7,723 persuadable voters leaning El-Sayed; 23,616 more ballots in 2024 than in 2022.
  5. St. Clair County +5,050
    Starting line Rogers +27. 5,371 persuadable voters leaning El-Sayed; 21,891 more ballots in 2024 than in 2022.

The top 12 counties carry 70% of the plan.

How

  • Persuade +102,720
    297,441 voters in play: ticket splitters now leaning El-Sayed, plus the 11 percent still undecided in the latest poll.
  • Mobilize +14,284
    At least 439,770 people who voted in 2024 but not in 2022, in cities and townships where Rogers's side wins at least 55 percent.
  • Consolidate +5,537
    Voters who went third party in recent races, in the counties that lean Rogers's way.

What has to be true

  1. Rogers wins 17 percent of the ticket splitters and undecided voters now leaning El-Sayed.
  2. In the cities and townships Rogers's side carries in St. Clair County and Allegan County, 12 percent of the people who voted in 2024 but skipped 2022 come back and vote.
  3. No national wave is required. The plan uses 35 percent of what recent Michigan results show is possible.
  4. The starting line comes from the published 2026 head-to-head polls between these two finalists, sponsored ones included; if the true gap is different, so is the plan.

The numbers behind it

  • Crossover regression
    4,417 precincts, slope 0.984, R² 0.994, 612 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 5-point baseline error. 90% range of the final margin with the plan: -363,932 to 380,911
  • Cash on hand, July 15
    Rogers $4,473,237 · El-Sayed $2,552,763
  • Call program
    217,578 households, 652,733 dials, about 34,203 live conversations and 11,663 voter IDs, at $0.2749 per household (published HyperPhonebank rate)
Upset Brief · MI-04 · Sean McCann (D) vs. Bill Huizenga (R) · incumbent

McCann can win. The plan uses 26 percent of what history allows.

Odds today50%
With the plan66%
Net votes to find6,921
Call program$5,185

Where

  1. Portage +1,822
    Starting line McCann +23. 7.7% split their ticket.
  2. Kalamazoo +723
    Starting line McCann +64. 4.8% split their ticket. About 7,843 more ballots in 2024 than in 2022 (estimate).
  3. Texas Township (Kalamazoo County) +719
    Starting line McCann +10. 7.6% split their ticket.
  4. Comstock Township (Kalamazoo County) +570
    Starting line McCann +9. 7.5% split their ticket.
  5. Benton Harbor +494
    Starting line McCann +93. 2.1% split their ticket. About 1,507 more ballots in 2024 than in 2022 (estimate).

The top 12 cities and townships carry 80% of the plan.

How

  • Persuade +5,899
    23,084 voters in play: ticket splitters now leaning Huizenga, plus the 7 percent still undecided in the latest poll.
  • Mobilize +1,054
    At least 33,797 people who voted in 2024 but not in 2022, in cities and townships where McCann's side wins at least 55 percent.
  • Consolidate +313
    Voters who went third party in recent races, in the cities and townships that lean McCann's way.

What has to be true

  1. McCann wins 13 percent of the ticket splitters and undecided voters now leaning Huizenga.
  2. In Kalamazoo and Benton Harbor, 9 percent of the people who voted in 2024 but skipped 2022 come back and vote.
  3. No national wave is required. The plan uses 26 percent of what recent Michigan results show is possible.
  4. The starting line comes from the published 2026 head-to-head polls between these two finalists, sponsored ones included; if the true gap is different, so is the plan.

The numbers behind it

  • Crossover regression
    299 precincts, slope 1.026, R² 0.990, 44 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 5-point baseline error. 90% range of the final margin with the plan: -21,840 to 36,026
  • Cash on hand, July 15
    McCann $1,686,240 · Huizenga $2,268,747
  • Call program
    11,620 households, 34,859 dials, about 1,827 live conversations and 623 voter IDs, at $0.4462 per household (published HyperPhonebank rate)
Upset Brief · MI-08 · Thomas J. Smith (R) vs. Kristen McDonald Rivet (D) · incumbent

Smith gets within 10,623 votes on campaign work alone. A 3.1-point national shift finishes it.

Odds today19%
With the plan33%
Net votes to find21,130
Call program$9,463

Where

  1. Saginaw Township (Saginaw County) +499
    Starting line McDonald Rivet +11. 4.1% split their ticket.
  2. Grand Blanc Township (Genesee County) +421
    Starting line McDonald Rivet +13. 3.6% split their ticket.
  3. Bay City +392
    Starting line McDonald Rivet +19. 5.2% split their ticket.
  4. Burton +368
    Starting line McDonald Rivet +10. 5.1% split their ticket.
  5. Midland (Midland County) +337
    Starting line McDonald Rivet +8. 2.4% split their ticket.

The top 12 cities and townships carry 37% of the plan.

How

  • Persuade +7,357
    7,357 voters who split their ticket in recent elections and lean McDonald Rivet today.
  • Mobilize +1,414
    At least 21,336 people who voted in 2024 but not in 2022, in cities and townships where Smith's side wins at least 55 percent.
  • Consolidate +1,736
    Voters who went third party in recent races, in the cities and townships that lean Smith's way.

Plus a 3.1-point national shift (10,623 votes) that no campaign controls.

What has to be true

  1. Smith wins 50 percent of the ticket splitters now leaning McDonald Rivet.
  2. In Fenton Township (Genesee County) and Thomas Township (Saginaw County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
  3. The national environment moves about 3.1 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
    302 precincts, slope 0.944, R² 0.991, 57 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -50,954 to 29,369
  • Cash on hand, July 15
    Smith $219,799 · Rivet $4,254,099
  • Call program
    28,297 households, 84,890 dials, about 4,448 live conversations and 1,517 voter IDs, at $0.3344 per household (published HyperPhonebank rate)
Upset Brief · MI-07 · William Lawrence (D) vs. Tom Barrett (R) · incumbent

Lawrence can win. The plan uses 80 percent of what history allows.

Odds today50%
With the plan62%
Net votes to find7,577
Call program$7,150

Where

  1. Lansing (Ingham County) +2,307
    Starting line Lawrence +47. 1.7% split their ticket. About 11,486 more ballots in 2024 than in 2022 (estimate).
  2. East Lansing (Ingham County) +978
    Starting line Lawrence +58. 1.3% split their ticket. About 4,601 more ballots in 2024 than in 2022 (estimate).
  3. Meridian Township (Ingham County) +461
    Starting line Lawrence +42. 1.5% split their ticket. About 3,840 more ballots in 2024 than in 2022 (estimate).
  4. Delta Township (Eaton County) +368
    Starting line Lawrence +23. 3.3% split their ticket.
  5. Delhi Township (Ingham County) +240
    Starting line Lawrence +20. 2.3% split their ticket.

The top 12 cities and townships carry 67% of the plan.

How

  • Persuade +3,584
    4,501 voters who split their ticket in recent elections and lean Barrett today.
  • Mobilize +3,181
    At least 30,237 people who voted in 2024 but not in 2022, in cities and townships where Lawrence's side wins at least 55 percent.
  • Consolidate +1,191
    Voters who went third party in recent races, in the cities and townships that lean Lawrence's way.

What has to be true

  1. Lawrence wins 40 percent of the ticket splitters now leaning Barrett.
  2. In Lansing (Ingham County) and East Lansing (Ingham County), 28 percent of the people who voted in 2024 but skipped 2022 come back and vote.
  3. No national wave is required. The plan uses 80 percent of what recent Michigan results show is possible.
  4. Published 2026 polls do not set the starting line here (every usable one comes from one side's sponsors, and the major ratings disagree); with Cook, Sabato and Inside Elections all at toss-up, the race is treated as a tie. An independent head-to-head poll would move it.

The numbers behind it

  • Crossover regression
    302 precincts, slope 1.021, R² 0.997, 46 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -35,702 to 52,074
  • Cash on hand, July 15
    Lawrence $383,116 · Barrett $2,871,872
  • Call program
    20,616 households, 61,849 dials, about 3,241 live conversations and 1,105 voter IDs, at $0.3468 per household (published HyperPhonebank rate)
Upset Brief · MI-10 · Christina Hines (D) vs. Michael Bouchard (R)

Hines can win. The plan uses 34 percent of what history allows.

Odds today43%
With the plan50%
Net votes to find4,254
Call program$3,850

Where

  1. Rochester Hills +1,366
    Starting line Hines +1. 7.1% split their ticket.
  2. St. Clair Shores +1,045
    Starting line Bouchard +4. 6.7% split their ticket.
  3. Eastpointe +461
    Starting line Hines +52. 4.3% split their ticket. About 4,862 more ballots in 2024 than in 2022 (estimate).
  4. Harrison Township (Macomb County) +416
    Starting line Bouchard +16. 6.6% split their ticket.
  5. Warren +387
    Starting line Hines +13. 6.0% split their ticket.

The top 10 cities and townships carry 100% of the plan.

How

  • Persuade +3,570
    10,468 voters who split their ticket in recent elections and lean Bouchard today.
  • Mobilize +331
    At least 7,864 people who voted in 2024 but not in 2022, in cities and townships where Hines's side wins at least 55 percent.
  • Consolidate +567
    Voters who went third party in recent races, in the cities and townships that lean Hines's way.

What has to be true

  1. Hines wins 17 percent of the ticket splitters now leaning Bouchard.
  2. In Eastpointe, 12 percent of the people who voted in 2024 but skipped 2022 come back and vote.
  3. No national wave is required. The plan uses 34 percent of what recent Michigan results show is possible.
  4. Published 2026 polls do not set the starting line here (every usable one comes from one side's sponsors, and the major ratings disagree), so the line is a weighted blend of official recent results. An independent head-to-head poll would move it.

The numbers behind it

  • Crossover regression
    288 precincts, slope 1.005, R² 0.982, 42 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -38,351 to 38,227
  • Cash on hand, July 15
    Hines $158,196 · Bouchard $1,144,520
  • Call program
    7,392 households, 22,175 dials, about 1,162 live conversations and 396 voter IDs, at $0.5208 per household (published HyperPhonebank rate)
Upset Brief · MI-01 · Callie Barr (D) vs. Jack Bergman (R) · incumbent

A long shot on the math: it takes a 15.1-point national wave.

Odds today<1%
With the plan2%
Net votes to find75,130
Call program$8,594

Where

  1. Marquette +524
    Starting line Barr +37. 4.8% split their ticket. About 2,147 more ballots in 2024 than in 2022 (estimate).
  2. Traverse City (Grand Traverse County) +402
    Starting line Barr +34. 5.2% split their ticket. About 1,206 more ballots in 2024 than in 2022 (estimate).
  3. Garfield Township (Grand Traverse County) +374
    Starting line Barr +4. 6.6% split their ticket.
  4. Sault Ste. Marie +240
    Starting line Bergman +5. 9.1% split their ticket.
  5. East Bay Township (Grand Traverse County) +226
    Starting line Bergman +6. 5.9% split their ticket.

The top 12 cities and townships carry 20% of the plan.

How

  • Persuade +13,245
    13,245 voters who split their ticket in recent elections and lean Bergman today.
  • Mobilize +607
    At least 7,061 people who voted in 2024 but not in 2022, in cities and townships where Barr's side wins at least 55 percent.
  • Consolidate +1,293
    Voters who went third party in recent races, in the cities and townships that lean Barr's way.

Plus a 15.1-point national shift (59,985 votes) that no campaign controls.

What has to be true

  1. Barr wins 50 percent of the ticket splitters now leaning Bergman.
  2. In Marquette and Traverse City (Grand Traverse 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 15.1 points toward Barr'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
    538 precincts, slope 0.973, R² 0.984, 79 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -105,792 to -13,673
  • Cash on hand, July 15
    Barr $530,725 · Bergman $660,163
  • Call program
    25,699 households, 77,097 dials, about 4,040 live conversations and 1,378 voter IDs, at $0.3344 per household (published HyperPhonebank rate)
Upset Brief · MI-03 · Terri DeBoer (R) vs. Hillary Scholten (D) · incumbent

A long shot on the math: it takes a 9.2-point national wave.

Odds today7%
With the plan9%
Net votes to find36,411
Call program$6,032

Where

  1. Grand Rapids +443
    Starting line Scholten +43. 0.7% split their ticket.
  2. Byron Township (Kent County) +423
    Starting line DeBoer +29. 1.0% split their ticket. About 3,093 more ballots in 2024 than in 2022 (estimate).
  3. Wyoming +311
    Starting line Scholten +10. 1.3% split their ticket.
  4. Allendale Township (Ottawa County) +295
    Starting line DeBoer +22. 1.1% split their ticket. About 2,967 more ballots in 2024 than in 2022 (estimate).
  5. Fruitport Township (Muskegon County) +225
    Starting line DeBoer +13. 2.9% split their ticket. About 2,014 more ballots in 2024 than in 2022 (estimate).

The top 12 cities and townships carry 62% of the plan.

How

  • Persuade +2,225
    2,225 voters who split their ticket in recent elections and lean Scholten today.
  • Mobilize +1,083
    At least 14,030 people who voted in 2024 but not in 2022, in cities and townships where DeBoer's side wins at least 55 percent.
  • Consolidate +1,249
    Voters who went third party in recent races, in the cities and townships that lean DeBoer's way.

Plus a 9.2-point national shift (31,854 votes) that no campaign controls.

What has to be true

  1. DeBoer wins 50 percent of the ticket splitters now leaning Scholten.
  2. In Byron Township (Kent County) and Allendale Township (Ottawa County), 35 percent of the people who voted in 2024 but skipped 2022 come back and vote.
  3. The national environment moves about 9.2 points toward DeBoer'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
    271 precincts, slope 0.995, R² 0.993, 40 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -72,079 to 7,874
  • Cash on hand, July 15
    DeBoer $361,355 · Scholten $1,356,573
  • Call program
    15,213 households, 45,638 dials, about 2,391 live conversations and 815 voter IDs, at $0.3965 per household (published HyperPhonebank rate)
Upset Brief · MI-02 · Ben Ambrose (D) vs. John Moolenaar (R) · incumbent

A long shot on the math: it takes a 25.4-point national wave.

Odds today0%
With the plan<1%
Net votes to find100,270
Call program$6,870

Where

  1. Mount Pleasant +370
    Starting line Ambrose +29. 5.7% split their ticket. About 1,708 more ballots in 2024 than in 2022 (estimate).
  2. Caledonia Township (Kent County) +260
    Starting line Moolenaar +29. 5.4% split their ticket.
  3. Algoma Township (Kent County) +162
    Starting line Moolenaar +26. 4.4% split their ticket.
  4. Union Township (Isabella County) +158
    Starting line Ambrose +6. 7.6% split their ticket.
  5. Cadillac +155
    Starting line Moolenaar +16. 7.0% split their ticket.

The top 12 cities and townships carry 17% of the plan.

How

  • Persuade +10,198
    10,198 voters who split their ticket in recent elections and lean Moolenaar today.
  • Mobilize +170
    At least 1,823 people who voted in 2024 but not in 2022, in cities and townships where Ambrose's side wins at least 55 percent.
  • Consolidate +1,012
    Voters who went third party in recent races, in the cities and townships that lean Ambrose's way.

Plus a 25.4-point national shift (88,890 votes) that no campaign controls.

What has to be true

  1. Ambrose wins 50 percent of the ticket splitters now leaning Moolenaar.
  2. In Mount Pleasant, 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 25.4 points toward Ambrose'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
    385 precincts, slope 0.981, R² 0.979, 64 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -129,399 to -47,605
  • Cash on hand, July 15
    Ambrose $21,868 · Moolenaar $1,905,561
  • Call program
    17,326 households, 51,979 dials, about 2,724 live conversations and 929 voter IDs, at $0.3965 per household (published HyperPhonebank rate)
Upset Brief · MI-05 · Christian Vukasovich (D) vs. Tim Walberg (R) · incumbent

A long shot on the math: it takes a 23.6-point national wave.

Odds today<1%
With the plan<1%
Net votes to find87,818
Call program$7,002

Where

  1. Bedford Township (Monroe County) +513
    Starting line Walberg +24. 6.4% split their ticket.
  2. Jackson +453
    Starting line Vukasovich +14. 6.0% split their ticket. About 3,709 more ballots in 2024 than in 2022 (estimate).
  3. Summit Township (Jackson County) +416
    Starting line Walberg +16. 7.2% split their ticket.
  4. Frenchtown Township (Monroe County) +330
    Starting line Walberg +24. 7.3% split their ticket.
  5. Blackman Township (Jackson County) +313
    Starting line Walberg +13. 8.6% split their ticket.

The top 12 cities and townships carry 33% of the plan.

How

  • Persuade +9,808
    9,808 voters who split their ticket in recent elections and lean Walberg today.
  • Mobilize +283
    At least 5,460 people who voted in 2024 but not in 2022, in cities and townships where Vukasovich's side wins at least 55 percent.
  • Consolidate +778
    Voters who went third party in recent races, in the cities and townships that lean Vukasovich's way.

Plus a 23.6-point national shift (76,949 votes) that no campaign controls.

What has to be true

  1. Vukasovich wins 50 percent of the ticket splitters now leaning Walberg.
  2. In Jackson and Albion, 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 23.6 points toward Vukasovich'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
    322 precincts, slope 0.982, R² 0.985, 42 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -115,411 to -38,776
  • Cash on hand, July 15
    Vukasovich $13,290 · Walberg $993,833
  • Call program
    18,837 households, 56,511 dials, about 2,961 live conversations and 1,010 voter IDs, at $0.3717 per household (published HyperPhonebank rate)
Upset Brief · MI-06 · Heather Smiley (R) vs. Debbie Dingell (D) · incumbent

A long shot on the math: it takes a 27.2-point national wave.

Odds today0%
With the plan<1%
Net votes to find108,687
Call program$4,671

Where

  1. Canton Township (Wayne County) +700
    Starting line Dingell +23. 2.3% split their ticket.
  2. Brownstown Township (Wayne County) +461
    Starting line Dingell +7. 5.5% split their ticket.
  3. Novi +389
    Starting line Dingell +17. 2.6% split their ticket.
  4. Huron Township (Wayne County) +334
    Starting line Smiley +18. 4.7% split their ticket. About 2,226 more ballots in 2024 than in 2022 (estimate).
  5. Ypsilanti Township (Washtenaw County) +314
    Starting line Dingell +52. 2.2% split their ticket.

The top 12 cities and townships carry 68% of the plan.

How

  • Persuade +4,398
    4,398 voters who split their ticket in recent elections and lean Dingell today.
  • Mobilize +230
    At least 4,034 people who voted in 2024 but not in 2022, in cities and townships where Smiley's side wins at least 55 percent.
  • Consolidate +820
    Voters who went third party in recent races, in the cities and townships that lean Smiley's way.

Plus a 27.2-point national shift (103,239 votes) that no campaign controls.

What has to be true

  1. Smiley wins 50 percent of the ticket splitters now leaning Dingell.
  2. In Huron Township (Wayne 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 27.2 points toward Smiley'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
    261 precincts, slope 0.929, R² 0.989, 39 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -146,680 to -59,892
  • Cash on hand, July 15
    Smiley n/a · Dingell $414,935
  • Call program
    10,469 households, 31,408 dials, about 1,646 live conversations and 561 voter IDs, at $0.4462 per household (published HyperPhonebank rate)
Upset Brief · MI-09 · Ray Pooley (D) vs. Lisa McClain (R) · incumbent

A long shot on the math: it takes a 26.7-point national wave.

Odds today0%
With the plan<1%
Net votes to find118,333
Call program$7,305

Where

  1. Macomb Township (Macomb County) +1,382
    Starting line McClain +26. 7.6% split their ticket.
  2. Chesterfield Township (Macomb County) +988
    Starting line McClain +23. 8.0% split their ticket.
  3. Orion Township (Oakland County) +822
    Starting line McClain +12. 7.4% split their ticket.
  4. Independence Township (Oakland County) +788
    Starting line McClain +19. 7.1% split their ticket.
  5. Washington Township (Macomb County) +527
    Starting line McClain +38. 6.4% split their ticket.

The top 12 cities and townships carry 48% of the plan.

How

  • Persuade +13,493
    13,493 voters who split their ticket in recent elections and lean McClain today.
  • Mobilize +0
    Few cities and townships give Pooley's side 55 percent or more, so turnout is a small lever here; the plan leans on persuasion instead.
  • Consolidate +1,240
    Voters who went third party in recent races, in the closest places on the map.

Plus a 26.7-point national shift (103,600 votes) that no campaign controls.

What has to be true

  1. Pooley wins 50 percent of the ticket splitters now leaning McClain.
  2. It would take a national shift of about 26.7 points toward Pooley's side, larger than any midterm swing in recent memory.
  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
    325 precincts, slope 0.984, R² 0.982, 63 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -148,455 to -58,064
  • Cash on hand, July 15
    Pooley $4,942 · McClain $1,536,890
  • Call program
    21,065 households, 63,195 dials, about 3,311 live conversations and 1,129 voter IDs, at $0.3468 per household (published HyperPhonebank rate)
Upset Brief · MI-11 · Ethan Baker (R) vs. Jeremy Moss (D)

A long shot on the math: it takes a 19.2-point national wave.

Odds today<1%
With the plan<1%
Net votes to find79,890
Call program$5,575

Where

  1. Troy +918
    Starting line Moss +12. 4.0% split their ticket.
  2. Waterford Township (Oakland County) +799
    Starting line Baker +3. 4.3% split their ticket.
  3. Farmington Hills +683
    Starting line Moss +32. 3.0% split their ticket.
  4. West Bloomfield Township (Oakland County) +646
    Starting line Moss +19. 3.1% split their ticket.
  5. Royal Oak +610
    Starting line Moss +36. 3.3% split their ticket.

The top 12 cities and townships carry 83% of the plan.

How

  • Persuade +6,541
    6,541 voters who split their ticket in recent elections and lean Moss today.
  • Mobilize +203
    At least 3,966 people who voted in 2024 but not in 2022, in cities and townships where Baker's side wins at least 55 percent.
  • Consolidate +744
    Voters who went third party in recent races, in the cities and townships that lean Baker's way.

Plus a 19.2-point national shift (72,403 votes) that no campaign controls.

What has to be true

  1. Baker wins 50 percent of the ticket splitters now leaning Moss.
  2. In White Lake Township (Oakland 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 19.2 points toward Baker'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
    287 precincts, slope 0.992, R² 0.992, 45 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -115,894 to -28,901
  • Cash on hand, July 15
    Baker $121,125 · Moss $352,517
  • Call program
    13,229 households, 39,687 dials, about 2,080 live conversations and 709 voter IDs, at $0.4214 per household (published HyperPhonebank rate)
Upset Brief · MI-12 · James D. Hooper (R) vs. Rashida Tlaib (D) · incumbent

A long shot on the math: it takes a 40.1-point national wave.

Odds today0%
With the plan0%
Net votes to find125,325
Call program$6,155

Where

  1. Dearborn +3,355
    Starting line Tlaib +21. 19.2% split their ticket.
  2. Livonia +2,149
    Starting line Tlaib +2. 7.1% split their ticket.
  3. Dearborn Heights +1,328
    Starting line Tlaib +23. 20.8% split their ticket.
  4. Westland +1,000
    Starting line Tlaib +20. 5.1% split their ticket.
  5. Garden City +524
    Starting line Hooper +2. 7.3% split their ticket.

The top 11 cities and townships carry 100% of the plan.

How

  • Persuade +8,940
    8,940 voters who split their ticket in recent elections and lean Tlaib today.
  • Mobilize +0
    Few cities and townships give Hooper's side 55 percent or more, so turnout is a small lever here; the plan leans on persuasion instead.
  • Consolidate +1,150
    Voters who went third party in recent races, in the closest places on the map.

Plus a 40.1-point national shift (115,234 votes) that no campaign controls.

What has to be true

  1. Hooper wins 50 percent of the ticket splitters now leaning Tlaib.
  2. It would take a national shift of about 40.1 points toward Hooper's side, larger than any midterm swing in recent memory.
  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
    408 precincts, slope 0.770, R² 0.737, 36 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -149,282 to -81,583
  • Cash on hand, July 15
    Hooper $8,381 · Tlaib $3,434,433
  • Call program
    14,605 households, 43,814 dials, about 2,296 live conversations and 783 voter IDs, at $0.4214 per household (published HyperPhonebank rate)
Upset Brief · MI-13 · Taras "T.P." Nykoriak (R) vs. Donavan McKinney (D)

A long shot on the math: it takes a 42.9-point national wave.

Odds today0%
With the plan0%
Net votes to find111,252
Call program$4,209

Where

  1. Detroit +1,747
    Starting line McKinney +82. 3.2% split their ticket.
  2. Taylor +636
    Starting line McKinney +10. 4.3% split their ticket.
  3. Allen Park +414
    Starting line McKinney +6. 4.8% split their ticket.
  4. Southgate +388
    Starting line McKinney +11. 5.0% split their ticket.
  5. Lincoln Park +376
    Starting line McKinney +18. 4.7% split their ticket.

The top 12 cities and townships carry 88% of the plan.

How

  • Persuade +4,996
    4,996 voters who split their ticket in recent elections and lean McKinney today.
  • Mobilize +28
    Few cities and townships give Nykoriak's side 55 percent or more, so turnout is a small lever here; the plan leans on persuasion instead.
  • Consolidate +936
    Voters who went third party in recent races, in the closest places on the map.

Plus a 42.9-point national shift (105,292 votes) that no campaign controls.

What has to be true

  1. Nykoriak wins 50 percent of the ticket splitters now leaning McKinney.
  2. It would take a national shift of about 42.9 points toward Nykoriak's side, larger than any midterm swing in recent memory.
  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
    429 precincts, slope 0.925, R² 0.989, 74 crossover precincts
  • Monte Carlo
    20,000 seeded simulations, 7-point baseline error. 90% range of the final margin with the plan: -133,870 to -76,887
  • Cash on hand, July 15
    Nykoriak n/a · McKinney $142,854
  • Call program
    8,935 households, 26,804 dials, about 1,405 live conversations and 479 voter IDs, at $0.4711 per household (published HyperPhonebank rate)

Independent analysis by The Political Group from official Michigan Bureau of Elections precinct returns, FEC filings and published polls. Baseline error and plan efficacy are stated assumptions shown on each race page. Not affiliated with any candidate or party. Text or call Frank Carrasco, 469.639.1600.