← kevinmdeluca.com  ·  PLSC 2219 · Yale University

Election Forecast Tracker

Six ways of forecasting the 2026 midterms, collected every day and shown side by side. The display is the disagreement between methods, not another trend line.

Election day 3 Nov 2026 days out Snapshot 2026-09-09 598 dates on the chart
This page has not updated in days. The daily collection may have stopped. Everything below is still the snapshot named above, not today's.
Contest

The seat that decides control

Every Senate seat, ordered most Democratic to most Republican, with each one's expected margin as a bar.

R+20+R+10EVEND+10D+20+MA — expected D+35.5, P(D) 100%. Democratic seat number 35.MADE — expected D+25.6, P(D) 100%. Democratic seat number 36.DERI — expected D+25.5, P(D) 100%. Democratic seat number 37.RIOR — expected D+25.0, P(D) 100%. Democratic seat number 38.ORCO — expected D+22.3, P(D) 100%. Democratic seat number 39.CONJ — expected D+19.5, P(D) 100%. Democratic seat number 40.NJNM — expected D+18.5, P(D) 100%. Democratic seat number 41.NMVA — expected D+18.2, P(D) 100%. Democratic seat number 42.VAIL — expected D+17.2, P(D) 100%. Democratic seat number 43.ILGA — expected D+11.1, P(D) 98%. Democratic seat number 44.GAMN — expected D+10.9, P(D) 98%. Democratic seat number 45.MNNH — expected D+10.0, P(D) 97%. Democratic seat number 46.NHME — expected D+7.3, P(D) 91%. Democratic seat number 47.MEMI — expected D+6.4, P(D) 88%. Democratic seat number 48.MINC — expected D+4.4, P(D) 80%. Democratic seat number 49.NCTX — expected R+3.0, P(D) 29%. Democratic seat number 50.TXIA — expected R+3.3, P(D) 27%. Democratic seat number 51.IAOH — expected R+7.6, P(D) 8%. Democratic seat number 52.OHFL — expected R+7.8, P(D) 7%. Democratic seat number 53.FLMT — expected R+9.4, P(D) 4%. Democratic seat number 54.MTAK — expected R+9.6, P(D) 4%. Democratic seat number 55.AKKS — expected R+12.2, P(D) 1%. Democratic seat number 56.KSSC — expected R+12.6, P(D) 1%. Democratic seat number 57.SCNE — expected R+15.9, P(D) 0%. Democratic seat number 58.NELA — expected R+16.6, P(D) 0%. Democratic seat number 59.LAMS — expected R+16.7, P(D) 0%. Democratic seat number 60.MSAL — expected R+17.6, P(D) 0%. Democratic seat number 61.ALKY — expected R+17.9, P(D) 0%. Democratic seat number 62.KYOK — expected R+21.8, P(D) 0%. Democratic seat number 63.OKTN — expected R+22.4, P(D) 0%. Democratic seat number 64.TNSD — expected R+22.9, P(D) 0%. Democratic seat number 65.SDAR — expected R+24.2, P(D) 0%. Democratic seat number 66.ARID — expected R+28.6, P(D) 0%. Democratic seat number 67.IDWY — expected R+31.5, P(D) 0%. Democratic seat number 68.WYWV — expected R+33.5, P(D) 0%. Democratic seat number 69.WVPolling: 49.1 Democratic seatsProfessional: 49.3 Democratic seatsAcademic: 49.7 Democratic seatsMarkets: 50.1 Democratic seatsFundamentals: 50.2 Democratic seatsThis class: 50.3 Democratic seats50th · TX51st · IA394449545964← 34 D seats not on the ballot31 R seats not on the ballot →Democratic seat count
Bar colour is certainty: safe D · toss-up · safe R
Polling 49.1 Professional 49.3 Academic 49.7 Markets 50.1 Fundamentals 50.2 This class 50.3 the seat that decides control
ThresholdSeatExpected marginP(D)
50th — a tie, broken by the vice-presidentTXR+3.029%
51st — an outright majorityIAR+3.327%

Margins beyond ±20 points are clipped and notched. Probabilities assume 34 Democratic holdover seats; one seat moves them about twenty points.

Polling model

Democratic-leaning

6 of the 13 unsettled races. Seats the model has the Democrats winning, closest first.

StateMarginP(D)Lean
NCD+4.480%R+1.4
MID+6.488%R+0.2
MED+7.391%D+3.8
NHD+10.097%D+2.0
MND+10.998%D+2.5
GAD+11.198%R+0.8
Polling model

Republican-leaning

7 of the 13 unsettled races. Seats the model has the Republicans winning, closest first.

StateMarginP(D)Lean
TXR+3.029%R+5.9
IAR+3.327%R+6.1
OHR+7.68%R+5.3
FLR+7.87%R+5.4
MTR+9.44%R+9.8
AKR+9.64%R+6.5
KSR+12.21%R+8.0

These are the polling model's numbers, not a consensus of forecasters — the election-day national tide it derives from the generic ballot, carried to each state by twice that state's partisan lean. Nothing in it knows who the candidates are. What the other three methods say about the same races is on the comparisons page. Races where the model puts the Democratic win probability between 1% and 99%. The rest are not close under any assumption this model makes.

The seat that decides control

The 45 districts around the majority line, ordered by expected margin. The 435 districts are ranked most Democratic to most Republican; safe seats at each end are collapsed.

R+10+R+5EVEND+5D+10+NY-17 — expected D+5.3, P(D) 69%. Democratic seat number 205.NY-17CA-06 — expected D+5.2, P(D) 69%. Democratic seat number 206.CA-06AL-02 — expected D+5.2, P(D) 69%. Democratic seat number 207.AL-02FL-25 — expected D+5.1, P(D) 69%. Democratic seat number 208.FL-25WA-03 — expected D+4.7, P(D) 68%. Democratic seat number 209.WA-03PA-01 — expected D+4.1, P(D) 65%. Democratic seat number 210.PA-01NC-01 — expected D+3.9, P(D) 65%. Democratic seat number 211.NC-01CA-22 — expected D+3.4, P(D) 63%. Democratic seat number 212.CA-22TX-15 — expected D+3.3, P(D) 62%. Democratic seat number 213.TX-15NE-02 — expected D+1.6, P(D) 56%. Democratic seat number 214.NE-02PA-07 — expected D+1.5, P(D) 56%. Democratic seat number 215.PA-07AZ-01 — expected D+1.2, P(D) 55%. Democratic seat number 216.AZ-01PA-08 — expected D+0.9, P(D) 53%. Democratic seat number 217.PA-08MT-01 — expected D+0.9, P(D) 53%. Democratic seat number 218.MT-01TX-35 — expected D+0.4, P(D) 52%. Democratic seat number 219.TX-35MI-07 — expected D+0.3, P(D) 51%. Democratic seat number 220.MI-07CO-08 — expected D+0.2, P(D) 51%. Democratic seat number 221.CO-08TX-09 — expected R+0.3, P(D) 49%. Democratic seat number 222.TX-09OH-09 — expected R+0.5, P(D) 48%. Democratic seat number 223.OH-09WI-03 — expected R+0.7, P(D) 47%. Democratic seat number 224.WI-03KY-06 — expected R+1.0, P(D) 46%. Democratic seat number 225.KY-06FL-14 — expected R+1.3, P(D) 45%. Democratic seat number 226.FL-14IA-02 — expected R+1.4, P(D) 44%. Democratic seat number 227.IA-02WI-01 — expected R+2.1, P(D) 42%. Democratic seat number 228.WI-01VA-02 — expected R+2.7, P(D) 40%. Democratic seat number 229.VA-02IA-03 — expected R+3.8, P(D) 36%. Democratic seat number 230.IA-03NH-01 — expected R+4.2, P(D) 35%. Democratic seat number 231.NH-01IA-01 — expected R+4.3, P(D) 34%. Democratic seat number 232.IA-01FL-22 — expected R+4.4, P(D) 34%. Democratic seat number 233.FL-22AZ-06 — expected R+4.6, P(D) 33%. Democratic seat number 234.AZ-06MO-05 — expected R+4.7, P(D) 33%. Democratic seat number 235.MO-05PA-10 — expected R+4.8, P(D) 32%. Democratic seat number 236.PA-10NJ-07 — expected R+5.0, P(D) 31%. Democratic seat number 237.NJ-07MN-01 — expected R+5.6, P(D) 30%. Democratic seat number 238.MN-01MN-08 — expected R+5.9, P(D) 28%. Democratic seat number 239.MN-08NJ-02 — expected R+6.0, P(D) 28%. Democratic seat number 240.NJ-02NC-11 — expected R+6.8, P(D) 26%. Democratic seat number 241.NC-11ME-02 — expected R+7.2, P(D) 25%. Democratic seat number 242.ME-02FL-09 — expected R+7.4, P(D) 24%. Democratic seat number 243.FL-09MI-04 — expected R+7.4, P(D) 24%. Democratic seat number 244.MI-04SC-01 — expected R+7.8, P(D) 23%. Democratic seat number 245.SC-01NC-03 — expected R+8.2, P(D) 21%. Democratic seat number 246.NC-03NC-07 — expected R+8.3, P(D) 21%. Democratic seat number 247.NC-07TX-32 — expected R+8.3, P(D) 21%. Democratic seat number 248.TX-32NY-01 — expected R+8.3, P(D) 21%. Democratic seat number 249.NY-01Polling: 224.2 Democratic seatsProfessional: 225.1 Democratic seatsAcademic: 229.2 Democratic seatsMarkets: 230.7 Democratic seatsFundamentals: 234.1 Democratic seatsThis class: 235.5 Democratic seats218th · MT-01209214219224229234239244← 204 safe Democratic seats186 safe Republican seats →Democratic seat count
Bar colour is certainty: safe D · toss-up · safe R
Polling 224.2 Professional 225.1 Academic 229.2 Markets 230.7 Fundamentals 234.1 This class 235.5 the seat that decides control
ThresholdSeatExpected marginP(D)
218th — a majorityMT-01D+0.953%

Margins beyond ±10 points are clipped and notched. District margins are our forecast; the partisan index behind them is not published — why.

Polling model

Above the line

Seats 205–227, ordered by expected margin.

SeatDistrictMarginP(D)
205NY-17D+5.369%
206CA-06D+5.269%
207AL-02D+5.269%
208FL-25D+5.169%
209WA-03D+4.768%
210PA-01D+4.165%
211NC-01D+3.965%
212CA-22D+3.463%
213TX-15D+3.362%
214NE-02D+1.656%
215PA-07D+1.556%
216AZ-01D+1.255%
217PA-08D+0.953%
218 ◀MT-01D+0.953%
219TX-35D+0.452%
220MI-07D+0.351%
221CO-08D+0.251%
222TX-09R+0.349%
223OH-09R+0.548%
224WI-03R+0.747%
225KY-06R+1.046%
226FL-14R+1.345%
227IA-02R+1.444%
Polling model

Below the line

Seats 228–249, ordered by expected margin.

SeatDistrictMarginP(D)
228WI-01R+2.142%
229VA-02R+2.740%
230IA-03R+3.836%
231NH-01R+4.235%
232IA-01R+4.334%
233FL-22R+4.434%
234AZ-06R+4.633%
235MO-05R+4.733%
236PA-10R+4.832%
237NJ-07R+5.031%
238MN-01R+5.630%
239MN-08R+5.928%
240NJ-02R+6.028%
241NC-11R+6.826%
242ME-02R+7.225%
243FL-09R+7.424%
244MI-04R+7.424%
245SC-01R+7.823%
246NC-03R+8.221%
247NC-07R+8.321%
248TX-32R+8.321%
249NY-01R+8.321%
Why so many districts share a margin. Every district margin is the same national tide plus twice that district’s lean, and the lean is carried at low resolution, so the margins land on a coarse grid: 435 districts take only 45 drawn values, and a dozen or more can share one. It is arithmetic rather than a bug, but it has a consequence worth knowing: inside a tied group the seat ORDER is arbitrary, so the district named at the majority line is one of several that could equally have been named. Read it as “the majority is decided around here”, not as a call on that district. How the mapping works.

◀ marks the seat that reaches the majority. Showing 45 of 435 districts; the rest are safe at either end. The full set is in the published data files.