OHIO ELECTION MODELED FORECAST: EV, VBM, VF, PH, CC, EDS How The Model Works: Every early voter in the Secretary of State’s file is matched to the statewide voter file, so we know who is voting early, county by county, against who voted early in 2024, 2022 and 2018. The model projects the final early vote from the mail ballots still out and the in-person pace, then estimates how that early electorate votes using each county’s own history. The early vote comes out to D+6.4 in projected votes. Election Day cannot be observed, so it is set as a scenario, and the four screenshots attached show four of them: an Election Day vote of R+6, R+8, R+10 and R+12.8. These are projected vote margins among the people who vote on Nov 3, not party labels. Polls are NOT represented in these models. Everything comes from SOS files and county canvass reports. Canvass data tells us what Election Day usually looks like. Official results broken out by voting method were gathered from 87 of Ohio’s 88 counties for 2024, from 51 counties for 2022, and from 26 counties for 2018. The 2018 sample is the thinnest and leans heavily Democratic, as big urban counties make up most of it, meaning it understates the Republican Election Day. In every race, in every county with a canvass on file, the people who vote on Election Day have been far more Republican than the people who vote early. The 2024 presidential Election Day vote was R+20. The 2022 governor Election Day vote was about R+30. Even the two best Democratic performances in the set, Sherrod Brown in 2024 and Tim Ryan in 2022, had Election Day votes of R+11 and R+13. The 2018 Democratic wave, with Cordray at the top of the ticket, still produced an Election Day somewhere near R+8. Image One: Model, Election Day R+6 An R+6 Election Day would be the worst Republican Election Day in Ohio in at least a decade. It is below the 2018 wave. It is below what Brown and Ryan faced. This is deliberately built as the most conservative case the data can defend, the one where Republicans simply fail to show up on Nov 3. Even then the model has Republicans winning the state by about 1.7 points, with Democrats carrying 11 of 88 counties. R+6 is NOT realistic, this is the floor. Image Two & Three: Model, Election Day R+8 & R+10 R+8 and R+10 are the generous end of the realistic range. R+8 matches the 2018 wave year and gives Republicans about 3 points. R+10 is still a point short of what Brown faced in 2024 and gives Republicans about 4.3 points, carrying 79 counties. Image Four: Model, Election Day R+12.8 R+12.8 is the scenario that qualifies as ordinary. It sits right where the two best Democratic candidates of the decade landed on Election Day, between Brown's R+11 and Ryan's R+13, and it is still 7 points better for Democrats than the 2024 presidential Election Day. At R+12.8 the model shows Republicans winning by about 6 points, 53 to 47, with Democrats holding 8 counties. If Election Day simply repeats the county-by-county gap measured in the 2024 canvasses, the result lands near R+8 to R+10, beyond even that. So what would Democrats need to win? Two things, both run counter to the data: First, the early vote would have to move the other way. It stands at D+4.5 by party label today and has fallen every week since it opened at D+18. It would need to climb back to D+10 or better over the final week, when the daily vote has been running about even and Monday went Republican. Second, Democrats would need an Election Day that is close to even or better. No measured Ohio race in the canvass data comes close to that. The only route is a candidate effect of the size Tim Ryan produced in 2022, which was worth about 20 points of Election Day margin, and nothing in the early vote suggests that kind of candidate effect is present this year. How we got here: This data has been run as many ways as possible: county composition models, measured Election Day gaps from the canvasses, historical anchors on 2018, 2022 and 2024, different weightings between midterm and presidential years. They only disagree with each other by a few points. None land on a Democratic win. Across the four scenarios the Republican margin runs from 1.7 points to 6 points, and the one scenario that resembles a normal Ohio Election Day is at the top of that range. The honest reading of the Ohio early vote is that it is very difficult for Democrats to win, and the more conservative the Election Day assumption, the more that conclusion stands out. This projection will be updated again Friday, and Monday next week. Monday will be the final modeled forecast.

Collins +3 at 50% in NYTimes OR Jackson +7 at 51% in Echelon Insights Choose your own adventure!