How to watch tonight's Wisconsin Supreme Court election like a pro

Get my county-level election benchmarks, and a Substack crossover watch party with The Downballot

Note: While this post is free to read publicly today, it is an example of premium work I’d give to paying subscribers in the future. If you like empirical political journalism or find election-night benchmarks helpful, you can signal your support for this work by pledging a subscription (at no initial cost to you) here.

---

Tonight, Democrats have their first chance to fight back against Donald Trump and reverse some of their party’s losses in the 2024 election — and Republicans have a shot to score a big judicial victory in a court currently controlled narrowly by liberals, and in a state that is key to the presidency and control of the U.S. House of Representatives. It’s a race everyone is watching and people are spending significant sums on. I’m talking of course about the Wisconsin Supreme Court race between liberal Susan Crawford and conservative Brad Schimel.

Odds are you have probably heard of the election from the coverage of Elon Musk’s involvement, which has included him spending $20 million in television and digital advertising as well as giving away checks for $1 million to random rally goers this weekend (which is, apparently and shockingly, not illegal). But the stakes are significant: The Wisconsin Supreme Court has recently decided cases on gerrymandering, campaign finance, and voting rights, and would have jurisdiction over a pending abortion case and important electoral cases before the 2028 presidential race. Across all parties, nearly $100 million has been spent on the race.

Prognosticators mostly expect Crawford to narrowly win the race, with room for uncertainty and a small Schimel victory. Crawford has led most of the polls conducted of the race, and the line at Split Ticket is that Republicans have an off-year turnout problem that tilts the scales against them. You can apply a similar logic from my ”dual electorates” piece and draw the conclusion that Schimel is likely to have a bad time, though a win is not impossible. The prediction market Kalshi (I know) gives Crawford an 84% chance (the markets tend to overestimate odds for losers, so her real odds might be higher than this).

Whatever the odds, what really matters is who votes, and here is how to watch the results like a nerd pro:

1. Compare results to my county benchmarks

One of the tools we used often on election nights at 538 was a huge Excel spreadsheet of estimated margins in each county that each candidate needed to hit in order to be on track for victory. Well, I don’t have a team of researchers and journalists to rely on to do that work here at Strength In Numbers, so I’ve put in the work of estimating what the Supreme Court elections results in each of Wisconsin’s 72 counties would be in the case that the statewide result is exactly tied.¹ And by way of credit: As far as I know, the idea of benchmarks was developed by Geoffrey Skelley at 538 for the 2022 midterms. You can read him at geoffreyskelley.com.

I have uploaded these benchmarks to a public Google spreadsheet that I will update throughout the night. Right now, all the data in the columns related to 2025 are based on randomly generated data, but I will update cells live starting at 9:00 PM ET tonight.

County-level benchmarks for the 2025 Wisconsin Supreme Court race

The way to use this spreadsheet is to open it in a tab on your computer, and check back every few minutes as I go through and update results based on other public sources.

The benchmarks are cool, but they do have one shortcoming: They are just estimates and come with uncertainty. When there are significant changes in turnout between presidential- and off-year elections, the margin each candidate needs to hit in each county in order to win can be quite different. For example, if turnout is lower in rural areas tonight than in Nov 2024, Crawford will need fewer votes from Milwaukee than the presidential-based benchmarks indicate. We could account for this by using old Supreme Court elections in our benchmark calculations, but this risks missing recent changes in subgroup voter loyalty.

This means you should take the benchmarks as a generally good measure of who is ahead, but not a perfect predictor. I wouldn’t use the benchmarks alone to call an election, for example, but if you can analyze turnout too and one candidate is beating the benchmarks pretty much everywhere, you’ll have everything you need.

2. Join the The Downballot-Strength In Numbers watch party!

One place to get that extra information is The Downballot. The fine folks there have a nice preview of the race out this morning, and will be hosting a live video stream reporting on the results. I’ll be joining them to track these benchmarks and share any thoughts (and charts!) on turnout and other factors that are relevant to predicting the election result. The live stream will be up from 9:00 ET until the race is called (or until we got to bed).

This should be fun, and I hope you all join. To be sure, while I know we like to watch the results come in on election nights, I acknowledge the stakes of the contest are high too, so it’s good to remember we’re watching democracy and history unfold, not just a horse race. More on the stakes here.

I hope you will all tune into the stream so we can claim the newsletter has critical mass, or at least pull up the benchmarks and my social feeds. If this is helpful in your toolkit for watching returns, let me know with a comment or email. Also, the data analysis behind this post and the live models takes time and computer infrastructure, so if you want to see me continue doing that work please pledge a paid subscription to the Substack.

1. I did this by taking the county-level 2024 presidential election results in Wisconsin and adjusting the Democratic margin in each county (by adding votes for the Democratic candidate, Kamala Harris, and subtracting votes for Donald Trump) until the two candidates were exactly tied statewide. The county margins in this scenario make up the benchmark of a truly tied race. If Crawford is beating her benchmarks in many counties, she’s likely on pace to win; vice-versa for Schimel.