
You may or may not know this, but we have a robust and very thoughtfully made midterms election hub on our website. It ingests polls, markets, models — everything you would need to know if the polls are right… or if you should believe the markets.
We are, of course, leaning with markets. And there’s something interesting going on there.
The example Dane Martinez used in his latest video was Alaska’s Senate race between Mary Peltola (D) and Dan Sullivan (R). Kalshi currently lists the race as 75% Peltola and 25% Sullivan. ProphetX has it 76.19% for Peltola and 26.32% for Sullivan. The polls, however, have it essentially a toss-up, with 50plusOne.com (which is run by former 538 staffers) reading the polls as Peltola up by about 2 points.

Keep in mind, the polls are just asking a small group of people who they’re voting for, and this is what people are answering. The markets aren’t saying the election gap will be that big, just that as of right now, considering the polls and other elements, there’s a 75-76% chance that Peltola wins. So there’s a 25-ish percent chance she doesn’t.
And just to keep us honest with the “well hey anything can happen and everyone could be wrong about everything!” curveball, I will offer this: We all remember Francesca Hong, right?

The point of today’s newsletter isn’t to play the “this is right or wrong” game with polls vs. markets. We already did that with Kansas last week.
Instead, it’s to dive into the midterm elections and see if there really are any toss-up races left that both the polls and the markets can agree on. Because they seem to be few and far between. And then we’ll dive into a smaller Wisconsin race and dissect what makes it a toss-up.
You will not benefit monetarily from today’s newsletter if you love longshots, like I do. The best I can offer is maybe a 44 cent buy! But this is interesting, and maybe the conclusion is this: after November 3, we will have a huge amount of new data to play with.
Let’s start in Los Angeles, where the mayor’s race has a 56/44 split for Karen Bass, but the graph has been moving — and flipping — over the last month:

The winner there, however, is relatively inconsequential compared to Senate and gubernatorial races. But it’s a close one, at least! Polls and markets can’t seem to figure this out, just yet. And, as you’ll see, “close polls” and “close markets” are infrequent bedfellows.
The biggest markets toss-up on the national stage is probably the Ashley Hinson (R)/Josh Turek (D) race for Iowa’s senate seat, currently at 58/42 Hinson.

50plusOne has it much tighter, with Turek currently in the polls lead:

What does this mean? Even with a race this tight in the polls, a $6 million market has Turek as an underdog, where $20 can bring back $46.80 on ProphetX if you go with Turek:

But again, the point here is that this is a toss-up in the polls and the markets. We’re just poking around and setting up the lab before combining everything into a beaker, which unfortunately won’t combust until 10 days from now. Highlight this one, though.
Probably the most interesting of the midterm races is in Georgia, where Keisha Lance Bottoms and Rick Jackson are running for governor and are in a 54/46 “deadlock” with several graph flips since June:

And as opposed to the Turek/Hinson parallel lines, this 50plusOne polling has some opposite movement:

Even if you comb through the numerous House races, there are very few running neck-and-neck on the markets. The close ones include:
…and on and on. This isn’t meant to be a comprehensive run-through of all the close races, but rather a look at how polls may be 2 points apart — and within the margin of error — while markets for the same race could have a 20% gap… or more. It gets interesting when you put a magnifying glass up to the close ones.
Steil vs. Berman, for instance, is a $166K market for the WI-01 seat.
Berman, you may have noticed, is wearing an American flag bandana here:

Pull that official campaign photo back and — surprise!

Berman is an emergency room nurse!
Here’s an awkwardly-seated debate between the two of them from Tuesday:

For those of you wondering — because I was wondering — the district is pretty normal; no giant cities and it “leans Republican,” having voted mostly for George W. Bush, Barack Obama, Mitt Romney, and Donald Trump thrice. But as you can see here, the margins have been tight:

It reminds me of what The 7 Oracles co-owner Gene Clemons said on a recent stream while we were talking about the Miami Dolphins — they haven’t been getting blown out. Yes, they’re losing, but the games have been relatively close, for the most part. And that’s something bubbling under the surface that goes beyond the bad record that some people may not care to notice.
WI-01 similarly has gone Republican a good amount, but… it’s been close. Enough frustration and annoyance bubbles up and you might see a Berman upset.
Unfortunately, for my fellow lovers of longshots, I think that chance is baked in. We just reverse-engineered that 56-44 split — but it is the autopsy of a markets “toss up” in a world of close polling but wildly different market numbers.
This is how you get a toss-up on Kalshi, in other words — by having a voting history that can be compared to the 2026 Miami Dolphins.
How will I throw my $20 around this week? I’d hold. But… if you really must pry(“I must! I must!”) — $10 on Turek and $10 on Berman.
Good luck everyone! We’re at the NEXTPredict conference the next few days. If you’re around, come say hello!