Benjamin Freeman

BIG NEWS: I’m joining the Kalshi team! My main goal is going to be growing Kalshi’s political / election markets in terms of visibility, volume, and influence, and showing the world why these markets are so important.

This means that I have given up trading on Kalshi. While I loved being a trader and loved giving away “alpha” to my followers, I can no longer do so due to Kalshi’s policies. It was a hard tradeoff for sure. From my days of interning at the White House, starting my own election prediction Youtube channel, to starting trading on Kalshi’s election markets in October 2024, politics, and specifically predicting elections has always been my passion. Since starting tweeting in October 2025, I wanted to show a new audience of why I believe in the power of election prediction markets. I’ve 20x’ed my followers since starting then, and have had a blast doing it along the way with y’all.

Now, my goal is to drive liquidity to these markets and make them a bigger part of the cultural zeitgeist and political landscape. Having very liquid, high volume political prediction markets makes our political system better. More accurate forecasts benefit traders and non-traders alike, and growth in political prediction markets translates into higher participation and interest in our political system, which is good for democracy. These markets are built on rationality and probabilistic thinking, which is desperately needed in today’s corrupted information environment. These markets can be incredibly useful for campaigns, political commentators, casual political viewers, fundraisers / bundlers, investors wanting to hedge political risk, investors wanting political exposure, small dollar campaign donors, academics, and everyone in between. Political prediction markets are for everyone. I’m really excited to be a part of the team that mainstreams these markets.

Feel free to reach out to me, if you have any ideas on how to grow Kalshi’s political / election markets! Thank you, all!

Before the Midterms, I kept on telling people that I thought the 2026 Midterms were the “easiest election cycle to predict.”

Not that picking the exact seat count for Democrats and Republicans in the Senate, House, and Governor was easy (I underestimated Democrats by forecasting 249 Democratic House seats, 54 Senate seats, and 29 Governor seats), but directionally, I thought the market and the models were underestimating Democrats (which I was right).

Here’s why I was so confident in a blue tsunami in 2026:
1. Worse R environment than 2018
2. 2025 election results
3. Candidate quality differential
4. Campaign quality differential
5. Early vote tea leaves

#1: Because this was the 2nd Trump Midterm, I thought it made absolute sense to set 2018 as one’s “base case” of what the 2026 Midterms would be.

But what was different from 2018?

Well, Trump’s approval was much lower, inflation was worse, and Republicans didn’t have as a strong of a reason to turnout.

In 2018, I thought Republicans’ closing message to turnout was actually decent. Trump would say at his closing rallies, get out and vote for “Kavanaugh, the caravan, and common sense.”

The 2018 Midterms came right after the Kavanaugh hearings which did energize Republicans, and the immigration issue was put more into picture with the “caravan.”

Also, the economy was roaring. Even though Democrats did well, it was not a blue tsunami like 2006.

In 2026, Trump’s approval was much lower, the perception of the economy was much worse, and immigration was way less salient of an issue.

Unlike 2018, where I thought Republicans had a decent closing message, it was unclear to me what was going to motivate the marginal Republican voter to get out.

For this reason, I thought Republicans’ floor was a D+10 generic ballot with the potential of it being a lot worse.

#2: In 2025, the polls vastly underestimated Democrats (especially in NJ Governor), and Hispanics swung sharply to the left. Many pollsters were making the same mistakes that the 2025 VA and NJ polls were making, so I thought it was pretty reasonable that polls would underestimating Democrats again in 2026 (not the opposite like in 2024)

+ the political environment had gotten worse for Republicans in Nov. 2026 compared to 2025

#3: Chuck Schumer did an exceptional job of recruiting Democratic Senate candidates in Alaska, Ohio, North Carolina, etc, and the most “electable” Democrats generally won their primaries.

The same cannot be said on the Republican side, but I do think the Democrats just did a very good job, and the idea that Republicans just had really bad candidates is probably overplayed. I think the national environment had a much bigger impact

#4: It was very challenging to find a winning message for Republicans these Midterms, but I really think they leaned way too much into the transgender issue and repeated the they/them ad to their detriment.

Democrats were on message with their ads focusing on the issues that voters demonstrated they prioritized (e.g., cost of living, corruption)

I would be interested to see what ads Republican candidates who over performed ran (e.g., Pete Ricketts in NE).

#5: Smart people who were looking at the early vote indicated that it was looking really good for Democrats (especially in states like NC).

Knowing that Democrats were “banking their vote” in Southern states and Republicans were not made me more confident in my “blue tsunami” hypothesis. I did not buy that Republicans would have record turnout on Election Day to offset the Democratic early vote advantage. That didn’t even happen in 2024.

Anyway, would love to get other people’s thoughts. I don’t think 2028 or future election cycles will be as easy to predict as 2026 was.

Weighting to 2024 was egregious.

This is why I was so confident Talarico was going to win Texas. Polls were absolutely herding around a Talarico +1-4 sample, with a Trump +10-15 2024 recall.

It was absolutely ridiculous to think that the Texas electorate was going to be double digits for ‘24 Trump in 2026.

I’m not blaming all pollsters, but we live in a world where the incentives to release a Talarico +2 poll are just stronger than releasing a Talarico +12 poll for a variety of reasons.

Credit to the pollsters who didn’t herd. Good post by @ChazNuttycombe

Roy Cooper’s attack ads on Michael Whatley.

Not only did Cooper win a Trump+4 state by 20%, but him leading the ticket also led Democrats to flip many Republican House seats that were seen as un-winnable

BREAKING: 2026 Midterms trading volume has surpassed $3 billion on Kalshi