the true meaning of election night

A preview of what Election Night will look like, and my relationship with politics, interspersed. By Jacob Cohen (@Conflux).

On Election Night 1948, as he began to take an unexpected lead in the election results, President Harry Truman was asleep. He’d gone home, eaten a sandwich, drunk a glass of milk, and gone to bed early. 

when i was a little kid, i didn’t understand election watch parties. wouldn’t it be easier to just wait until all the votes were counted, and react to everything then? what was all the hurry and fake drama?

In this newsletter, I’m going to preview what Election Night (and specifically, Manifold’s coverage) will look like. However, I feel honor-bound to state that, if you find election-watching stressful, you can go to sleep as early as you want! I won’t judge you. But I will do my best¹ to write you another newsletter, summarizing what we know, and it’ll be in your inbox in the morning. And I will be jealous of you, because you’ll be more well-rested than me. So please: take care of yourself. Don’t watch the election ... unless you find it fun.

“Election Day is my Super Bowl,” i wrote in a high school scholarship application. i had changed my mind since being a little kid — and had so much fun watching the 2018 midterm returns come in. “As I pepper my politics Google Hangout with news flashes, notice FiveThirtyEight’s models fluctuate, finish a large bowl of guacamole, and watch John King examine the map on CNN, I can’t help but feel the thrill of the moment. However, whatever happens tonight will have a major impact on the country, and even the world, for years to come.” 

There’s a 32% chance the election will be officially projected tonight (by 1am PT = 4am ET, the end of the Manifold Election Day Watch Party) and a 61% chance by end of tomorrow. It may come down to polling error: a normal-sized 3-point correlated error could cause a projection on Election Night, but accurate polls, or polling error running in opposite directions, would mean a razor-thin race. (The Senate situation is similar; the House we’re quite unlikely to know, since many critical races are in California, my lovely home state but one which counts ballots quite slowly.) If you’re looking for a state-by-state overview, 538 has a great one — as does MSNBC’s fashion icon Steve Kornacki. But the gist is that meaningful results will start coming in around 4pm Pacific Time, and diminishing returns will probably set in around 10pm. Here’s a graphic from Jamelle Bouie (linked on Manifold by The Official):

on the same application, there was a spot for “dream job.” i’m sure some people put astronaut, or scientist, or president. me: “political analyst (Nate Silver-style).” i can’t make this up. my middle school dream job was to be a POLITICAL ANALYST.

As part of my current job as CHIEF POLITICAL ANALYST FOR MANIFOLD MARKETS², I will be xeeting up a storm at twitter.com/ManifoldMarkets, including making official projections on behalf of the all-new Manifold Markets Decision Desk. That’s right, in addition to an election livestream at manifold.markets/tv, we will be issuing an official projection once a candidate gets a 98% chance of winning that state, according to our market probabilities, superforecasters’ analysis, and data. Because our markets react so quickly and accurately to news, we’re expecting to be faster than traditional networks without getting a call wrong! Well, you can bet on it: 

 

2020 was a crazy time. but i got lucky. i didn’t really like my school, or have many friends, so i hadn’t lost much of a social life. i volunteered for a city council campaign, which was energizing. and i was transferring to a new school (the manifold-famous proof school) which felt like such an upgrade for my math-nerd brain even though it was online. i loved the math, and i loved the non-math too. an incredibly thought-provoking literature class about susan gubar’s “antimorphosis” and the ethics of making light of tragedies like the Holocaust. a fascinating world history course using “modernization theory” to analyze the rise of democracy across brazil, nigeria, china, and the usa. some wonderful new friends to talk to online.

Make new friends and come with your current ones to the Manifold Election Day Watch Party in Berkeley! I’ve tried to make other names happen (like Manifold’s Political Party and Electionpalooza), but a plurality of you who filled out the form wanted to stick with Manifold Election Day Watch Party, and it seems like that’s what people are calling it. Based on form data (I still think my election-themed social game and costume contest would be fun, but I see now that this is not the venue) we will have the following exciting lineup:

• 2:30 pm - 1 am: Election returns watch party. Streaming all day from multiple locations across the venue!

• 3 pm: Election Past-casting: Jacob Cohen (Conflux) (hey that’s me!), Manifold's newsletter writer and lifelong election aficionado, will regale audiences with the tale of a randomly selected election in US history — with names and details anonymized. This will allow the audience to hone their election forecasting skills and bet on the twists and turns of this election!

• 3:45 pm: Election trivia / estimathon: A casual competition on your knowledge of current events, political history, and/or Manifold probabilities! The winner will get a prize.³

• 4:30 pm: Prediction Markets and the Dilemma of Donald Trump with Pratik Chougule: Explore tension of how Trump is the friendliest president and major party candidate in modern history toward prediction markets, but also his embrace of a style that is anathema to the values of the forecasting community.

• 5 pm - 10 pm: Key election returns

PLUS

• Physical election needles with live-updating odds! (main building)

• Trading Floor (Bayes Ground): a place for dedicated election bettors to chat

• Politics-free discussion zone (Bayes Attic)

• Lightning talks at 9 pm (Gyroscope): Give a 2-minute rapid-fire talk on any topic of your choice! ⚡ (Stephen thinks this will be unpopular even though it was the top-scoring event on the form, so you should all come in order to prove him wrong! Also, it will be emceed by the inimitable Ricki Heicklen.)

Or if you live in the Boston area, you could instead go to Andrew’s Anti-Watch Party, and join the trend — which includes TBTBTB’s iconic market — of trying to escape the politisphere and not learn who won.

january 6, 2021. at first i was happy: democrats seemed to be pulling off upset victories in the georgia runoffs! but as i saw the other news, that of the capitol insurrection, my joy turned to terror. i feared that congresspeople would get killed, or that biden’s rightful win would somehow be overturned. my overriding feeling was that This Wasn’t Supposed To Happen. maybe it would happen in nigeria, or another developing country i was learning about in my world history class, but not in the united states of america. we discussed the next day in class how the usa’s “polity score” had dropped to 5, meaning it was no longer a democracy and instead an “anocracy.” i was scared of what could happen. suddenly i didn’t want to follow politics anymore, i didn’t want to be consumed with stress or blog about something i couldn’t control.⁴ so i wrote about “grapes; for the worms shall eat them.” i went deep into an uncontroversial nerd hobby: constructing puzzles.

The 2024 election — let me be real — quickly devolved into a boring, depressing affair. After Ron DeSantis’s campaign fell flat (remember when he was actually the favorite to be the nominee?), the primaries were for show. It became a foreordained rematch between two men with record-breaking ages and underwater approval ratings. Somehow Trump had rehabilitated his image again, and the polls were razor-thin and absolutely impossible to move. Everyone knew which man they thought was the lesser evil.

Source: AP.

i joined manifold when it first launched. i liked to think of play-money betting like editing wikipedia: using my expertise to make a communal information source more accurate. soon i ended up spreading the website to my proof school friends, and simultaneously became a well-known community figure (“trustworthy-ish,” whales v. minnows resolution council, ...) It was only a matter of time before i was betting big on my old hobby: elections. with my old forecasting knowledge, the strategy of simply copying fivethirtyeight, and a healthy dose of luck, i placed third in the 2022 midterm forecasting tournament. i was very proud of winning ninety-three dollars despite spending nothing.

PredictIt faced legal troubles (rumored to have been linked to rival Kalshi, though Manifold thinks it’s unlikely to be confirmed). FiveThirtyEight had major layoffs, and lost Nate Silver. In 2023 it seemed like Manifold would be the savior of 2024 election coverage, shining among the fallen.

as i got more involved with manifold, they offered me a position writing newsletters on politics. i was grateful but didn’t want to commit. there was no news! the 2024 election was so boring! i just didn’t want people stressing about meaningless minutiae. i didn’t want to have to follow the meaningless minutiae closely enough to write the newsletter! but they let me try out a sample newsletter, on some random political fantasy that ezra klein was talking about: the idea that joe biden would somehow magically be replaced as nominee. it was straight out of the west wing. but i said maybe there was a chance.

Then there was the debate. I swear, when I was watching it, I thought it was normal. Just “Deja Vu But Worse,” as Weird Al described it in a largely pre-written songification. (Technically I was watching the Sarah Cooper live-reaction, because I thought it would be unbearable to watch the debate by itself. I think my underreaction to the debate itself, by the way, was because I’d seen clips of 2024 Biden, thought the country had priced in his inability to speak fluently, and was still struck by Trump’s total disregard for truth.) But CNN’s tone afterward was like nothing I’d ever seen. Suddenly, after a historic few weeks (which included an assassination attempt against Trump), the unprecedented was happening. Joe Biden was being replaced — after he already had enough delegates to be the presumptive nominee — with Kamala Harris. This didn’t just fall out of a coconut tree. It was brat summer. Politics was fun.

i wanted my newsletter to be fun, but i didn’t want to express controversial opinions. i still don’t want to. in my blue bubble, i know there’s pretty much nothing controversial about me saying “I endorse Kamala Harris for President of the United States.” but I do. though really, the closest thing to my views right now is this acx endorsement of harris in a swing state and “anyone but trump” in a safe state. 

Somehow it’s also a renaissance for prediction markets. Polymarket, Kalshi, Manifold sweepstakes: betting on the election is bigger (and more mainstream) than ever. And although FiveThirtyEight is a shadow of its former self, Nate Silver has created its spiritual successor, Silver Bulletin.⁵

i feel guilty. i wish i knew more about local races, wish i could tell you what to do about the propositions. as matt glassman likes to say, real politics is not about following the political soap operas in the news: it’s about the pta, and local government, and the things you can help make better by showing up — but the part of politics where no one wants to. in the next four years, whoever wins, i promise to have more courage to say hard things, and to get my hands dirty. hold me accountable.

Manifold sends this newsletter to anyone who has ever given us their email and has not unsubscribed. There are over 90,000 recipients. Even though I’m sure most of you don’t click through, this is still surely the biggest audience any of my writing has ever had.⁶

i wrote this newsletter in lighthaven, a magical place, in a three-hour flow state. sometimes i hate writing, but this time, at least, my writer’s block storm abated. my stress about the election results has sort of been replaced by the stress about all the things i have to do, and i feel productive, so this is perversely good for my stress level. humans are weird. even though i’m tired, i’m excited for the pageantry of this strange holiday. and, even if it’s not my long-term plan, i’m thrilled to be living my middle-school dream of being a political analyst!!

Whether you’re voting for Harris or Trump or RFK or not at all, whether you’re coming to the Manifold Election Day Watch Party, whether you’ll be watching manifold.markets/tv or reading our live-tweets or increasing our accuracy by betting on our markets or all or none, whether you are planning to pull a Harry Truman and go to bed early — from the bottom of my heart, on this day that has meant so much to me: may you find solace, may you find joy, and may you feel like the world will still be okay. <3

The latest xkcd comic.

i’ve been jacob cohen (aka conflux), a student at stanford university. i blog at tinyurl.com/confluxblog, make puzzles at puzzlesforprogress.net, and don’t release any new episodes of the market manipulation podcast on your favorite podcast platform. check out manifold politics at manifold.markets/election, on twitter/x @manifoldmarkets, and at manifold.markets/tv. and if you enjoyed this newsletter, don’t forget to like, share, and/or subscribe to above the fold on substack! 

PS: The enter key on my computer is broken due to an angry wave (and unfortunate accident) at the beach. I’m the type of person who finds it an interesting puzzle, and I have this website with a newline bookmarked for copy-pasting, but if any of my tweets seem to have fewer newlines than you’d expect, this is why.

recommended supplementary reading: astral codex ten, “Mantic Monday: Judgment Day”

1. I might be totally wrecked, in terms of stamina. If there’s no newsletter, I’ll at least try to have a long tweet at the top of twitter.com/ManifoldMarkets.

2. Strictly speaking, this is not my official title (yet). Stephen can I please have it?

3. I don’t know yet what the prize will be, but there will be one. (Let me know if you have ideas!)

4. This story makes it sound as if the Capitol insurrection was my only motivation. It definitely was one, but two other significant ones were my obsession with the problematic nature of the musical Bloody Bloody Andrew Jackson and also a deepening fear of saying things online which could potentially offend people.

5. Someone who wanted to be identified as “Not Allan Lichtman” wants me to mention the many other prediction markets that have lived and died.

6. I have written Upon Reflection puzzles which were published in The New York Times Magazine, but that doesn’t really count. Also, I don’t think very many people do those puzzles; I’ve never met anyone who recognized those puzzles without knowing who I was.