The Market Manipulation Podcast
Five Episodes, Five Stories at the Conflux of Serious and Meme
by Jacob “Conflux” Cohen
Have you ever longed for a podcast discussing new implementations, betting strategies, and events surrounding Manifold? Craved lighthearted yet serious commentary, storytelling, and trivia gameplay? Yearned for the dulcet tones of my voice and highly-reviewed banter with my long-term co-conspirator and inspiration Andrew Imbens, plus guests including Manifold’s cofounders? Bawled for a Manifold podcast to crystallize at last? Entered this newsletter with no expectations and are currently assessing whether you want to keep reading?
Regardless, the five episodes of The Market Manipulation Podcast are now available on YouTube, Apple Podcasts, Spotify, Overcast, Google Podcasts, Amazon Music, and RSS! Search for "market manipulation" or "market manipulation podcast" (under Podcasts & Shows if applicable), and we should show up as a top result.
The podcast will satisfy all the above desires, and is themed around Manifold, the website that has become my full-time obsession. Episodes release irregularly, and the podcast’s future is unclear, but that makes it all the more low-commitment to binge all five episodes! (Although beware that Episode 1 received an award from Isaac King for the worst sound quality he’d ever heard on any podcast, and will test the constitution of your eardrums — you may wish to opt for reading the transcript instead, or start with one of the other four episodes. You don’t have to listen in order.)
The podcast is geared toward the Manifold community, but anyone can listen! If you’re reading the Manifold newsletter, I’m assuming you’re familiar with things, but otherwise, some background on Manifold is in this footnote.1
So Why Is It Called "The Market Manipulation Podcast”?
Well...funny you should ask...
On Pat Scott’s recent Manifold survey (which I recommend), there’s a question about whether Manifold is a serious website. I had a hard time answering it, because…I’ve been on Manifold since the early days, and it’s always been right at the border of serious and meme: the conflux, if you will.
I made a market which inspired in-depth mathematical investigation…about whether 69 was the only “nice number”. I served on the resolution council to ensure a fair ruling of one of Manifold’s most highly traded and controversial markets…Whales v. Minnows, a self-resolving market non-predictive of real-world events (except the final digits of Bitcoin hashes). My biggest profit so far was on a Senate election market, part of the 2022 midterms prediction tournament…while my biggest loss was on whether Levi was HMYS, “teaching me market dynamics.”
So yes, it’s true, The Market Manipulation Podcast started as a joke “podcast” created to manipulate a Manifold market. But on Manifold, jokes have a way of becoming real.
I will tell five stories — five stories for five episodes of Manifold’s premier podcast — stories, like the podcast, which live in the conflux of serious and meme.
The Tale of Episode One: I Make a Podcast Exclusively to Manipulate a Manifold Market
The date is Tuesday, February 28, 2023. It's about 9:15pm Pacific Time. Since the end of the month is a pretty natural time to end your market, many of them are closing soon, and I'm scrolling through, looking for quick mana opportunities. Suddenly, I see one...
Its title: "Will anyone create a podcast centered around discussions on Manifold Markets?"
Its creator: wannabe Proofnik and self-described "best market creator" Wobbles, fresh off a rebrand from "DesTiny." (Many fans of the streamer Destiny use Manifold, since there was previously a meme about creating "stock" markets for different "orbiters", or guests, but Destiny himself doesn't have an account — but many people have thought DesTiny was him.)
Its current displayed percentage: a mere 4%.
The market has hovered around 70% for most of its life, as longtime whale BTE and Wobbles discussed starting a podcast together, but it’s fallen as no podcast emerged. And 4% makes sense, considering that the market provides a fast-approaching deadline: end-of-day on February 28, in Central Time. Which is in 45 minutes.
But with a twinkle in my eye, I carefully look at the resolution criteria. "Resolves YES if a podcast is created for the purpose of discussing new implementations, betting strategies, or events surrounding Manifold Markets. Must have atleast one episode uploaded for public viewing that is 10mins or longer by Feb 28th (other requirements might be added)."
New implementations, betting strategies, or events surrounding Manifold? I can discuss all those topics. Uploaded for public viewing? I'll put it on YouTube. One episode that's 10 minutes or longer? I have 45 minutes: that's almost five times as many as 10. I buy 100 mana worth of YES, moving the probability from 4% to 33%.
I improvise the whole thing, recording with my computer microphone, providing live commentary as I continue to scroll through closing-soon markets. Once I clear the 10-minute mark, I wrap up the podcast and start listening back, to make sure nothing was too bad. The sound quality is whiplash-inducing, but it’ll have to do. Investing M$300 more, I move the probability to 78%.
To upload to YouTube, I need to convert my sound file into a video file. Unfortunately, I only have the iMovie app on my phone, not my computer. So I frantically AirDrop the file. Meanwhile, I channel all my graphic design passion into designing a beautiful logo in seven minutes, whose graphic design motifs include the meme stonk guy and a tagline heavily influenced by the market's resolution criteria. This logo will go on to…still be used.
With time running out, I take a picture of the logo on my computer with my phone (I know, it's horrifying), and put it into iMovie over the sound file. I export it back to my computer, and start uploading it to YouTube at like 9:50, throwing together a brief video description.
But by 9:57, I'm panicking. YouTube provides a link to where the video would be, but says it’s still uploading. I can link that early and pretend it was already available, but it feels unethical to do this if the video is inaccessible. Maybe 45 minutes wasn't enough time after all, at least with my tenuous plan. I sell some of my YES shares, dropping the market to 58%.
In my adrenaline, though, I think of a backup plan. I upload the sound file to Google Drive, share it with "anyone with the link", and copy that to the comments. Tackier than YouTube for a podcast, but time is of the essence.
As the market is about to close, the Google Drive file finally uploads. As I'm about to link it in the comments, I realize, dismayed, that the market has closed. But wait: it's only 9:59:20! The market's close time is set to 9:59:00 PST, but the end of February 28th in Central Time is 10:00:00 PST. I’ve made it in time. As it happens, the video finishes uploading to YouTube shortly after, and I'm able to link it at 10:00:37.
Pros of listening: It's only ten minutes, and even less if you speed it up. Also, the content isn't actually that bad considering it's 100% improv. And it's a podcast with me on it!
Cons of listening: bingeworthy wrote "for the next podcast you record please rub the microphone against sandpaper more often. i think that was something that this one lacked.", while Isaac King wrote "I would like to issue you an award for worst sound quality I have ever heard in a podcast, or any online audio recording really." I'm serious about the sound quality being awful. Listen at your own risk.
Anyway, I turned a profit of M$764 for this podcast (and another M$73 on Wobbles' other market on whether a podcast would be made by end of March) — but what's priceless is the everlasting contribution to Manifold history and culture.
The Tale of Episode Two: The Silver Tongue of My Long-Term Co-Conspirator Is Used
So that was all for the memes...right?
I'm not actually planning to spin this into a real podcast. But it's funny how things happen.
Wobbles makes another market on whether a podcast "that is somewhat high quality" would be released in March. And I can feel the allure. So I make a market on whether I'd release a second episode in March.
And while I'm not the most qualified person to start a Manifold podcast at this point, I'm not the least qualified either. As previously mentioned, I've been an active user since the beginning, I've come in third in Manifold’s 2022 midterm election forecasting tournament, I’ve spearheaded an investigation into nice numbers like 69.
Relevantly, I've also somehow convinced tens of people at Proof School, the wonderful full-curriculum math-focused school which I attend, to join the platform, partially due to a market on the love life of person-who-enthusiastically-agreed-to-this-and-whose-girlfriend-at-the-time-did-also Andrew Imbens.
However, BTE and Wobbles' other podcast idea is starting to take fruition, and they’ve written up a detailed grant proposal for their podcast, now with a name, Crystal Ballin', on Manifund (Manifold's charitable granting arm). I have no desire to compete with another podcast, which I'm frankly excited to listen to, and maybe even be a guest on at some point.
Really, it’s Andrew’s fault. We’ve been friends for a long time (and even co-created a school newspaper some years ago). While he isn’t as active on Manifold as me, he’s always been interested in starting a podcast, and plus that way he can represent the average user who didn't know that much. He has some sound equipment and expertise, and volunteers to edit the podcast. He suggests we just try one serious episode, and see how we feel afterward.
I give in to the call of the podcast.
We take it seriously, brainstorming topics, segments, and theme music ideas. Ultimately, we settle on covering the dramatic saga of Manifold's rise and popularity at Proof, though we also incorporate a Manifold News segment covering calibration plots and shirts. We also work to provide context on all the Manifold things we mention, though I ruin an anecdote midway through by not saying that I mentioned to Audrey that I'd made 200 fake dollars recommending the 3B1B soundtrack, inspiring her to join Manifold.
It's a legit podcast episode, if I do say so myself! Thanks to the upgraded sound quality and Andrew’s great editing, it is widely considered "listenable,” although, due to uploading the wrong version of the sound file at the end, you may have to turn the volume all the way up on YouTube. Alternatively, simply listen on the other podcast platforms, like Apple Podcasts, Spotify, Overcast, Google Podcasts, Amazon Music, and RSS.
(We changed the tagline to reflect Wobbles' new market, which now resolved YES. Although we were inconsistent in replacing "market close" with "Feb 28" last time but not replacing it with "Mar 31" this time. This actually really bothers me. Also, we kind of abandoned the idea of theme music later, oops.)
While we’re editing, a new piece of Proofnik Manifold drama occurs, too late to mention. In a bid to "become a Proofnik", Wobbles emails Proof School's admissions director, and then spends real money shipping samosas and Almond Joys (favorite foods of the dean of humanities and head of school, as informed by Proof graduate, long sentence writer, and person-known-to-wake-up-at-4am-to-ride-public-transit-to-the-middle-of-nowhere-and-back-on-purpose-for-fun stq) to the school building. (I originally wrote 6am, but when I asked for his permission to be described this way, he told me that “6 am isn’t a particularly early start for me.”) The samosas, unwisely shipped to our downtown San Francisco campus at 2pm on a Sunday, get stolen, but the Almond Joys arrive, and are displayed during our morning meeting by the head of school — who says they were from an "applicant." It’s very unfortunate that we don’t cover this in our podcast (but we go on to mention it later in Episode 5).
Albeit without this incredible anecdote, the podcast is done! It was fun to make it, but also, shock surprise, more work than I expected, and not quite as fun. We’ve received a few nice pieces of feedback, but not enough to justify an Episode 3 in our minds. Plus, Crystal Ballin' is on the horizon, complete with an official Manifold account and tons of markets about their first few episodes. Manifold veteran Isaac King also joins their project.
So, all things considered, we're not really planning to do a third episode.
The Tale of Episode Three: We Get Funded with Mana
But The Market Manipulation Podcast is dedicated to its ideals, and those ideals involve excuses to get free fake money.
So at the end of Episode 2, we did a segment about how we would guarantee a third episode if we were "funded" with 6,000 mana. This was a blatant grab for mana, and also mostly a joke. I'd made a market about whether we'd achieve the "funding goal" within a week, and it resolved NO — our only investment offer was a M$100 "sponsorship" from nice-numerologist Joshua Brower. I'd also made a market about whether we'd achieve it in two weeks.
Much like Wobbles' original market about a Manifold podcast, it reaches a low watermark of 4%...but it resolves YES when Manifold cofounder Austin Chen sends us a manalink for all M$6k we had requested just in the nick of time. (Of course it’s a Manifold employee who can afford to give us that much, at that time...)
(By the way, though we joked on the podcast about splitting the mana unevenly, I do give half of it to my cohost Andrew. And while Joshua's sponsorship would've been fun, sadly we pledged to make the podcast ad-free if we received funding...)
And let it not be said that The Market Manipulation Podcast flakes on its promises!
Plus, Manifold had drama that needed discussing, and Crystal Ballin' was yet to release their first episode.
Self-resolving markets (non-predictive profits, whalebait, meta markets, they go by many names) have always had a presence on Manifold. My first ever market, back in February 2022, was "Will at least ten users trade on this market within the next week?" Personally, I think they're kind of fun, and a cool way to learn some market dynamics (yes, there are others), though they get stale if you repeat the same concepts over and over. Also, as I like to say, bet on gambles and your profit's in shambles.
Anyway, I make one that would become known as simply "The Market": open for a week, and resolves YES if its average percentage was above 50%, otherwise NO. It goes "viral" on Manifold, with 458 traders, M$2 million volume, the then-biggest spike in Manifold revenue (you can buy mana with real money...), and a front-page feature during April Fool's Day. (Ironically, my wealth gained from unique bettor bonuses as well as day-trading on The Market, among other things, makes M$6,000 feel much less significant to me.)
In addition to touching on self-resolving markets (a topic Andrew finds boring, a boredom we discussed), we also covered my mana grant program, dating markets, personal goal markets, stocks, making bots using the Manifold API, and various choices made by friend of the podcast and alleged chatbot Audrey, who also cameoed as a call-in guest! We prepared a bit less, and recorded later in the day, decisions which caused us to have more fun, and made the episode better as well, I think.
But the thing about this episode was, at the time we recorded it, Andrew was super busy (as was I). So though we found time to record on April 19, editing the episode got delayed and delayed, and it took us until June 6, forty-eight days later, to send it out.
And time did wonderful things to Episode 3. We recorded it when The Market was still the most prominent self-resolving market, but a new one, Whales vs. Minnows, would soon overtake it. For a rundown, watch bingeworthy's excellent YouTube video explaining everything, or firstuserhere's shorter, memier but still accurate video — but here's the main story, partially plagiarized from my own writeup on the market:
The Story of Whales vs. Minnows
Isaac King created Whales vs. Minnows as a fun little market about whether Team YES (the whales) would amass at least 10,000 times as many shares as the number of bettors for Team NO (the minnows). 10,000 is an enormous number, but Isaac surfaced as the main whale, so bettors suspected he had a trick up his sleeve. Furthermore, there emerged concerns about highly ambiguous resolution criteria — initially the market would wait to resolve until no one was betting, except if "after a week it looked like it wouldn't be closing any time soon", at which point it would close once a Bitcoin block was mined with hash ending in 00. A week after what? It also was unclear how exactly alt accounts would be counted and removed from counting for the Minnows. Therefore, Isaac delegated resolution to a resolution council of myself, Martin Randall, and Jack, who sold our shares to remove conflicts of interest. I recorded Episode 3 of the podcast with Andrew the day I was put on the resolution council.
In any case, the Minnows had taken a healthy lead — each NO bettor required an investment of M$10,000 on YES, after all — and the user tornado embarked on a plan to spend real money to hire people from developing countries to make accounts and bet NO. However, Isaac had also been hiring people, and he had them give him his API keys, allowing him to bet for them using a computer program. He initially had them bet NO, to make it seem like the Minnows' lead was safer, but once the random resolution process began, he had them all sell and bought enough YES to put Team Whale into the lead. He spent USD$29,000 on mana for this market, perhaps thinking of it as an elaborate way to make a charity donation: after all, he couldn't lose...right? The market hit 100.0%. We on the resolution council lamented that we couldn't buy NO, but debated whether we'd sell at 96 or 99.
However, Isaac (and we) underestimated the resilience of Team Minnow's recruiting efforts; in particular, recruiting from the Destiny community, whose conceptually flawed early "stock" markets were often won by Isaac. In any case, he then realized that spending additional thousands of dollars to gamble on Manifold was unwise. Manifold graciously decided to refund Isaac $25k of his purchase. It went on to take various steps to counteract self-resolving markets, such as the creation of a highly subsidized Showcase group, and the ignoring of "non-predictive profits" in their tabulations for Manifold Leagues (where they put you in a group of users and you can promote, demote, and win prizes based on your profits and bonuses during a month — I think they're a bit silly and gamify Manifold, but many people like them).
I do wanna say for the record that although Isaac pushed the boundaries of Manifold, he contributed so many creative and important markets, and apart from occasional exceptions (like WvM), he provided clear resolution criteria and followed them to the letter. He's also a Magic: The Gathering judge who runs a cool blog called Outside the Asylum that the MTG rules nerd in me loves, especially learning about layers and gaps in the game's allegedly-comprehensive rules.
As for the Podcast...
Suffice it to say that criticism of self-resolving markets aged well. Additionally, Isaac King left Manifold after losing so much on Whales vs. Minnows. A loss to Manifold, but also to Crystal Ballin', which he'd previously committed to help with.
Which still hadn't released an episode.
But anyway, The Market Manipulation Podcast had done all it promised! We were now savvy enough to set the funding goal for a fourth episode to the much more ridiculous M$100,000. No one gave us that.
So we were absolved. Why on earth would we release Episode 4?
The Tale of Episode Four: The Stars Align and We Record a Podcast at a Firepit in Ocean Beach That We're Reserving for Austin and Rachel's Wedding Bonfire
Since Proof School and Manifold HQ are both in San Francisco, Manifold is sort of a local small business. And I organize an event where Austin and other Manifold employees come to visit Proof and do a Q&A. I end up being out with covid, but the event happens nevertheless, and from what I hear, people have a great time. There was a seventh-price sealed-bid T-shirt auction.
After a whirlwind romance told in prediction markets, Manifold cofounder Austin Chen and Rachel Weinberg are getting married on Friday, June 9, and planning to have a bonfire at Ocean Beach. Austin asks me if I or other Proofniks, no longer total strangers, were interested in helping by guarding/squatting on a firepit to ensure that it's available during the evening. He says that we’d be invited to stay for the bonfire, and they'll provide dinner and maybe s'mores.
The last day of the Proof School 2022-23 school year is going to be Friday, June 9. We're going on a field trip to the San Francisco Zoo, and then hanging out at Ocean Beach. It's the perfect excuse for an event.
The event’s official title becomes the Joshua Brower Memorial Nonfire (in honor of former potential podcast sponsor Joshua Brower, who wanted to go but couldn't). For legal reasons, but I am compelled to mention that the genius name "nonfire" was the timeless epiphany of potential math major who was right once Allison. (Favorite guest Sophia, of sophialiao.com math blog fame, tells me that "nonfire" is a bad name, but, dear reader, as soon as I speak now that Allison was the mastermind behind it, that’s when she decides to change to saying it’s a good name. Long story short, this hoax corroborates other established evidence that she's Chaotic Evil.) (Sophia authorized the publication of this anecdote.)
In any case, I'm talking to Allison about ideas of ways to occupy ourselves for the four hours at the beach, and pitch "market manipulation podcast episode 4, live from the nonfire?!?!?" I mean it unseriously, but she's enthusiastic about the idea, suggesting we run a joke style episode, since it’d be less work. I'm skeptical that it would really be less work (or that it would be entertaining).
However, I become enchanted by the idea to make a Manifold trivia game, and have everyone play it live on the podcast at the beach! It'll be fun, I love when other podcasts play games, like Guess What Americans Think on The FiveThirtyEight Politics Podcast, or Wait Wait...Don't Tell Me as long as you forget that it was supposed to be a game with scoring.
This marks a pivotal turning point in enthusiasm for the podcast. Previously, it was external factors (Wobbles' market, Andrew's lobbying, Austin's funding) spearheading the push to make episodes. But now it’s ME! I'm the one who wants to do a podcast and run a Manifold trivia game during the nonfire because that’s the fun of me.
Indeed, during the nonfire, momentum shifts away from the idea of recording a podcast. Over halfway through, I go to the nearby Safeway to go to the bathroom and acquire some much-needed vegetables (in addition to a platter, I purchase a whole tomato and avocado for myself, because that's how I roll), and return to a group happily playing a party game, looking impossible to tear away. But somehow I pull it off.
The sound equipment Andrew's brought is missing an SD card, so we have to record the podcast as a voice memo on AP Bio 5 scorer Amelia's iPad. The sound quality is what you expect from a podcast recorded at a beach whose less-Manifold-podcast-interested guests are coming and going. I modify the rules and scoring of the trivia game, which has not been playtested, on the fly. Nevertheless, the episode might be my favorite!
You can play along with the trivia game (note that the first section contains Proof School-specific questions), or simply enjoy this unique time capsule of Manifold and nonfirestalgia.
It's after the release of this episode that I'm inspired to get The Market Manipulation Podcast on proper podcast platforms.
It's all through RSS.com, which makes it easy to add other places to listen — let me know if you prefer listening to podcasts somewhere else, and I can probably add The Market Manipulation Podcast there! I found RSS.com while searching for a free site that enables creation of RSS feeds for podcasts. It has a nice interface, and I was fully satisfied with it right until I tried uploading Episode 2, when I discovered that a certain feature required a premium subscription to RSS.com.
That feature? Having multiple episodes.
That's right, RSS.com is a totally free podcast creation platform, as long as your podcast only has one episode. But your second episode — now that's a bit much, don't you think? (Maybe I’m being too harsh. How’s a freemium website gonna get subscriptions?) Fortunately, I was able to get a free trial; also, by that point I was satisfied enough with their interface that I probably would've paid anyway.
The Tale of Episode Five: It's Summer and I'm Writing This Blog Post
Two of my summer plans have fallen through after Manifold markets gave them an >80% chance: qualifying for the International Linguistics Olympiad in Bulgaria and attending the Program on AI and Rationality in England. It's a near miss in both cases, but I can't help but feel that the markets were a bit too confident in me.
I'm in a productivity funk. I have applications to write and projects to work on and not enough of it is getting done. I fail two markets, losing thousands of mana, about finishing a draft of this very blog post; and as I first write these words, I'm a mere thirteen minutes away from my self-imposed deadline for the third (which I end up making, with seconds to spare); but the failed markets, and the few deadlines that are real right now, feel like the only things keeping me on track.
But sometimes being unable to write something is a sign that it's not the best version of itself, that you have to let it percolate for the right idea to come. This was originally just gonna be a post on my personal blog linking to the Market Manipulation Podcast. Then it was gonna mention the podcast, but just also talk about Manifold for my readers. Then it was gonna be an FAQ that worked all the information about Manifold into the questions and answers. Only drafting it now did I hit upon the idea of structuring it as five tales. And only later would I have the idea to even ask to host this on the Manifold newsletter. (After all, there’s been podcasting, and there’s been manipulating, but there hasn’t yet been all that much marketing!)
Anyway, I felt like the blog post needed a crown jewel. And wasn't five episodes a nicer number than four? Wouldn't it be a great way for the podcast, initially a joke, to gain even more legitimacy? Shouldn't I do something to get out of the house, didn't I live near the Manifold office, and hadn't the Manifold employees been super kind and super generous with their time? It was worth asking, at least.
They said yes! So Andrew and I make the journey to Manifold's San Francisco office, metaphorically open the door with my now-septuply-signed Manifold Key, and sit down (separately) with cofounders Austin Chen and Steve Grugett, as well as developer Sinclair Chen and cofounder James Grugett. We edited their words together to form the fifth episode.
We discuss a wide range of topics, many suggested via bounty market (whose bounty I apportioned based on airtime in the episode): the rebrand of markets as questions (and whether we need to rebrand as the Question Manipulation Podcast), Manifold’s name backstory, the upcoming Manifest conference, Austin’s relationship markets, dogfooding, Manifund, Stephen's prediction market inspiration, bounty markets themselves, Mana-chan, banning philosophy (as applied to Levi Finkelstein), the real reason Manifold employees sometimes write bad resolution criteria, Wobbles' shenanigans, Sinclair's hot takes, catgirls among us, and James' theory about what Manifold will look like in a million years.
During the episode, I also ask the Manifold guests to predict the chance that Crystal Ballin’ would release an episode before September. They all guessed probabilities in the teens. The current market percentage is…well, it’s been fluctuating, but somewhere in the seventies or eighties. I hold almost 20,000 NO shares, while BTE has nearly 25,000 YES. BTE is the cohost of the podcast, so this is both worrying for my Manifold portfolio and exciting for the Manifold podcasting world.
(To upload Episode 5, it’s a new month, so I do have to pay RSS.com — which is fair, they’ve been a good website. But when I go to reactivate my RSS.com subscription to post the episode of my podcast, I find that in order to get my subscription back, I have to contact customer service — since even though they allow a monthly plan, they’re not a big fan of cancelling and uncancelling subscriptions for each month in which you release episodes. However, customer service responds promptly, and I pay them the monthly fee.)
Making this podcast is fun, but it’s been a lot of work, and Andrew and I are gonna be busier heading into the school year. Perhaps Episode 5 will be the last episode, and then we’ll all get to listen to Crystal Ballin’ as it assumes the position of world’s premier podcast created for the purpose of discussing new implementations, betting strategies, or events surrounding Manifold. But, as always, we’ll see what actually materializes…no promises.
In the meantime, go listen to The Market Manipulation Podcast on YouTube, Apple Podcasts, Spotify, Overcast, Google Podcasts, Amazon Music, or RSS. I hope you enjoy!
Thanks for reading to the end. I’ve been Conflux, and I blog at tinyurl.com/confluxblog and make puzzles at puzzlesforprogress.net. I’m a huge fan of Manifold and prediction markets, and if you ever want to chat for any reason, you can email me at chromaticconflux@gmail.com or DM me on Discord @chromaticconflux. Also, I don’t want to spoil anything, but if you subscribe to my YouTube channel, there will be a fun mathy video dropping within the next few days!
Manifold (recently rebranded from "Manifold Markets") is like Wikipedia for the future: it’s a website and app where anyone can create a "market" (or "question") about a future event — from “Who will win the 2024 US Presidential Election?” to “Will I have a date to prom?” — and see crowdsourced probabilities of the options.
How are the probabilities crowdsourced? With bets of mana (which I will write as M$), Manifold’s currency! (You can buy mana with US dollars at a rate of 1 cent = 1 mana, but you can’t convert it back except through matched charitable donations.) The non-mathematical explanation for how it works is that each bet changes the probability, and it all works out so it’s optimal for each market to reach its true probability!
(The mathematical explanation: to make a YES/NO market, you pay 50 mana, which becomes 50 YES shares and 50 NO shares. If the market’s at 69%, then a YES share (which pays out M$1 on YES) costs M$0.69, and a NO share costs M$0.31. If you believe the probability is really 69%, you pay M$0.69 for a 69% chance at M$1, or M$0.31 for a 31% chance, which is a net expected value of zero. But if the true percentage is 80%, you pay M$0.69 for an 80% chance at M$1, for an expected profit of M$0.11. But beware: the probability, and thus the price per share, changes continuously as you bet! So there’s diminishing returns — although less so if it’s a highly liquid market, like one with many previous bettors or one that’s received a subsidy (an infusion of shares into the pool). This makes sense considering the pool of shares: to bet M$10 on YES, the 10 mana becomes 10 YES shares and 10 NO shares. You keep the YES shares, and trade the NO shares for more YES shares. But if there’s fewer shares remaining, you won’t get as many, since some shares must always be leftover. This is all done fractionally, by the way, with mana amounts rounded in the display. I’m a math person and find this all fascinating, so I’d love to go into even more detail about the system, but I don’t wanna scare away everyone else, so I’ll leave it at this.)
Anyway, if news breaks, the markets should move accordingly — allowing us to contextualize which news is important and which isn’t. (See the Base Rate Times!) These are quantitative predictions of the future instead of vague words: and if you think they’re wrong, you can profit! (At least in Manifold’s currency.) Scott Alexander’s FAQ gives more details about the practical applications and resilience of prediction markets more generally.
Manifold can also be used to gain more information about your personal life, motivate yourself on personal goals (by betting on yourself), and much more: it’s got sell orders and limit orders, PROB resolutions (which provide fractional payouts for YES and NO shares) and N/A resolutions (which cancel the market and revert all transactions, including sells), subsidies (which increase market liquidity, allowing more potential profits) and clever new multiple choice mechanisms, bounty markets and poll markets, automatic loans and unique bettor bonuses, leaderboards and leagues, charity donation and news tabs, an admittedly laggy website and capricious search feature, thriving comment sections and a Discord community where its employees are very responsive, an About page and this newsletter, and of course, The Market Manipulation Podcast.
Loved the post :)
great post! super funny :D