sofiechan home

Prediction Group House Idea

anon_pici said in #4966 7d ago: received

Years ago, before the rise of Polymarket and Kalshi, I read Phil Tetlock's Superforecasters and thought hard about the implications. I've my more or less my entire career investing and the book resonated with me immediately. The basic concept is straightforward:
Experts are generally speaking horrible at making predictions. Their personal biases and lens through which they view the world leads to extremely precise and extremely wrong predictions.
Superforecasters, by contrast, take in disparate information sources, make predictions that are consistently better than average that start from a base rate, biased with new information.

The important implication: one could spend 30 years studying geophysics and refinery crack spreads, raising $100mm to launch a commodity hedge fund attempting to predict oil prices. However, they would be in competition with thousands of other analysts trying to predict oil prices. A better approach would be to use superforecasting skills to predict real world events.

The next implication was the most electrifying: instead of competing with Bridgewater for investment alpha and Bridgewater for investment talent, why not create a residential program for bright, aimless, young men and train them
to be superforecasters.

I call it a Prediction Group House. Basically a monastery in a rural area with a dormitory, gym, kitchen, shooting range, shrine, library lecture hall, and workshop.

Young men would:
Study diligently in the art of Superforecasting
Engage in group fitness
Hit the gym
Learn sparring and grappling
Eat together

In exchange they would be given:
Room and board, a minimal stipend, and once they had mastered the basics, they would be given a grubstake to trade on the prediction markets.

Young men would be given a slope like at a hedge fund, in the beginning, they invest zero percent of their own capital, but earn 10% of their profits. As they acquire a track record of successful predictions, they would be given a greater share of the profits from their trading. As they gain even more experience, they could buy into the partnership and thus earn a share of all of the house profits.

The goal would be for young men to leave the house mentally, physically, spiritually and financially better off than 99% of other young men, perhaps going on to launch new houses (potentially grubstaked by the first house).

Let me know if you see any major flaws of this idea

referenced by: >>4970

Years ago, before th received

anon_feka said in #4969 7d ago: received

It's an interesting idea. The key questions are
1. whether you can train generic smart guys to have a positive edge on Polymarket / whether "superforecasting" is actually a learnable skill, and
2. whether there's enough volume on Polymarket to make this a profitable exercise

Many of the interesting markets on Polymarket are surprisingly low volume. Look at this market. Very interesting question, clearly important, less than $2m total volume, even according to Polymarket's extremely generous methodology for counting volume.
https://polymarket.com/event/khamenei-out-as-supreme-leader-of-iran-by-june-30-747?tid=1767845219462

Far more of you have heard of Polymarket than the other big crypto market success story, Hyperliquid, but Hyperliquid is the one with 20x+ Polymarket's the volume and open interest.

This is unfortunately a structural weakness. Prediction markets overperform as media entities. As much as CNBC tries to make "news" out of daily Nasdaq dips and bumps, the kinds of the events Polymarket covers are fundamentally more interesting. But prediction markets severely underperform on liquidity, in significant part because these same interesting markets are the ones most susceptible to insider info. For media, this is a great thing! Having a source reveal, a few hours ahead of time, that we're about to capture the president of Venezuela is kino breaking news. But for liquidity, it's terrible: nobody wants to make markets or trade in these shark waters where the people who come to trade against you are the ones who know something you don't. Some trad markets, like commodity futures, have this problem too, but to nowhere near the extent that prediction markets do.

TLDR; there are two risks to your idea.
One is that the most interesting prediction markets are unfortunately micro-caps for structural reasons.
Two is the format... if "gym-monestary in the middle of nowhere" was the right way to do this, why are the big trading firms all in skyscrapers in NYC?

Maybe you are right, though. Maybe prediction markets are too small to build a big fancy hedge fund or prop firm, but just the right size to build a kind of specialized hacker house with staked individual players. I've seen houses where everyone is "training" for "esports" and on house in Vegas full semipro poker players, so there's precedent. Would be a fun experiment for sure.

referenced by: >>4973

It's an interesting received

anon_swqy said in #4970 7d ago: received

>>4966
I think the rationalists tried stuff like this (getting underemployed young people to train in superforecasting). They do still play around on the prediction markets, but I haven't heard of any systematic success. They had a lot of epistemic and cognitive advantages, so I take that as evidence against this idea. Their group houses seem to more reliably revolve around doing programming work, especially in AI. Who knows though it might work if you had the right master and the right students.

I've been thinking of a different application for a similar idea though: start a school. We suppose we have an oversupply of underemployed NEETs who need to be trained in doing the reading, pursuing gymnastics, music, and mathematics so they can become guardians and philosopher kings. But they aren't good enough to do world class prediction or anything like that. What is an intellectual activity where you don't have to be best in the world, you just have to be available and adequate? Teaching children. The NEETs will train the next generation of guardians. This, if arranged with seriousness, could be a sustainable private education model for kids in our network.

referenced by: >>4972

I think the rational received

anon_pici said in #4971 7d ago: received

1. I would strongly bet that it is, if your raw material is 95th percentile or better IQ.
2. This was where the idea broke down last time. Before Polymarket and Kalshi there was zero liquidity. Now the question is scalability, you could an put on a lot of $10,000 trades but no ten million dollar trades.

I agree with your point on liquidity, you need informed traders, market makers, and flow traders to make a market. Hopefully with better advertising, you'll se a lot more flow traders on the prediction markets which would be enticing to market makers who can diversify event risk away.

I think partly because their business model requires being in NYC. If you want to be a credit hedge fund, there is a 6 block radius where all the talent for credit hedge funds is located. If want to recruit 2-3 year experienced analysts out of bulge bracket banks, they are all there. This model obviates that need by targeting thoughtful men between 18-22, college totally optional.

The esport/poker/hacker house concept is an interesting one, thanks!

1. I would strongly received

anon_pici said in #4972 7d ago: received

>>4970
That's a pretty good piece of negative evidence. can you DM me a name on Signal (assuming you know who I am) so that I could hear from them what worked and what didn't?

I think I have one major advantage over the rationalist version: there'd be no trans/fetishism/drugs, none of that freakshow stuff.

referenced by: >>4973

That's a pretty good received

anon_swqy said in #4973 7d ago: received

>>4972
I can DM you on here if you vouch me back. But I don't think we need to. I don't know any details, just that the rationalists were obsessed with this stuff, obsessed with prediction markets, and CFAR tried all kinds of stuff to make training work (in stuff like superforecasting etc), and consensus is none of it really worked that well. That said, maybe because as you note the liquidity has increased, the returns on marginal success would be higher now so you might get further before giving up.

>one of that freakshow stuff.
I don't think the freakshow stuff took root until later (late 2010s), and I don't think it affected their ability to deliver on these fronts.

>>4969
>But for liquidity, it's terrible: nobody wants to make markets or trade in these shark waters where the people who come to trade against you are the ones who know something you don't.
Good point. Prediction market trading is almost pure adverse selection unless you really know something. It's not rational for most people to trade.

I can DM you on here received

anon_pici said in #4975 6d ago: received

>It's not rational for most people to trade.

couldn't the same be said for literally all forms of gambling? You know the house has an edge but you gamble anyway. How are prediction markets different? People buy and sell prediction contracts for all kinds of idiosyncratic reasons, maybe their buddy who is a national guardsman heard from his cousin who is a Navy SEAL that we are invading Iran so he buys "Yes," but his cousin is a liar. I still believe that in aggregate most people have no idea and bet anyway

couldn't the same be received

anon_pici said in #4976 6d ago: received

Now that prediction markets a proliferating on numerous platforms you might not even need an edge, maybe probability arbitrage is profitable! You find Iran Yes trades at 25 cents on Platform A and 35 cents on Platform B but the contracts are worded identically, that's an arb. This is basically how FTX made their first buck before they became a fraud.

Now that prediction received

anon_joxu said in #4978 5d ago: received

Young men would:
Study diligently in the art of Superforecasting
Engage in group fitness
Hit the gym
Learn sparring and grappling
Eat together
+ Engage in serious contemplative practice

It could be like the Christian monks who sell beer for a living. A group lifestyle business.

referenced by: >>4979

Young men would:... received

anon_feka said in #4979 5d ago: received

>>4978

Is this a clanker summary? …
…
…

Why?

referenced by: >>4980

Is this a clanker su received

anon_joxu said in #4980 5d ago: received

>>4979

I'm unsure whether the goal of the house would be to just have a crypto startup with a bunch of jacked frat guys or to actually achieve something spiritually legitimate.

I'm unsure whether t received

You must login to post.