alan

@situationist Without subsidies, these markets naturally have an adverse selection spiral risk: spread widens to protect against informed flow -> elastic uninformed flow drops -> flow becomes more toxic -> spread widens -> 🔄

the scaling problem of current prediction market designs: the accuracy/benefits of prediction markets are sublinear with size, whereas the costs for liquidity providers scale linearly with size, because informed traders will exploit stale bids/asks for the full quantity offered.

@ito_markets I’ve also been thinking about how to do incentives in a more nuanced and targeted way 👀 And a certain kind of fund I think might emerge Will try to share in future posts! Keep on building 🫡

Cross - Subsidization in this context can be done with some of the methods talked about in this article. It can only be solved at scale by incentivizing a certain type of order flow and liquidity to flow into underfunded markets. LP Rewards with differentials to skew towards markets that are seeing less attention but are perceived to be widely important is one way of incentivizing the liquidity flow. This can be done by simply offering more rewards on MMing or LP Providing (in the case of AMM backed or pseudo AMM backed markets); and as the article pointed out these need not be permissionless, although they could be. It would also work in CLOB based centralized models with time variable rewards to limit orders placed * proximity to liquidity thin spots that are also close to the central limit This incentivization is what will grow the current 70/30 split between sports markets and everything else in PMs right now Which is obviously important if the thesis is that PMs are advanced financial instruments that allow you to hedge niche and complex risk profiles (think geopolitical and macro with a small exposure to "hail mary" markets within those profiles (one off rare events). We think these strategies will move the needle but the primary problem to solve is the liquidity in these markets of course. "It'll come from markets that are useful to people who never trade them" - this implies that theres a market segment that has demand but faces barriers + constraints to entering the market" That segment is institutions / funds , think boutique firms, quant dev shops, hedge funds, enterprise insurance, family funds etc. As ETF's solved this for Equities in the 90s (Toronto Stock Exchange and then Wall Street) - we saw a liquidity and flow explosion and the institutional interest and volume boomed That same layer - the ETF layer; is being built for prediction markets You bring the institutions in and you solve the need to rely on subsidization only Basketize an amalgamation of markets under powerful theses across platforms; determined by the desires of institutions - which is a representation of what the people want (their clients and the unique and interesting propositions they ask for). IE: A client wants to hedge "Legacy Tech underexposure to AI Adoption Risk" (could take either directional side on this, the client might want a directional play to short this basket implying that the risk isn't there because AI as a sector is trading way above the multiple of value its providing while a legacy tech giant such as Apple is still printing revenue even though its underexposed to AI - or vice versa) The baskets are then made up of a mixture of thin liquidity markets and higher liquidity markets, with weights attributed to this, we might skew towards the thinner markets in the basket with strong value props to "force" the institutions and buyers to fund that market, they are incentivized to do so and care less as the basket itself when looked at without its underliers has sufficient liquidity to trade the contract size they want to trade. We take directional plays on all the markets, we might stack the same market with different contract expiries in different directions due to conditional probability downstream effects as one market in the baskets outcome affects another, we then rebalance / redistribute after running an update step (analogue to training NNs), to recalculate the weights, cash balance + new markets to inherit if any. This also takes the binary expiry issue out the door as PMs rollover and now operate as futures contracts and in some cases depending on the basket could operate as perpetuals if theres a guarantee there will always exist demand for that basket and there will always be replacement markets to exist. That's what we've been building at Itô We come from a background of QR / QD at T1 IB/QF as well as extensive startup experience building the -1 to 1 in the prediction market space - both the tradfi and defi aspects. We met in undergrad as Mathematics x Computer Science majors that did applied research within the department of economics, we also ran our own startups at the time that leaned more towards the AI pricing / modeling side of niche data sources where we had moat. After some successful exists and some dissapointing failures - this is where we are at. Our conviction in this has never been stronger. The experience, agency, speed and knack for transfer learning to adjacent nascent markets we bring lead us to believe that without a shadow of a doubt, this is the right time, we are the right team and that we can grow this industry from 50 billion dollars to a trillion dollars by the end of this decade. Thank you for attending our TED talk. More to come soon.

@0xFleck Trading crypto & equity perps while hedging via option-like event contracts and real-world events all in one place Hyperliquid.

props to @alanwu. this is one of the clearest pieces on where prediction markets actually go next my top 3 takeaways (esp. for builders): 1. prediction markets scale when non-traders rely on them not when more degens show up → build for signal consumption, not just trading 2. the best markets wont sustain themselves, the price is a public good → you need intentional subsidies (fun markets funding useful ones) 3. prediction markets arent just probability engines they’re about hedging + capital-backed truth → in an AI world, money > opinions if you’re building: focus on markets the world would actually use if they existed worth reading the full piece 👇

@0xForecaster This is such a thoughtful summary! Appreciate the kind words 🤝

@TruebieMarkets Facts. Good resolution design is so important

"capital-backed specificity becomes more valuable as AI makes words cheaper" mmmmmm dat tldr everything else here - cross-subsidization, hedging, public goods - only holds if the resolution criteria survives contact with a contested $10M market without trust, you only have a fragile truth.

@pangramlabs @pseudoSeneca 💀reminds me of @Turnitin from high school

@pseudoSeneca @alanwu We believe that this document is fully human-written pangram.com/history/64bf27…

@goodbreffis te extraño got me 💀

Blink-Juan82 🤟🤟🤟

@lDrewcifer @Riot_Kassadin Old School RuneScape but for League

@Riot_Kassadin I just want a game mode with Classic League. Literally call it “Classic” and it has all the old champs with their kits from forever ago. It would be so nostalgic!

League of Legends about to look like this in a year 💀

@Y2K_DOTCOM Makes me think of Kongregate and Adventure Quest too 🥲

they all just vanished one day

@blockchainucla paradigm pilled

Tomorrow's meeting will be about OpenAI and Paradigm's newly released EVM Bench, a benchmark for evaluating AI agents in web3 auditing and what that means for the AI agent and Web3 ecosystem! 6 - 7 pm @ The Connection Lab. See you there!

@Jun1on @sudolabel x.com/taikimaeda2/st…

@sudolabel bro forgot to sell the farm token

back in 2021, there was a coin that was mined with storage (chia) so i put a petabyte in my parents' basement for ~$10k anyways, i never sold a coin and naturally, it's down 99.8% since

@Azzapp_LoL surprised this one isn’t in the comments yet!

In the recent blog Riot asks which other items they should bring back. There is only one candidate. Twin Shadows is the best item ever designed in League of Legends and there is no contest. Bring the boys back home!

@SaskioLoL and sometimes, under certain circumstances, your jungler can wingman you

The talking stage with a girl is like laning against a new opponent in League of Legends you load into lane. no data. no opgg. no idea what this person is capable of level 1 you're just farming small talk. surface level. "what do you do" "where are you from" you're not actually listening for the answers you're watching HOW they answer fast reply? aggressive player. trades back immediately. likes short trades. slow reply? freezing the wave towards them. controlling tempo. making YOU walk up. long message with questions? 3 wave crashing into you. she WANTS interaction. level 3 you start testing you throw out something slightly personal. not a full combo. just a Q to see if it lands. if she trades back? the lane is active. keep going. if she backs off? she's not ready. paly the wave. try again later. now you're tracking cooldowns how long between texts. does she double message. does she leave you on read then come back. does she initiate or wait for you to walk up every single time. you're reading every movement for intent she randomly goes cold? enemy jungler opposite side, play aggressive she suddenly gets warmer? her jungler is nearby. play safe or accept the 1 v 2 then there's the moment every laner knows the all in window happens something shifts. she says something real. drops the act. shows you what's underneath the trading sequences. this is the moment you pull the trigger or you don't and if you don't? the window closes. the wave resets. and you go back to farming safely waiting for summoners wondering what would have happened but if you DO the lane is never the same, you lose all priority that first all in changes everything. the spacing is different. the energy is different. you both know something now. every interaction after is built on that one commit some players farm the entire laning phase. never trade. never all in. never take the risk. and they wonder why they're stuck in the same elo forever the talking stage isn't about being smooth it's about reading your lane opponent well enough to know the EXACT moment to go all in and not being a pussy when the window opens Study the Saskio way

@ycombinator @Heypocket @AkshayNarisetti @gabrieldymowski they shipped Norma

Pocket (@heypocket) is your notetaker for real world meetings. In the last 5 months, the team has delivered over 30k units with a $27M annualized run rate, growing 50% month over month. Congrats on the launch, @AkshayNarisetti and @gabrieldymowski! ycombinator.com/launches/PaX-p…

@escapefrommelos His son will be Hyperliquid Chen.

I am sorry but WHAT

@llamaonthebrink Ppl committed capital on defined terms… then the terms got redefined 🥸 Immutability would make these instruments more credible. Poor resolution criteria gets priced in and the liquidity will reflect it

A small contingency of people are in favor of prediction markets providing additional context in the form of “clarifications” mid-market. But I’m firmly against this. A contract’s rules should be immutable. If the rules are poorly drafted, the contract won’t see serious action. The market effectively self corrects for the potentially missing context. @Trueo_app is designed this way, all rules are embedded as strings onchain, making them immutable once an event contract is deployed.

@OldSchoolRS Dungeoneering party with Alysa??

Gz Alysa Liu on the Olympic gold 🥇Even the penguins broke cover for this one.

@dionrraylol How thoughtful of them to speedrun his 0/10 powerspike