MagicMaker

@Eltonma Cozy in 4th with no incentives. Just pure flow. x.com/0xMakingMagic/…

Seems new month, the rewards are no longer sports heavy, I am back to #1 spot (BY A LOT).

@sbinfluencer @JonSchweppe $1300 every bet

@JonSchweppe I used to love betting on sports. Then I saw this post, and I realized how much the sportsbooks have been screwing me. I have you to thank for that.

Just a reminder: sportsbooks absolutely screw their customers. $3,000 to win $100. $100 to win $1,300. That $1,700 difference? That’s their take, lol.

@PineTreePredict @base @jessepollak Dms aren't open ser

@0xMakingMagic @base @jessepollak Sounds good DM me when you can

The house always wins. Until the house is removed. No limits on winners, and zero commission. We aren’t building a sportsbook. We’re building the infrastructure for the next generation of p2p event contracts on @base. 🌲⛓️$PTP #BuildOnBase #DeFi #PredictionMarkets @jessepollak

Switching from back testing to live. Many such cases.

@glitchtruth @Kalshi Prediction markets are designed to surface insider knowledge and information asymmetry.

Prediction markets have an insider trading problem nobody talks about. If you know a Fed decision before it's announced, you can bet on @Kalshi legally. If you do the same thing on the NYSE, it's a federal crime. The legal protection exists because they're technically 'event contracts,' not securities. The economic incentive to exploit non-public information is identical. The regulatory treatment is not.

@withqwerty @mplsoccer_dev @kloppy_dev Beautiful. Will take a look.

thrilled to announce that campos ⚽🧤 is now in public beta: campos.withqwerty.com the aim: quickly and painlessly go from raw football provider data to beautiful, interactive charts. think of it as @mplsoccer_dev + @kloppy_dev for the web. good presets, endlessly customisable, sensibly interactive. campos adapters turn opta, statsbomb, wyscout and the rest into tidy events you drop straight into campos components - no more coordinate flips, no stress. build by hand or with agents paired with nutmeg.withqwerty.com skills open source, forever. contributions welcome - new chart types, more adapters, new features, bug fixes, anything. gracias, jorge.

@cjhtech @CryptoDiffer @trylimitless Inevitable.

interesting

@trylimitless Nice!

We just crossed $3.5 billion in total traded volume Almost $300 million in the last 5 days alone What did you get done this week?

@NickPreszler Token2049 didn’t happen, so I’m now overdue a conference… but probably won’t make it.

Who's going to Miami next week? If you're building or trading prediction markets, let's find a time to chat. DM me or drop a comment and i'll reach out.

@decenterghost Good post

Polymarket Brier 0.063 is excellent calibration on a curated set of markets. Not on all markets Brier 0.063 doesn't mean Polymarket is a "truth machine" It means PM is well-calibrated on the markets it chooses to list Three filters PM applies that polls don't: > Liquidity filter No market for "will this startup fail in 18 months" -no verifiable outcome, no volume. Polls don't get to opt out of hard questions. PM's selection criterion isn't difficulty -it's verifiability. That's a different bias, and a structural one > Base rate skew A 0.90 market has a Brier floor of 0.01, but an expected score of 0.09. A 50/50 market has a floor of 0.25. If PM systematically lists high-confidence markets, Brier 0.063 is partly an artifact of mix, not accuracy. The number looks sharp. The denominator is doing work you're not seeing > Time resolution Brier score doesn't distinguish predictive power from speed of convergence. A market that moves to 0.99 five minutes before resolution scores nearly perfectly and tells you nothing useful ex ante. Polls measure uncertainty before the fact. PM measures convergence toward the fact. Comparing them through a single Brier score is a methodological error, not a calibration triumph PM is sharp. But "truth machine" is a narrative for people who've never had to model what happens when the oracle itself is the trade Price on PM is information and information that feeds back into the behavior of the agents it's supposed to measure. That's not a prediction market anymore. That's a reflexivity machine with a good Brier score

@ChairmanSelig @CFTC I'm going to write an article about how modern Prediction Markets developed, and how they are explicitly designed to surface information asymmetry and punish those who are uninformed.

Claims that “insider trading is rampant” in prediction markets because the @CFTC's antifraud rules are “fuzzier” than others are simply untrue. During my first 100 days leading the agency, we’ve strengthened and modernized our strategies for detecting illicit activity and taken legal action against bad actors. Read my full letter to @WSJ correcting the record: wsj.com/opinion/predic…

@ToddFuhrman Use an exchange.

If a sportsbook won’t offer you a 2 way market then you’re almost always getting gouged *any time goal scorer *any time TD scorer *player to hit a homerun *racing placement markets Not saying these markets never offer value but they’re the very definition of buyer beware

@cjhtech Awesome thanks.

@0xMakingMagic docs.limitless.exchange

Limitless in April 2026 > $1.5B in monthly volume > crossed $20M annualized revenue > #2 largest onchain prediction market > #3 largest globally including US regulated markets > football markets achieving emergent PMF > continued growth of short duration prices contracts, a prediction market segment @trylimitless pioneered in 2024 which now represents 40% of all onchain prediction market revenue across platforms > too many product & API upgrades to mention it's not too late to become limitless. but the window of opportunity to be early is closing.

@cosmicgateAI If you’re focused on Sport, let’s chat. x.com/0xMakingMagic/…

A common question is whether new prediction market startups can realistically compete with incumbents such as Kalshi and Polymarket. The better framing is not whether they can displace incumbents outright, but whether they can capture meaningful share in a category that is still early and structurally underdeveloped. Prediction markets are projected to reach a $ 325B+ run rate in 2026, yet the market remains far from saturated. That leaves room for new entrants, especially those that do not approach the category as just another exchange. For new startups entering the space, the opportunity is unlikely to come from copying existing models. It will come from building differentiated infrastructure and distribution around specific user behavior. Today, most prediction markets still share the same core limitations: - similar product structures - weak social engagement loops - cold-start liquidity challenges and - limited GTM differentiation We think the next generation of category leaders will be defined by three things: 1. Defensible product moat Not just market access, but a product experience that is difficult to replicate. 2. Distribution advantage In prediction markets, GTM is not secondary to product. Distribution is often the deciding factor in whether a platform reaches escape velocity. 3. Agentic infrastructure As AI agents become mainstream, there is a growing opportunity to re-architect how markets are created, how signals are aggregated, and how users interact with prediction systems. Most incumbents have not yet fully integrated AI into the execution layer, intelligence layer, or settlement workflow. We believe the strongest opportunities in prediction markets sit at the intersection of verticalized distribution, social participation, and agentic intelligence. At cosmicgate.ai we are building that agentic layer for sports and entertainment, turning fan attention into consumer markets.

@KalshiNewsroom Prediction markets were specifically designed for “insider trading” and exposing information asymmetry. What’s the definition?

"Claims that 'insider trading is rampant,' and that our insider trading rules are 'fuzzier' than others are simply untrue. I’ve made it clear time and time again that anyone who engages in insider trading will be found and prosecuted to the full extent of the law." wsj.com/opinion/predic…

@Trueo_app What about: “prediction markets are just markets”

>“Prediction markets are just gambling” Everyone: “Awckshually, there is real societal and epistemological value to free market information aggregation…” Polymarket:

@Vicmetax They feel broken for sport too. Some insane settlement outcomes on known results.

Yo, pull up real quick I’ve been diving deep into prediction markets lately and had to share this with you. We all saw Polymarket and Kalshi crush it with billions in volume during the elections. Skin-in-the-game forecasting feels like the future. But if you’ve actually tried using them for anything beyond politics or sports… it still feels broken for normal people like us. Here's the full gist 🧵 🔻

@BlondiePredicts Sharp sports bettors exist, and have repeatable edge. This isn’t the distinction. They’re the exact same.

To be successful in prediction markets month over month, you need real skill and a real edge. Who in gambling has a positive PnL month over month? Pressing a slot machine button isn’t a skill you can improve