taetaehoho runs an empirical analysis of Polymarket's liquidity rewards and sponsorships, finding they can tighten spreads only when daily spend exceeds roughly 1% of the existing book. Even above that threshold, incentive size barely predicts which markets respond; pre-existing liquidity conditions and pre-trends are far stronger determinants of uplift than reward intensity.

Compares spreads across DraftKings, FanDuel, Polymarket, and Kalshi on identical sports markets to isolate how counterparty identification versus maker competition affects pricing. Finds PM prices are 100-300 bps better on liquid markets but long-tail markets on prediction markets have 10-50% spreads. Argues sportsbooks tighten spreads through counterparty-aware pricing while PMs compensate through maker competition and order book transparency.

Analyzes Polymarket's short-duration crypto up/down markets, which now account for 16% of monthly volume but generate ~40% of platform fees. Finds professional algo traders (19 addresses) extract consistent profits through paired trading and maker execution while 69% of retail traders lose money. Argues prediction markets are becoming a game of volatility rather than forecasting.

Isaac Rose-Berman dissects Kalshi's claim that as an exchange rather than a bookmaker, it has no stake in whether users lose. Through a detailed walkthrough of maker-taker fees, adverse selection and the zero-sum structure of prediction markets, he argues the business depends on ordinary users systematically losing to sophisticated market makers just as surely as any sportsbook. A coin-flip bet that should be fair becomes a 3.4% expected loss after fees.

A trader who ran ~10,000 automated trades through Polymarket dissects why most prediction market terminals fail: execution mirages, non-synthesizing research layers, and strategy tabs that are deck screenshots disguised as features. Lays out the baseline infrastructure real traders need and argues only two terminal types survive the next 24 months: institutional API rails and a trader-native terminal built by someone who has actually lived the workflow.

Argues that central limit order books fixed the capital destruction problem of early Polymarket AMMs but introduced a new pathology: passive participation is impossible, and only professional market makers can quote. Kalshi reportedly has 23 active market makers with the top three providing 70% of election-contract liquidity, meaning any market those firms ignore is dead on arrival. Positions this as the reason prediction markets remain concentrated in politics and sports while entertainment, science, and culture verticals stay empty, and makes the case for peer-to-peer architecture that lets the first participant seed liquidity for the second.

Part 1 of a Ranger Global research series on onchain prediction market microstructure. Walks through why CLOBs beat constant-product AMMs for binary events, the YES/NO minting and merging invariant that lets depth expand whenever matched counterparties exist, and probability-scaled dynamic fees that shrink near 0 and 1. Closes with a regression of prediction market midpoints against BTC spot, finding PM traders systematically underreact to spot moves by 10-20% and that latency under 100ms now captures 73% of arbitrage profits.

Survey of recent research on why conventional market making fails in prediction markets. Covers cross-venue fragmentation (the same contract at 58-67 cents on different platforms), the January 2026 Polymarket XRP exploit that paid $231K on thin weekend liquidity, Kalshi's structural longshot bias, and evidence from 150M Polymarket trades that the top 5% skilled traders earned $228M while spread capture barely moves P&L. Concludes passive LPs on these venues behave more like underwriters of terminal risk than classical market makers.

Response to Axios and More Perfect Union coverage framing prediction markets as gambling. Park argues the investing/gambling line is about +EV, not the game itself, and that the flip side of speculation is always insurance. Contends prediction markets are structurally different from other derivatives because they are precise (binary payoffs create clean basis risk to truth) and have finite expiry, and that fears about insider trading are overblown since liquidity in obscure asymmetric markets will be negligible. Closes with a media-criticism argument that prestige outlets attack prediction markets because the markets threaten institutional control over truth.

Technical research on adapting the Avellaneda-Stoikov market-making framework for prediction markets. Classical models fail because prediction market prices are bounded probabilities (0 to 1) rather than unbounded asset prices, creating non-constant volatility and guaranteed terminal convergence. After a logit-space transformation lost $1,114 in backtest, XO Labs iterated to a probability-space engine with inventory-skewed spreads, volatility regime detection, and multi-outcome coordination that turned profitable at $453. Includes the full mathematical framework and open problems.

On-chain analysis of Polymarket's fast crypto markets from September 2025 through March 2026. Five-minute contracts overtook 15-minute in weeks ($2.3B vs $795M in notional volume), bots control 55-62% of volume across fast markets, and Bitcoin drives 77% of turnover. The piece argues that with $23.7M in taker fees collected in 83 days through a maker-rebate fee model, Polymarket has structurally converged with a derivatives exchange.

Applies the sportsbook industry's square vs. sharp distinction to prediction markets. Argues Kalshi (3.5% take fee) and Polymarket (1.5%+) are building square prediction markets that monetize price-insensitive retail takers, while sharp prediction markets focused on trading efficiency will thrive outside the US. Notes that raising fees from 0.5% to 0.75% on 4casters had no material impact on volume, suggesting sports bettors are less price-sensitive than assumed.

Landscape report comparing four models for adding leverage to prediction markets: lending pools (Gondor/Morpho-style), prime brokers (Ultramarkets), synthetic desks (CFD counterparties), and perpetual futures (dYdX TRUMPWIN). Sizes the fee revenue opportunity at $15M base case to $50.7M bull case, with 87% driven by financing revenue on open interest rather than trading fees. All four models share a structural dependency on CLOB venue architecture that degrades during jump events.

Traces the parimutuel betting model from its 1867 origins in French horse racing to modern prediction markets. Argues that parimutuel pools solve the cold start problem by bootstrapping liquidity without market makers, but identifies three friction points that need upgrading: locked positions, timing constraints, and price readability.

Categorizes every active Polymarket wallet across trade frequency and volume tiers, producing a seven-persona map of platform participants. Finds that 2% of users (high-frequency, high-capital wallets) generate roughly 90% of all platform volume, with crypto markets dominated by algorithmic execution and politics markets driven by casual event-driven participants. Draws implications for fee design, category expansion, and why optimizing for user growth versus volume growth requires fundamentally different product decisions.

Wall Street equity research analysis of anonymized trading data comparing prediction market and sports betting returns. The median prediction market user has an ROI of -8%, worse than sports bettors at -5%, with only traders above $500K in volume achieving positive returns (+2.6%). Finds that prediction markets attract sharper competition than regulated sportsbooks, creating worse outcomes for casual retail participants.

Challenges the smart-versus-dumb money dichotomy in prediction markets by synthesizing research from Snowberg/Wolfers/Zitzewitz, INSEAD's BIN model, and Wharton's cognitive search framework. Argues that noisy traders fund the probability space rather than serve as exit liquidity, and compares how binary CLOBs versus continuous probability markets decompose and harness noise differently.

Data-driven deep dive into Polymarket's order book structure using 600M+ raw datapoints filtered to a 343M research dataset. Categorizes order flow into soft (retail), hard (professional), and AI flow, revealing that Polymarket's liquidity is episodic and attention-driven: the p95 peak hour shows hundreds of millions in open interest while the p50 median is thin. Order book analysis shows surface symmetry at top-of-book but systematic ask-side skew at deeper levels, and market impact data confirms that medium-to-large orders hit liquidity cliffs. Argues the core problem is trapped capital — dollars reserved multiple times against mutually exclusive outcomes — and that better netting and capital efficiency, not more money, is the fix.

Uses the case of a trader sniping Polymarket's geopolitical strike markets at 10 cents to argue that continuous order books are structurally broken for binary assets. In traditional markets, sniping costs basis points; in prediction markets, it costs 80 cents on the dollar because prices jump from 0.10 to 0.99 on a single tweet. Proposes frequent batch auctions (citing Budish, Cramton, and Shim) to shift competition from speed to price accuracy, and introduces the concept of a 'liquidity mirage' where the highest social-value markets are precisely those where passive liquidity is unsustainable.

Comprehensive research report on AI agents for prediction markets, proposing a four-layer architecture (data, analysis, execution, learning) for autonomous trading systems. Maps the ecosystem of existing agents, compares Kelly criterion vs fixed-fraction bet sizing, surveys arbitrage strategies across platforms, and outlines business models (agent-as-a-service, liquidity mining, data sales). Argues that AI agents will become the dominant market participants within two years, transforming prediction markets from retail-driven speculation into infrastructure for probabilistic information.

A Reforge co-founder's bearish case against prediction markets, structured as 23 distinct failure modes. Covers structural constraints across capital efficiency, liquidity mechanics, adverse selection, oracle governance, and regulatory fragmentation. Argues that prediction markets face fundamental limitations that perpetual futures markets do not, making institutional scaling unlikely under current designs.

Traces the quantitative finance toolkit from backtesting (Deflated Sharpe Ratio, combinatorial purged cross-validation) through factor models, Black-Litterman portfolio optimization, Bayesian regime detection, and machine learning, then argues each technique transfers directly to prediction markets. The core claim is that prediction markets are the purest testing environment for investment theory because binary resolution eliminates the unobservable noise that obscures strategy quality in traditional finance. Uses LMSR's mathematical identity with the softmax function to bridge quant finance and prediction market pricing.

Analysis of 149 CPI prediction markets on Kalshi from 2021 to 2026 finds that trading volume explains less than 1% of variance in forecast accuracy, challenging the assumption that more liquidity improves market quality. Introduces Minimum Viable Liquidity (Cost of Expertise divided by Price Gap) as a framework for determining the threshold of liquidity needed to attract informed traders. Argues platforms should prioritize breadth over depth, running many thin markets rather than concentrating volume in few contracts.

Argues that insider trading in prediction markets is structurally different from traditional securities markets because prediction markets can make almost anything tradable, often in contexts where relevant confidentiality duties are unclear. Proposes solutions across three layers: platform-level detection and position limits scaled to account size, market design mechanisms like dynamic spread widening and market maker insurance pools, and legal frameworks from updated corporate compliance policies to CFTC guidance.

Decomposes 222 million Polymarket trades into directional and execution components and finds that forecasting accuracy does not predict profitability. Traders who pick the right side still lose money because they arrive late and pay unfavorable prices, while automated traders with near-random directional skill profit by paying 2.52 cents less per contract.

Compares Kalshi and Polymarket's NFL game markets during the 2025 season. Finds Kalshi reprices faster (median 7-second lead) while Polymarket has deeper liquidity requiring 3-4x more volume to move prices comparably. Uses Kyle-style market impact analysis to quantify the price discovery vs. liquidity depth tradeoff between centralized and on-chain order book architectures.

Summarizes research analyzing 72 million Kalshi trades. Identifies three persistent biases: longshot bias (5c contracts win only 4.18% of the time), maker-taker asymmetry (makers outperform at 80 of 99 price levels), and YES/NO asymmetry (YES buyers average -1.02% returns vs +0.83% for NO buyers). Finance markets are most efficient (0.17% spread) while crypto is least (2.69%).

Builds a formal framework to decompose why prediction markets have late volume: is it because information arrives late (hazard), or because early entry is punished by adverse selection (toxicity)? Introduces LOX, a metric computed from on-chain trades that measures whether new entrants hesitate more than volume alone would predict. Explains why boxing markets cluster with news markets despite being categorized as sports.

First-person account from a seven-figure prediction market trader. Covers strategy as a 'bond mule' (locking capital for small premiums on near-resolution markets), OSINT information sources (non-English media, Telegram), fractional Kelly sizing, and trading personality-driven markets. Notes edge has compressed as space matured.

Analyzes 72.1 million trades ($18.26B volume) on Kalshi and documents a systematic wealth transfer from takers to makers averaging 1.12% excess returns on each side. Takers disproportionately buy YES longshots, accepting returns 64 percentage points lower than equivalent NO positions. Shows this transfer only emerged after Kalshi's October 2024 legal victory attracted professional algorithmic market makers, and that market efficiency varies sharply by category: finance markets are near-efficient while entertainment and media markets show gaps exceeding 7 percentage points.

Argues price prediction markets (short-term expiries like 'will BTC close above $100k?') represent a new asset class. Compares AMM vs CLOB mechanics, notes Limitless achieves 50-400bps spreads (better than onchain options at 1000+bps). Outlines how prediction markets enable synthetic covered calls, structured hedging, and volatility expression.

Argues that First-Come-First-Served order matching in prediction markets creates perverse incentives: latency wars between traders and defensive spread widening by market makers. Proposes priority batch auctions that process cancellations before maker orders before taker orders, allowing market makers to quote tighter spreads and improving price quality for all participants.

Examines how to distinguish wash trading from legitimate market making on Polymarket using network analysis. Wash traders exhibit homophily, trading only within their collusive group, while market makers trade indiscriminately with diverse counterparties. Describes an algorithm developed by Columbia researchers that identified a cluster of 200 wallets generating $113 million in volume with just $57.86 in aggregate losses.

Proposes a unified stochastic kernel (logit jump-diffusion) for prediction markets analogous to Black-Scholes for options. Treats traded probability as a risk-neutral martingale, exposing belief volatility, jump intensity, and correlation as quotable risk factors. Defines derivative instruments (variance swaps, correlation swaps, first-passage notes) for hedging belief risk.

Applies adverse selection concepts specifically to prediction markets. Compares market making difficulty across Indian options (easy), crypto (medium), and prediction markets (legendary). Argues gap risk is effectively worse than any other asset class because informed counterparties can have near-perfect information and take out entire order books.

Compares the traditional sportsbook house model with prediction market exchanges to explain why exchanges offer better odds. Uses data showing Betfair's ~3% overround versus bookmakers' ~12% to argue that peer-to-peer exchange models produce fairer pricing and welcome all winners, unlike sportsbooks that limit successful bettors.

Educational deep-dive on adverse selection and market making fundamentals. Uses the classic Jane Street coffee interview question to illustrate why conditional on someone trading with you, you should be less confident your trade was good. Explains how market makers profit from retail flow while avoiding toxic informed counterparties.

Introduction to a series arguing prediction market mass adoption is threatened by structural barriers to market maker participation. Claims binary markets are frequently unhedgeable and suffer from adverse selection, making them a professional market maker's nightmare compared to options or crypto.

Skeptical take arguing prediction markets are dangerous policy despite theoretical appeal. Claims they lack heterogeneous risk preferences necessary for efficiency, relying instead on continuous retail losses. Warns that large-scale prediction markets exhibit reflexivity, potentially incentivizing manipulation toward negative, sensational outcomes.

Argues prediction markets face fundamental barriers to scaling beyond sports and elections due to misaligned incentives: speculators demand rapid resolutions (42% of election volume in final week), investors prefer wealth-building assets over locked capital, and market makers need consistent retail flow that doesn't exist for niche topics.

Argues that prediction markets' unpopularity isn't due to regulation but fundamental demand-side issues. Markets need savers, gamblers, or sharps to function, but prediction markets attract none: they're zero-sum (no savers), have long resolution times (no gamblers), and are too small for professional traders (no sharps).