IOSG | 70% of Top Profitable Wallets Are Bots, but AI Hasn’t Taken Over Prediction Markets Yet
TechFlow Selected TechFlow Selected
IOSG | 70% of Top Profitable Wallets Are Bots, but AI Hasn’t Taken Over Prediction Markets Yet
Prediction markets are wealth-transfer machines, and bots are their operators.
Navigating Web3 tides with focused insightsAuthor: Jeff, IOSG
Executive Summary
The core data fueling panic around prediction market bots is strikingly straightforward: on Polymarket, 5% of wallets classified as “bot-like” account for 75% of trading volume. Since January 2025, 823 wallets have each netted over $100,000, collectively withdrawing $131 million in profits from Polymarket. Among the top 20 profit earners, 14 are classified as bots (per Stacy Muur leaderboard inspection). A University of Toronto study—covering 2.4 million users and $67 billion in trading volume since 2022—found that 68.8% of users are unprofitable, while the top 1% captured 76.5% of all profits.
This yields a compelling narrative: prediction markets are wealth-transfer machines, and bots are their operators. The data is accurate—but the framing is half-biased.
Core Thesis
1. The central flaw in the bot narrative lies in conflating “trading volume concentration” with “capital extraction.” The fact that 5% of Polymarket wallets generate 75% of volume reflects only the distribution of account activity—not direct evidence that retail capital is being extracted by bots.
2. Group-level data is more telling. AI agent wallets achieve a ~37% win rate; human wallets, only 7–13%. This 3–4× structural gap at the population level is real evidence of systemic advantage; the fact that 14 of the top 20 profit earners are bots (per Stacy Muur leaderboard inspection) is merely the right-tail projection of that distribution—not independent evidence.
3. Bot advantage operates on structural—not judgmental—dimensions. Bots dominate three market types: price-feed latency arbitrage, automated betting-state execution in live sports, and cross-platform combinatorial arbitrage. Their commonality? None require subjective judgment about real-world outcomes. Once market resolution depends on synthesizing multi-source real-world information, bot advantage systematically erodes.
4. Polymarket’s category composition has shifted over the past 12 months—from “Politics 42%” to “Sports 50%.” The fastest-growing categories are precisely those where bots hold no structural edge: long-horizon event markets. The platform’s overall retailization trend is unmistakable.
5. Forward-looking view: Bot share will continue rising as deployment costs fall—but the scale of capital extraction from humans by bots will peak *before* bot share does—because bots cannibalize each other faster than they extract from human accounts.
6. Investment strategy: Equity opportunities at the platform layer (Kalshi + Polymarket collectively hold >97% share) are effectively closed to venture-scale checks. Value is migrating upward to L2 agent infrastructure layers (e.g., Olas / Valory models) and venue-agnostic middleware—and downward to C-end bot products and L3 data / pricing layers, neither of which fits venture criteria.
I. Market Scale Outweighs Bot Panic
Three quantitative anchors define the scope of this report.
First, Bernstein revised its forecast for the prediction market sector to $240 billion for 2026E on April 14, 2026, with consensus across sell-side firms now pointing toward a $1 trillion market by 2030.
Second, Kalshi and Polymarket combined YTD trading volume surpassed $60 billion in mid-April 2026—already exceeding the sector’s full-year 2025 total of $51 billion.
Third, Robinhood launched over 1,000 Kalshi contracts, with its 1M+ customers having traded 9 billion contracts cumulatively. Robinhood’s prediction market business generates ~$350 million ARR, up from $150 million in 2025 and projected at $586 million for 2026E—the company’s fastest-growing product line.
Collectively, these figures point to one conclusion: prediction markets are no longer a crypto-native niche—they behave more like a TradFi distribution problem. The “retail investors being extracted from,” assumed in the bot narrative, are not primarily crypto natives, but retail investors entering via traditional broker channels.
Hence, the contextual bias in bot panic: the sector isn’t being value-extracted by automation—it’s being flooded with traffic by mainstream finance at a pace far exceeding any automated extraction speed.
II. The Real Data That Matters: 37% vs. 10%
The most frequently cited data point in the bot narrative suffers from sample selection bias.
The claim that “14 of the top 20 profit earners are bots” relies on a small, pre-sorted sample ranked by profitability. This sample reveals only how bots occupy the right tail of the distribution—not whether bots outperform humans at the population level.
Population-level data (sources: Polystrat / Valory disclosures, cross-validated with multiple on-chain Polymarket analytics):
A 3–4× win-rate differential at the population level is genuine evidence of bot structural advantage. The 14/20 statistic should be interpreted as a downstream manifestation of that win-rate distribution—not as standalone causal evidence.
III. Where Bots Win
Bot extraction is highly concentrated in three market types. All three share a key trait: none require subjective judgment about real-world outcomes—instead, they rely on latency or pricing advantages relative to the platform’s matching engine.
Price-feed latency arbitrage
Representative case: wallet 0x8dxd, which turned $313 into $437,600 in just 15 minutes trading BTC up/down binary contracts in January 2026, achieving a 98% win rate.
Strategy logic: monitor spot prices on Binance and Coinbase, then enter positions on Polymarket when its quotes lag behind CEX prices. Polymarket introduced a taker fee (peaking near 3% for ~50% of trades) for 15-minute crypto contracts on January 7, 2026, specifically to neutralize this strategy. The wallet’s cumulative win rate has since fallen to 54.7%.
Conclusion: Bot advantage in price-feed markets is real—but confined to an extremely narrow time window, and significantly compressed as platforms introduce friction costs.
Automated betting-state execution in live sports
Data source: cancun2026 team’s Polymarket wallet classification (Dune query 6648075, https://dune.com/queries/6648075, last 7 days, through May 11, 2026).
Source of advantage: bots react to in-game events significantly faster than retail users relying on live streams (30-second latency). Additionally, trading terminals such as Kreo and PolyCop democratize this edge via copy-trade and auto-follow features—meaning measured bot share includes human capital routed through bots.
Cross-platform combinatorial arbitrage
Data source: IMDEA Networks paper “Unravelling the Probabilistic Forest: Arbitrage in Prediction Markets” (AFT 2025, dspace.networks.imdea.org/handle/20.500.12761/1941).
The study covers ~$40 million in arbitrage extraction on Polymarket between April 2024 and April 2025, driven by two primary patterns: (i) rebalancing YES/NO shares within the same market; and (ii) cross-platform combo trades (e.g., buying YES on Polymarket and NO on Kalshi when their implied probabilities sum to less than $1). This pattern requires rigid multi-platform infrastructure—and shrinks as matching engines across venues converge.
IV. Where Human Accounts Win—and Their Constraints
The lowest bot share occurs not because “retail picks better”—but because profitability in those markets hinges on synthesizing multi-source real-world information—a domain where automation remains structurally disadvantaged versus humans.
Two independent studies corroborate this.
Joshua Della Vedova (University of San Diego)’s on-chain behavior research (jdellavedova.com) finds: retail users select winning outcomes at a higher frequency than bots; bots’ edge lies in execution—e.g., entering at $0.55 when retail buys YES at $0.72, locking in $0.17 per share in unrealized gain.
A working paper by University of Toronto / HEC Montréal / ESSEC (Akey et al., SSRN 6443103, March 18, 2026) shows: 56% of losing users place orders in extreme price ranges (<10¢ or >90¢), whereas only 28% of the top 0.1% profitable users do so. Losing users typically “chase 5¢ odds for 20× payouts” or “buy at 95¢ for late-stage certainty”; profitable users build positions in the middle of the probability curve.
Together, these studies indicate: retail judgment is broadly underestimated—but timing execution and order structure are systematically suboptimal.
V. Forward Path: Four Forces Shaping the Bot/Human Dynamic
Over the next 12–24 months, the critical variable is not today’s bot/human share—but its trajectory. This report identifies four countervailing forces.
Further collapse in bot deployment cost
Coding agents like Claude Code and Codex, open-source frameworks like Hermes, and Polymarket’s own MIT-licensed Polymarket Agents framework have collectively lowered the engineering bar for strategies like 0x8dxd—from “serious project” to “weekend prototype.” Copy-trade services further plug human capital into bot infrastructure, mechanically inflating measured bot share.
Peer-on-peer cannibalization of bot unit economics
The 823 profitable bot wallets represent the right tail of a much larger cohort of unprofitable bots. As the number of wallets running similar strategies rises, each bot’s profit window narrows. 0x8dxd’s 98% win rate is structurally non-replicable—not because inefficiency vanished, but due to peer competition + platform fee adjustments. Capital extraction from humans by bots will likely peak *before* bot share does.
Platform category composition tilting toward retail
Polymarket’s April 2026 category mix: Sports 50%, Crypto 24%, Politics 16%, Others 10%. In April 2025: Sports 29%, Crypto 12%, Politics 42%.
Sports trading volume grew 11× YoY in absolute terms. New volume flows overwhelmingly into long-horizon event markets—where retail dominates. Bernstein forecasts sports’ share of sector-wide volume will decline from 62% today to 31% by 2030, replaced by economics-, politics-, and corporate-event contracts—further expanding the category exposure where bots hold no advantage.
Natural category-based fragmentation across platforms
Hyperliquid’s HIP-4 launched on May 2, 2026, offering daily BTC binary contracts, zero open-position fees, USDH collateral, unified perpetual/spot settlement, and validator-slashable market deployment (1M HYPE per slot, ~$42.76M at current price).
This is a classic bot-advantaged market type deliberately spun off onto a dedicated platform. Day-1 volume came predominantly from arbitrage capital—consistent with historical BTC binary contract distribution. If HIP-4 later expands into sports/politics markets and integrates trusted oracles, its bot share may converge toward Polymarket’s. For now, its role is to isolate bot-friendly traffic onto a separate platform—accelerating Polymarket’s drift toward retail-centric categories.
VI. Platform Landscape & Valuation Snapshot (Mid-2026)
▲ Source: Bernstein note (April 14, 2026), Polymarket / Kalshi public disclosures, HIP-4 launch announcement
Conclusion: Kalshi + Polymarket hold >97% combined share. Platform-layer equity opportunities are effectively closed for venture-scale checks. Investable value is shifting both upward (trading terminals, quant strategy services, agent infrastructure) and downward (capital efficiency, arbitration, oracles).
VII. Risk Disclosures
Risk 1: Regulatory tail risk. Three bills submitted by Schiff (the DEATH BETS Act, Public Integrity Act, and Prediction Markets Are Gambling Act), Nevada’s TRO against Kalshi, and Arizona’s March 2026 criminal charges create federal/state jurisdictional friction. Kalshi’s 89% revenue concentration in sports exposes it most acutely—sports or war/death-themed contracts face realistic tail risk of full-category bans.
Risk 2: Oracle & arbitration failure risk. Polymarket integrated Chainlink for price-based markets in 2025—but subjective markets still rely on UMA. UMA’s token economy currently generates only ~$600k annual economic flow against a $37M FDV; post-MOOV2, proposer rewards are restricted to ~37 whitelisted addresses—most affiliated with Polymarket. Any high-profile contested ruling could trigger broad re-evaluation of trust across the sector.
Risk 3: Sports share reversal risk. Polymarket’s 2026 sports growth is seasonal (driven by NBA, NFL Super Bowl). If sports share retreats, the dynamic of “rising bot share + expanding retail” could reverse.
VIII. Implications for Builders & Investors
The bot debate is ultimately about one question: Within Bernstein’s $240B 2026 prediction market forecast, which layer captures value? Across four layers, value density varies sharply.
L1 — Agent trading products. Strategy advantage is decaying; C-end automated trading bears compliance risk. Not recommended as a standalone bet.
L2 — Agent infrastructure (Olas / Valory models). Toll-road economics—fee capture regardless of which agent wins. This is the cleanest investable option.
L3 — AI-native data, pricing, market creation. Most is absorbed internally by platforms—or captured by incumbent Web2 players (Kensho, Bloomberg, Dataminr). Remaining investable windows are narrow.
L4 — Arbitration & resolution. Economic flow is real but small-scale today. To become a Tier 1 venture asset, a redesigned token model is required—but no such redesign appears on the public roadmap.
Emerging edges worth tracking:
- PM-DeFi composability (e.g., Morpho collateralizing PM positions—currently 2x leverage, roadmap targets 4–5x, impacting capital efficiency)
- Trading terminals & copy-trade services (e.g., Kreo)
- PM-native quant firms
- New market primitives (impact markets, futarchy, conditional markets)
Conclusion: Bots Win Categories, Humans Win Markets, Platforms Win Structure
Bots have not taken over prediction markets. They’ve saturated specific market types—and any platform’s bot-to-human trading volume ratio is fundamentally a downstream reflection of its market-type composition. The headline “5% wallets / 75% volume” conflates volume concentration with capital extraction. Polymarket’s 2026 growth is driven largely by sports markets—where bots hold no structural advantage—while the $131M in bot profits was extracted primarily in short-window crypto markets where retail participation is low.
Future-winning platforms must deliver three capabilities: credible arbitration enabling diverse market types; balanced capacity to host both bot and human traffic at appropriate ratios; and cross-category user retention. Polymarket currently occupies this position: Bitget’s Q1 2026 research shows organic multi-category user growth—average categories per user rose from 1.45 to 2.34, and active days per user from 2.5 to 9.9.
Bots stay within their structural advantage zones; human capital running bots keeps migrating to the next event; and the ultimate winner is the platform that best hosts both types of traffic, across the widest range of market types, at optimal proportions.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
IOSG Ventures
