“A study on NFL bettors from the 2023 - 2024 season [found that] 87% of bettors accounted for a total of 1% of sportsbook revenue, which means that the vast majority of bettors are able to play safely. Then there's that other percent of players, like 3% of players that account for like 80%+ of revenue. Some of them are really rich and good for them - they can spend their money however they want. Some of them are not really rich and their dopamine pathways are rewired and they're addicted. (source)”

The house always wins, even when a bettor finds an edge, operators limit them or close accounts. And those who keep losing end up having their funds siphoned.

Crypto fixes this (?)

Onchain systems can bring transparency, global liquidity, and even guardrails, like proof-of-funds that enforce bankroll limits, or reverse leaderboards to help those that are in need.

Liquidity provisioning can position anyone to “be the house” and earn on trading fees.

Smart AMMs and odds algos can ensure that even is someone finds an edge, toxic flows are limited without the need of banning accounts.

How can crypto make it more efficient?

Today we’ll cover:

1. The scale of sports betting

2. Web2 and Web2.5 players

3. Prediction Markets

4. Web3 onchain players

5. What it takes to win

Subscribe

Yearly, nearly 1 in 6 adults, over 882 million people, have gambled online. Sports betting is a cultural fixture and a financial giant.

In 2024 the global sports betting industry:

• Generated around $100 billion (~10% YoY) in revenue, and it’s projected to nearly double by 2030.

• Europe has long been the largest market (~35%) due to mature regulations and sports culture, while the U.S. is rapidly catching up after the 2018 PASPA repeal legalized state-by-state sports wagering.

• From mid-2018 to 2025, Americans wagered over $500 billion legally on sports, yielding nearly $50 billion in revenue for sportsbooks. The U.S. surge is 23.7% and on track to lead the world by 2030.

• The catalyst for growth in the US was the 2018 regulation, where the federal ban for gambling was revoked. Today, platforms need a state-by-state DFS licensing, or if they’re prediction markets they need CFTC clearance, which both Kalshi and Polymarket have, allowing for a sort of regulatory arbitrage.

Sportsbooks have evolved from physical betting shops and casinos to mobile-first experiences. Today online betting is the dominant channel, making up ~78% of revenue. Technology has enabled features like:

• In-play (live) betting

• Props: Micro-betting on every play

• Parlays

• Fantasy sports

The cultural acceptance of betting has grown, major sports leagues (NFL, NBA, English Premier League, etc.) now partner with betting companies, and media outlets frequently cite odds on games and elections alongside polls.

Football is by far the most bet-upon sport globally, it consistently holds the largest market share ~24%. The runner-up is horse racing. Other sports include basketball, NFL, tennis, cricket, and baseball. Even esports betting is emerging fast, projected to be the fastest-growing segment (>14% CAGR) as younger generations bet on video game competitions.

To appreciate the scale: leading sportsbook operators handle tens of billions in bets annually. Companies like Flutter Entertainment (owner of FanDuel, Betfair, etc.) and Entain (Ladbrokes, part of BetMGM JV) are multi-billion dollar enterprises. In the US, DraftKings alone generated $4.7–$4.8 billion in revenue in 2024, and the combined “handle” across U.S. operators was ~$500B within 7 years. This dwarfs the volumes seen on newer crypto prediction markets, underscoring both the potential and the challenge for the next generation of platforms.

For founders, sports betting matters because it is one of the rare consumer categories with proven TAM and recurring demand. The incumbents are massive, but they are centralized, expensive, and constrained.

The incumbents dominate through licenses, brand equity, and relentless marketing. Their scale is staggering:

• Flutter Entertainment (FanDuel, Betfair, Paddy Power): 55B market cap. $11.7B in 2024 revenue, 50% U.S. share through FanDuel, 12M MAUs, $162M net profit. Differentiates by running both the #1 U.S. sportsbook and Betfair’s global betting exchange, scale, diversification, and marketing power make it the industry leader.

• DraftKings: 25B market cap. $4.8B revenue in 2024, ~35% U.S. share. $507M net loss, though first full-year positive adjusted EBITDA signals an improving trajectory. Differentiates through daily fantasy roots, product innovation, and aggressive promotions in U.S. sports.

• Entain (Ladbrokes, Coral, BetMGM JV): $6B+ annual revenue. -$461M net loss in 2024 due to one-offs, but operating EBITDA was positive at $790M. Differentiates via omnichannel retail + online integration and its U.S. BetMGM joint venture.

• Bet365: ~$4B revenue (privately held), dominant in Europe. Returned to profit in 2024 with ~$146M operating profit and ~ $800M pre-tax profit. Differentiates with the deepest in-play markets and pioneering live-betting UX, built entirely in-house.

These Web2 giants protect margins by limiting winners, restricting high-volume bettors, and sticking to regulated geographies. Flutter leads the pack because it combines scale across regions, the #1 U.S. brand (FanDuel), and diversified product lines, a portfolio approach that no competitor matches.

The Web2.5 operators show what happens when crypto rails are bolted on:

• Stake.com: $4.7B in 2024 revenue, crypto-only, BTC/ETH/USDT. Privately held, estimates suggest healthy margins (20–25% EBITDA) given low compliance overhead. Differentiates with huge limits, instant crypto payouts, and global sponsorships (Premier League, Drake).

• Sportsbet.io: BTC/USDT betting. Profitability not disclosed, but estimated in hundreds of millions. Differentiates with fast cashouts, live-streaming, and a user-friendly UX blending fiat and crypto.

• BC.Game, Cloudbet, BetOnline: Casino + sportsbook hybrids. Significant but opaque profitability, relying on casino margins; likely in the hundreds of millions annually. Differentiate by serving crypto users with speed, privacy, and wide stablecoin support.

Great insight from @simononchain

• It’s important to mention both Rollbit (135M FDV) and Shuffle (300M FDV), while they are crypto casinos, they do offer sports betting as well, and are ranked on the top revenue generating protocols in crypto overall. Both are centralized and custodial.

These Web2.5 players are custodial and centralized, but their traction proves crypto demand is real, especially stablecoins as the preferred betting rail.

Prediction markets are not sportsbooks per se, but they are expanding into sports and very quickly gaining traction. 5 of the top 20 active markets ranked by volume are sports related, and the top market is the F1 champion. This signals a huge demand for more sports betting.

• Kalshi: CFTC-regulated, raised at a $2B valuation in 2025. 2B in sports betting volume for H1 2025. 70% of the volume is sports related. 240M revenue annualized. ~100x smaller scale than FanDuel.

• Polymarket: Crypto-native on Polygon, and abstracting crypto away. A lot of its volume is sports related, but not as consistent as Kalshi’s, with ~40% being sports related. No revenue yet, but would be ~200M with a similar fee structure to Kalshi.

Kalshi proves compliance can scale, and are crafting tight partnerships with Robinhood and X. Polymarket proves raw crypto-native appetite. Both are expanding into sports even more, alongside politics, economics, and culture.

The defining difference of Web3 sportsbooks with web2 counterparts or prediction markets is their permissionless and onchain design. Odds are algorithmic or oracle-based, market making is open to anyone, governance can be DAO driven, and many protocols expose composable building blocks for others to build on. This creates a betting ecosystem that looks less like a closed bookie and more like an open financial primitive. Among onchain sportsbooks, three stand out by traction: Overtime, Azuro, and SXBet.

In preparation for this article, I’ve used all three products to bet on the Premier League Liverpool vs. Arsenal match. On Overtime I had 3.33x odds for singe match, and if I had hit my parlay including Machester City and Newcastle winning, I had a 12.8x. SX bet had a 3.38x, and the parlay was 13.6x. Azuro (via Bookmaker) paid 3.25x and 12.74x on the parlay. Polymarket had 3.33x, and Kalshi had 3.35x, none parlays.

Basically, all had very similar odds, and offered a very similar user experience. The Polymarket UX is way superior than the others. On Azuro I could do prop bets on things like goals, or corners. But on none I could do complex prop bets like MVP, number of goals per player, saves, and many others, which is common on trad web2 sportsbooks. Overtime does offer prop bets but for NFL games only, and Kalshi

Metrics wise, Overtime is ahead of the crowd, and undervalued compared to the others. The last couple of weeks have even seen more growth, nearing a 2x revenue multiple. The key question is the stickiness and growth potential of this revenue.

Overtime

• Data: Dune dashboard

  • $309M+ cumulative volume, ~30M last mo, 10M last week

  • 250 DAUs, 38k bets, $790 avg/bet

  • 1.5M in fees YTD, 200k last mo

  • Team burn is around ~120k/mo according to their discord

• Features: Overtime runs a parimutuel/AMM hybrid model, it offers “SportsAMM” liquidity pools that take the opposite side of bets with automated per-market single side liquidity rebalancing based on demand.

  • Users can provide USDC or ETH liquidity to collateralize the platform and be exposed to the "house" performance.

  • Users can place straight bets, parlays (multi-bets), same-game parlays (correlated combos), futures (long-term outrights), and even live in-game bets.

  • The platform covers over 100 sports and leagues worldwide, from major leagues (NFL, NBA, soccer leagues, MMA, etc.) to esports (it recently boasted being the #1 on-chain book for esports props).

  • Overtime is arguably the most user-friendly decentralized sportsbook to date, thanks to deliberate design choices. With integrating Particle's Chain Abstraction SDK, users can fund their Overtime Accounts seamlessly from any network, even Solana.

  • Users can deposit funds (crypto or via credit card fiat on-ramp) to their account’s address and start betting.

• Go to market: Overtime has been creative in building a community and engaging users. One pillar is “Overdrop”, a loyalty rewards program where users complete betting challenges to earn points, free bets, and even token airdrops. Overtime also doesn’t implement the negative social aspects of traditional books. The platform proudly advertises “no limits, no bans, no withheld payouts”, which has resonated with sharp bettors who often get limited by CeFi sportsbooks.

• Licensing & Legal: Overtime takes a DeFi-centric stance, it is a fully non-custodial, on-chain protocol and does not do KYC or restrict users on the dApp level.

• Differentiators:

  • The closest to DraftKings in UX but without limits on sharp bettors

  • Onchain scalability: Their custom merkle tree design that pulls sports data from chainlink nodes, allowing Overtime to scale its offering and odds updates to thousands of different sports, leagues, bet types and markets each day, without getting performance bottlenecks.

• Token: OVER. Market cap ~$15 million USD, nearly fully circulating. $OVER is the core betting collateral, providing the best odds and user experience. 100% of Overtime's fees and revenue fuel a buyback and burn program, reflecting platforms growth to all token holders.

• Roadmap:

  • Yield bearing assets as collateral

  • P2P features, such as permissionless Request-For-Fill orders

  • Full Chain Abstraction UX

Azuro Protocol

• Data: Dune dashboard

  • $225M+ cumulative, 2.5M last m

  • ~1.3k MAU, ~150 DAU, 25k bets, $100 avg. bet

  • Negative revenue for the past 6 months, and 73k+ in revenue last month: they had issues with arbitragers that siphoned the fees out of the system which they solved last month (link)

  • Liquidity Providers: 4,400+

• Features: Decentralized betting protocol rather than a single consumer app. It provides the “predictions layer” for Web3, allowing any front-end or operator to tap into Azuro’s smart contracts for liquidity, odds, and settlement.

  • The protocol’s core innovation is a peer-to-pool model with an automated odds engine. Liquidity providers contribute to large betting pools (liquidity vaults) that act as the counterparty to bettors, similar to an AMM (automated market maker) for sports outcomes

  • Azuro supports a wide variety of sports and events: football (soccer) dominates volume (~69% of bets), followed by basketball (~19%)

  • The protocol is chain-agnostic and multichain, initially launched on Gnosis Chain, later expanded to Polygon and even Chiliz’s chain (to capture sports fan token ecosystems)

  • Some Azuro-powered sites have integrated features like fiat on-ramps or custodial accounts to simplify onboarding for non-crypto natives.

  • Transaction costs are low (especially on Gnosis Chain) and typically bets settle quickly after matches, with smart contracts paying out winners transparently.

  • Resolution relies on oracles, Azuro either uses its own oracle network or Chainlink feeds.

SX Bet

• Data: Dashboard

  • 240M in the last 12mo. ~20M in the last month

  • ~3k MAU, ~100 DAU, 60k monthly bets, $322 per bet

  • Not charging any fees

• Features: First to bring esports onchain, governance via SX token, P2P style markets, on their L2. This is to become the best place to find the best odds, where bettors can pay higher gas fees in the native token to boost their order, appealing to the prosumers and more professional bettors.

  • It functions similarly to a Betfair-style exchange, users back or lay bets against one another, with SX acting as the matching engine via smart contracts. By eliminating the traditional “house,” SX Bet dramatically reduces the vig (fee), targeting a ~0.25% commission versus ~5% industry. All bets and market resolutions are recorded on-chain for transparency.

  • SX offers markets on major sports like American football (NFL), basketball (NBA), baseball (MLB), hockey (NHL), plus soccer and niche events (esports, politics). This broad selection makes SX akin to a hybrid of sportsbook and prediction market. In practice, users can wager in USDC (primary), as well as ETH or the native SX token, with peer-to-peer odds often more favorable than traditional bookies.

  • In mid-2023 SX introduced the first on-chain parlays system, showing that even complex bet types can be done P2P. Users do need a Web3 wallet and must bridge into SX Network (an Arbitrum-derived L2).

  • To simplify onboarding, SX built a gasless bridge that lets users move various tokens into SX Network. Still, compared to Web2 sportsbooks, there is a learning curve in managing a crypto wallet and bridging funds. There is no native mobile app in app stores, but the web app is mobile-responsive.

  • SX has also leveraged referral programs and an open API to encourage affiliates and quants to plug into the exchange.

Beyond the top three leaders, dozens of projects are experimenting with onchain sportsbooks, exchanges, fantasy hybrids, and betting infrastructure. Collectively they illustrate the breadth of the design space:

• Football.fun: Fantasy + DeFi on Base. Hit ~$100M TVL in weeks post-launch in 2025. Tokenizes football player performance.
  

• Sorare: Global NFT fantasy sports (soccer, NBA, MLB). 3M+ users, ~$200M yearly card volume. More fantasy than sportsbook, but significant.
  

• BRACKY: In feed ai agent betting on Farcaster.
  

• BetDEX: Solana-based betting exchange, licensed in Isle of Man. Founded by ex-FanDuel execs. ~$0.5–1M monthly volume. Regulatory-first approach.
  

• Onit: Prediction market focused on sports, leveraging miniapps on Base and Farcaster as distribution.
  

• Dexsport: Multi-chain sportsbook and casino with tokenized incentives (DESU).
  

• Betswap.gg: Hybrid sportsbook + P2P exchange. Lets users both bet and set odds.
  

• LEVRBet: Leveraged sports betting for prosumers on Monad.
  

• Hedgehog Markets: Solana no-loss prediction/betting pools, using yield from DeFi to pay out winners.
  

• Aver Exchange: Solana betting exchange focused on peer-to-peer orderbooks. Modest liquidity to date.
  

• Wagerr: Early decentralized sportsbook with its own blockchain. Once promising, now low traction.
  

• PeerBet: Early peer-to-peer betting dapp, niche adoption.
  

The incumbents have set the bar: instant payouts, live odds, slick mobile apps, deep liquidity. To compete, onchain platforms must match that baseline while offering what Web2 cannot: transparent odds, composable liquidity, global reach, and no arbitrary limits.

The winning products will deliver:

1. Amazing UX: Crypto fully abstracted, easy to log in, great mobile experience, haptics, everything you’d expect from a Robinhood or Duolingo like product experience.
   

2. Protocol approach: Going onchain means anyone can spin up a frontend and tap into the underlying odds and liquidity model, and earn trading fees. I want a football focused client. That said, companies should start by building out the best client they can themselves, before outsourcing their distribution to third party devs.
   

3. Prop bets: These are the highest margin products, and Kalshi is the first massive mover in the category. I want to bet on the top goalscorer for the Copa Libertadores final.
   

4. Parlays: Allowing for users to bet on low probability, high value outcome bets. This is effectively giving the chance of fans to dream and become millionaires of a single bet. It’s hard to do, given that creating an onchain risk model that “cancels” out probabilities and actually is able to manage the liquidity and odds to avoid getting drained by a single low probability winner. Adhi from Kalshi has a great article on this.
   

5. Managing toxic flow: Similar to the parlays, odds arbitraging is real, and having a model that allows for low latency, and a smart way to adjust odds based on the liquidity model provided, so the house is minimizing the risk of getting siphoned due to market arbitrages.
   

6. Decentralizing liquidity: Even though order books are models that seem to work for prediction markets and others, smart liquidity pool models may be more efficient at capital allocation and risk management. Web2 bookies usually post the liquidity/collateral themselves, and manage it with complex risk models. This could be brought onchain, and external LPs could be a “part of the house”.

Sports betting is a massive TAM, and crypto founders are rewriting its mechanics. Overtime, Azuro, and SX Bet show different archetypes. The next wave will determine whether the future DraftKings is Web2.5 custodial, or truly onchain.

We’re exploring this market actively, and happy to chat with anyone building or interested in the space.

This was done in conjunction with JohnGall from PTC, providing great thoughts and examples. Also, thanks to mikey from 1kx for thoughts, Newsletter reads from

Adhi Rajaprabhakaran

, simononchain tweets and market research. Intros by Dan from Nascent led to conversations with the Sxbet and Overtime teams to clarify their models.

Subscribe

This post is for informational purposes only, and does not constitute a recommendation to buy or sell securities or to pursue any particular investment strategy. This post should not be relied upon in evaluating the merits of any investment or any particular investment strategy. You should consult your own advisers as to business, financial, tax, legal, and all other related matters concerning any investment. The views expressed in this post reflect the current opinions of the authors and do not necessarily represent the opinions of Social Graph Ventures LLC