Prediction markets vs sports betting: the structural differences that matter
They look similar on the screen. The economics, regulation, and unit profile are different in ways that change how each business scales.
Note, November 15, 2026: this analysis was written in May 2026. Figures in it are estimates that we have not traced to a primary source, and some details have changed since. For current rules and status, see our regulation news and state table.
TL;DR:
A prediction market is a peer-to-peer exchange regulated (in the US) under federal derivatives law. A sportsbook is a dealer market regulated under state gambling law. The difference shapes margin structure, customer acquisition cost, jurisdictional reach, and how each business is valued in the public markets.
- Sportsbooks earn from the spread (vig); exchanges earn flat fees on volume.
- Sportsbook GGR margins run 7–10%. Exchange take rates are typically 1–4%.
- Sportsbooks are licensed state-by-state. Prediction markets argue for national, federal-preemptive reach.
- Customer acquisition cost is lower for exchanges because there's no state-by-state geofence rebuild.
- Investors should not value a prediction-market platform on sportsbook comps — the unit economics differ.
The five differences that actually matter
On the surface, the two products look almost identical. You pick an outcome, you stake money, you get paid if you're right. The structural differences sit underneath the UI, and they are what determine which business model will compound over the next decade.
1. Margin structure
A sportsbook is a dealer. It sets the odds slightly worse than the true probability, takes the bet, and on average keeps a percentage of every dollar wagered. Industry-wide, that hold rate sits between 7% and 10% of handle in the US (slightly lower in mature European markets). Revenue scales with handle, but the operator is exposed to bad outcomes on any given day — a heavily favoured team losing can cost a book millions in a single afternoon.
A prediction market is a broker. It earns a take rate — typically 1% to 4% on each trade — and never holds inventory in the contract itself. The book's revenue is a function of volume, not outcomes. There are no "bad weekends" for an exchange in the way there are for a sportsbook.
Implication: at the same handle, a sportsbook generates more revenue per dollar wagered. At the same revenue, an exchange does it with less volatility and less capital tied up against risk.
2. Liquidity dynamics
Sportsbook liquidity is essentially infinite — the house will take any reasonable bet because it manages risk across the whole book. Exchange liquidity is conditional: it exists only where another user is willing to take the opposite side.
This is the single biggest product weakness of prediction markets today. NEXT.io's Great Divide report (March 2026) identifies thin liquidity in long-tail markets as the structural constraint on category growth. Big political markets and major NFL games trade like real exchanges. Tuesday-night college basketball does not.
The first platform to solve liquidity at scale — through market-maker incentives, dark pools, or guaranteed-fill mechanics — wins the category.
3. Regulatory perimeter
In the US, sportsbooks are licensed state-by-state. Each license takes 12–24 months, costs millions in fees and integrations, and binds the operator to that state's tax regime (anywhere from 6.75% in Iowa to 51% in New York). To operate nationally, a sportsbook needs ~30 separate licences.
Kalshi, as a CFTC-registered Designated Contract Market, argues that its event contracts are federally regulated derivatives and therefore preempt state gambling law. If that position holds in court, a single federal registration grants access to all 50 states. That is a different business.
The litigation is unresolved. But the upside case for prediction markets is a national footprint without state-by-state taxation. For more, see the event-contract legal battles.
4. Customer acquisition
Sportsbook CAC in the US has been brutal. Public filings from DraftKings and FanDuel show acquisition costs of $300–$700 per depositing customer in newly opened states, driven by sign-up bonuses, free bets, and brand-marketing spend.
Prediction-market platforms have so far acquired users at materially lower cost. Polymarket's growth has been substantially organic and earned-media driven — financial press coverage of election markets, crypto-native word-of-mouth, social distribution of viral market screenshots. Kalshi's sports launch leveraged content partnerships and creator integrations rather than scorched-earth bonus offers.
The longer that gap holds, the more attractive the unit economics look to investors comparing the two categories.
5. Product breadth
A sportsbook offers sports. A prediction market offers anything that can be unambiguously resolved: elections, Fed decisions, weather, box-office openings, who wins the Oscars, whether OpenAI ships a new model before December.
That breadth has two effects. First, it diversifies revenue away from sports seasonality. Second, it creates a discovery problem the sportsbook category doesn't have — users need to be educated on what they can trade. This is a content and SEO opportunity that sports-betting affiliates have not had to solve.
The valuation question
Public-market sports-betting operators (DraftKings, Flutter, Entain, Penn) trade at EV/Revenue multiples in the 2–4x range, reflecting low margins, high capex, and concentrated state-tax exposure.
Prediction-market venues, if they're valued correctly, should trade closer to exchange comps — CME, ICE, Nasdaq — at 6–12x EV/Revenue, reflecting capital-light operations, recurring fee revenue, and federal-preemptive distribution.
The investor mistake to avoid is anchoring a prediction-market platform to a sportsbook multiple. They are different businesses.
Where the affiliate sits
Affiliate-media properties earn revenue from both categories: a CPA or revenue-share fee from the sportsbook, and a similar arrangement from the prediction-market venue. From an affiliate's standpoint, the two products are complements, not substitutes.
The opportunity is to be the trusted educator that explains the difference, ranks the venues, and routes the user to whichever product fits their use case. That is the playbook we describe in unit economics of a prediction-market affiliate site.
In regulated European markets — Sweden (SGA), Malta (MGA), the UK (UKGC) — exchanges and sportsbooks have coexisted under unified gambling regulators for two decades. The market split there is instructive: exchanges command a meaningful but minority share (5–15% depending on jurisdiction), concentrated in sophisticated bettors. Casual users still prefer sportsbooks. If the US category settles in the same way, prediction markets will be additive to sports betting — not a replacement.
#category
#explainer
Related