Are Prediction Markets Gambling? The Math Says It Depends on You

The Question Everyone Asks Wrong

Are prediction markets gambling? The question gets asked constantly. Politicians use it to justify regulation. Traders use it to justify tax treatment. The media uses it as a framing device. Almost everyone asking the question has already decided the answer before doing any math.

Here is the quantitative reality: prediction markets, sportsbooks, and stock options all involve risking capital on uncertain outcomes. The mechanics are structurally similar. The regulatory labels are different. And the math does not care what you call it.

The useful version of the question is not "are prediction markets gambling?" It is: "What separates a gambler from a trader?" The answer is the same across every risk market. It comes down to three things: expected value calculation, position sizing discipline, and systematic edge verification. If you do all three, the regulators call you a trader. If you skip all three, you are gambling. The platform is irrelevant.

The Structural Comparison: 3 Markets, 1 Framework

Consider three platforms where you can risk money on a binary outcome.

Sportsbook. You bet $110 at -110 odds on an NFL game. If you win, you profit $100. If you lose, you are out $110. The sportsbook bakes a 4.8% margin into the odds.

Prediction market. You buy a Kalshi contract at $0.55 on the same NFL game. If you win, you receive $1.00 and profit $0.45 (minus approximately 7% fee on winnings). If you lose, you are out $0.55.

Stock option. You buy a call option for $5.50 on a stock trading at $100 with a strike price of $105. If the stock goes above $110.50 by expiration, you profit. If it stays below $105, you lose $5.50.

All three are binary-outcome wagers on uncertain events. All three have explicit or implicit fees. All three attract participants ranging from disciplined quantitative traders to people following gut instinct. These are all risk markets. The differences are regulatory, not mathematical.

The math framework that determines profitability is identical across all three:

EV = (Probability of Win x Net Profit) - (Probability of Loss x Amount Risked)

Run every trade through this formula. If the result is positive, the trade makes money over time. If negative, it loses money over time. Whether you call it a "bet," a "trade," or a "position" does not change the output. The EV calculator works on any of these instruments.

The Loss Rate Data: 91%, 80%, and 95%

The numbers tell a consistent story across every risk market.

• 91% of Polymarket traders lose money (based on published wallet-level analysis)
• 80% of retail options traders lose money (SEC and academic research)
• 95% of sports bettors lose money (industry estimates)

These numbers look like evidence that all three are "gambling." They actually reveal something more specific: most participants in every risk market skip the math. They trade on narrative, emotion, or entertainment value.

The inverse is more interesting. The 9% of Polymarket traders who profit, the 20% of options traders who profit, and the 5% of sports bettors who profit share common traits. They calculate expected value before entering. They size positions based on edge magnitude. They track results over sample sizes large enough to distinguish signal from noise.

The dividing line is not the platform. It is the process. For how to apply that process on prediction markets specifically, the pipeline starts with probability estimation and runs through fee-adjusted EV calculation using the PM EV calculator.

Worked Example: Same Event, Two Platforms

The Super Bowl is priced on both a sportsbook and a prediction market. Here is the same event compared across platforms.

Sportsbook pricing: Team A -110 / Team B -110. Standard vig. Total implied probability: 52.4% + 52.4% = 104.8%. The book holds 4.8%.

Prediction market pricing (Kalshi): Team A Yes at $0.53 / Team A No at $0.49. Total price: $1.02. The 2-cent gap is the market maker's spread.

Suppose you believe Team A wins 55% of the time. Calculate EV on each platform.

Sportsbook at -110:

• Profit if win: $100
• Risk: $110
• EV = (0.55 x $100) - (0.45 x $110) = $55.00 - $49.50 = +$5.50 per $110 wagered
• Edge: 5.0%

Kalshi at $0.53 with 7% winner fee:

• Profit if win: $0.47 x 0.93 = $0.437
• Loss if lose: $0.53
• EV = (0.55 x $0.437) - (0.45 x $0.53) = $0.240 - $0.239 = +$0.002 per contract
• Edge: 0.4%

On this specific event, the sportsbook offers a 5.0% edge versus Kalshi's 0.4% edge. The sportsbook is the better trade after accounting for prediction market fees.

Now flip the scenario. A presidential election is priced at -200/+170 on a sportsbook (implied probability 66.7% and 37%, total 103.7%) and at $0.62 on Kalshi.

Sportsbook at -200:

• De-vigged true probability: approximately 64.3%
• Profit if win: $100
• Risk: $200
• EV = (0.643 x $100) - (0.357 x $200) = $64.30 - $71.40 = -$7.10

The sportsbook price implies the market is accurate. No edge.

Kalshi at $0.62 (you believe true probability is 68%):

• Profit if win: $0.38 x 0.93 = $0.353
• Loss if lose: $0.62
• EV = (0.68 x $0.353) - (0.32 x $0.62) = $0.240 - $0.198 = +$0.042 per contract
• Edge: 6.8%

On the political event, Kalshi offers a real edge because the prediction market participant pool is less efficient than the sportsbook market on sports. This is exactly why platform selection matters. Route each trade to the platform with the best risk-adjusted pricing.

The Regulatory Framework: 4 Different Labels

The same activity receives different legal classifications depending on the platform.

Kalshi: CFTC-regulated event contracts. Kalshi operates as a Designated Contract Market under the Commodity Futures Trading Commission. Event contracts are classified as derivatives, not wagers. The CFTC won a legal battle in 2023 to regulate these markets. Kalshi must follow the same compliance rules as futures exchanges. Gains are reported as capital gains, not gambling income.

Polymarket: Offshore/crypto. Polymarket operates outside US regulatory jurisdiction for US persons (though enforcement is evolving). It uses USDC on Polygon. There is no US regulator directly overseeing the platform. Tax treatment is ambiguous and depends on your tax advisor's interpretation.

DraftKings Pick6 / Sports betting: State gaming commissions. Sports betting is regulated state by state. Classified as gambling in most jurisdictions. Winnings are reported as gambling income on a W-2G. Losses are deductible only as itemized deductions and only up to the amount of gambling winnings.

Robinhood / Options: SEC-regulated securities. Options trading falls under SEC and FINRA regulation. Gains are capital gains. Losses can offset other capital gains and up to $3,000 of ordinary income per year.

The regulatory label matters for one practical reason above all others: tax treatment. Gambling losses have worse deductibility than trading losses. If your prediction market activity is classified as trading, you can deduct losses against capital gains. If classified as gambling, deductibility is limited. The platform you use and how the IRS classifies that platform's instruments directly affects your after-tax return.

A trader who loses $10,000 on Kalshi event contracts and gains $15,000 on other investments pays taxes on $5,000 net capital gains. A bettor who loses $10,000 on a sportsbook and gains $15,000 on investments pays taxes on $15,000 in investment gains (unless they itemize deductions and claim the gambling loss separately). Same dollar amounts. Different tax bill.

For how event contracts compare structurally to options, see event contracts vs options. For the mechanics of how to trade event contracts, the execution starts with market selection and probability estimation.

The Math Is the Dividing Line

The honest answer to "are prediction markets gambling?" is that the math is the same whether you call it gambling or trading. The formula does not care about the label. EV is EV. Probability is probability. Fees are fees.

What separates the 9% who profit from the 91% who lose is not the platform. It is a three-step process:

1. Calculate expected value before every trade. Not after. Not approximately. Run the actual numbers. The PM EV calculator takes a contract price and your probability estimate and outputs fee-adjusted EV in seconds.

2. Size positions based on edge magnitude. A 2% edge and a 10% edge require different position sizes. Overbetting a small edge is how profitable strategies go broke. This is Kelly Criterion territory, and it applies identically to prediction markets, sportsbooks, and options.

3. Verify your edge over meaningful sample sizes. 20 trades tells you nothing. 200 trades starts to show signal. If you are not tracking your results across hundreds of positions, you do not know whether your edge is real or you got lucky.

Anyone doing all three is trading. Anyone skipping all three is gambling. The platform, the regulation, and the label are secondary. This is the core framework behind how prediction markets work, and it applies whether the market is regulated by the CFTC, the SEC, or nobody at all.

State-by-State Access: Where the Gambling Label Hits Hardest

The gambling-versus-trading classification is not academic. It determines which platforms you can use and where.

CFTC-regulated platforms (federal jurisdiction). Kalshi, DraftKings Predictions (via CME), FanDuel Predicts (via CME), Robinhood, Coinbase (front-end for Kalshi), ForecastEx (via Interactive Brokers), and PredictIt all operate under CFTC oversight. Federal regulation means access is not gated by state gambling laws. A trader in a state that bans sports betting can still trade event contracts on Kalshi.

The exceptions are narrow. Some CFTC-regulated platforms voluntarily restrict certain contract types. Kalshi does not offer sports event contracts in all states. But the baseline access is federal, not state-by-state.

FanDuel Predicts operates in all 50 states because its CME Group partnership routes contracts through a federally regulated exchange. This is the broadest access of any prediction market platform in the US. For the full FanDuel breakdown, the fee structure and contract types matter more than availability.

State-regulated sportsbooks. Sports betting is legal in 38 states plus DC as of early 2026. Each state sets its own rules on operators, bet types, and tax rates. If your state does not have legal sports betting, you cannot legally use DraftKings Sportsbook, FanDuel Sportsbook, or any state-licensed book.

Polymarket. Not available to US persons under the platform's terms of service following the 2022 CFTC settlement. Enforcement is blockchain-level, not state-level. For the full legal picture, see is Polymarket legal in the US.

The practical takeaway: if you live in a state without legal sports betting, CFTC-regulated prediction markets are your only legal option for trading on event outcomes. The "gambling" label on sportsbooks creates a state-by-state patchwork. The "derivatives" label on event contracts creates federal access. Same math, different doors. For a side-by-side of all 8 platforms, see the best prediction market comparison.

Why the Label Still Matters

If the math does not care about labels, why does the gambling-versus-trading question matter? Three reasons.

Tax treatment. As outlined above, the classification determines how gains and losses are taxed. This directly affects your net return. A trader paying capital gains rates on Kalshi profits keeps more than a bettor paying gambling income rates on sportsbook winnings, even at identical gross returns.

Access. Gambling is regulated at the state level in the US. Some states ban sports betting. Prediction markets regulated as derivatives operate under federal jurisdiction, which can mean broader access. The regulatory landscape is changing, but platform availability varies by classification.

Self-assessment. This is the practical one. If you call yourself a trader but skip the EV calculation, you are lying to yourself. If you call yourself a gambler but run rigorous probability models and size positions with Kelly, you are selling yourself short. The label you choose should match your process, not your ego.

For the full comparison of how prediction markets differ from sportsbooks in practice, including fee structures, liquidity, and market types, see the sportsbook vs prediction market breakdown. And for a direct comparison of the two largest prediction market platforms, the Kalshi vs Polymarket comparison covers the dimensions that affect your edge.

Use the break-even calculator to find the minimum probability threshold for any contract price, the fee calculator to quantify platform costs, and the arbitrage calculator to check when cross-platform price gaps create guaranteed profit.

Frequently asked questions

Are prediction markets legal in the US?

CFTC-regulated platforms like Kalshi, DraftKings Predictions, FanDuel Predicts, and Robinhood are legal derivatives exchanges accessible across the US under federal jurisdiction. Polymarket is not available to US persons following a 2022 CFTC settlement. Sports betting is regulated state by state. Legal status depends on the platform's regulatory classification.

How are prediction market profits taxed?

Kalshi event contracts are generally treated as capital gains/losses for tax purposes because they are CFTC-regulated derivatives. Sports betting winnings are taxed as gambling income with limited loss deductibility. Polymarket's tax treatment is ambiguous. Consult a tax advisor for your specific situation.

What percentage of prediction market traders lose money?

Approximately 91% of Polymarket traders lose money based on wallet-level analysis. This is comparable to other risk markets: roughly 80% of retail options traders and 95% of sports bettors lose money. The loss rates reflect the percentage of participants who trade without a quantitative framework.

Is trading on Kalshi considered gambling?

Legally, no. Kalshi is regulated by the CFTC as a derivatives exchange, not a gambling platform. Event contracts are classified as derivatives. However, if you trade without calculating expected value and sizing positions based on edge, the mathematical outcome is the same as gambling regardless of the legal classification.

What is the difference between prediction markets and sports betting?

The core math is identical. The differences are structural: prediction markets use order books with tradeable contracts, sportsbooks offer fixed odds against the house. Prediction markets cover politics and economics, sportsbooks focus on sports. Fee structures differ. And the regulatory classification (derivatives vs gambling) affects tax treatment.