Arbitrage betting — sometimes called "arbing" or "sure betting" — is the strategy of placing bets on every possible outcome of an event across multiple sportsbooks, structured so that no matter who wins, you come out ahead. In theory, it's risk-free profit. In practice, it works less well than the math suggests, and the reasons matter.
This article walks through what arbitrage is, the math behind it, and the structural and operational barriers that prevent most bettors from making real money with it.
The simple theory
Different sportsbooks set odds independently. Sometimes their odds on the same event diverge enough that betting both sides at different books produces a guaranteed profit.
The textbook example. A tennis match between two players. Sportsbook A offers Player 1 at +120. Sportsbook B offers Player 2 at +110. The implied probabilities:
Player 1 at +120: 45.5% Player 2 at +110: 47.6%
Total: 93.1%.
Because the total is less than 100%, the market across the two books is mispriced — there's no margin built in for the sportsbook to take. By splitting your stake between the two books in the right proportions, you can guarantee a profit regardless of which player wins.
If you bet $475 on Player 1 at +120 and $476 on Player 2 at +110:
If Player 1 wins: $570 profit minus $476 loss = +$94. If Player 2 wins: $524 profit minus $475 loss = +$49.
Both outcomes are profitable. The bettor has captured the mispricing across the two books. That's arbitrage.
What the theory misses
In a vacuum, arbitrage is unbeatable math. In actual sportsbook operations, four structural barriers usually eat the theoretical profit.
Sportsbooks ban arbitrage bettors aggressively. Major U.S. sportsbooks — DraftKings, FanDuel, BetMGM, Caesars, ESPN BET — explicitly prohibit arbitrage in their terms of service. Most state their services are intended for "recreational bettors," which is regulatory code for "we will close any account that consistently bets profitably."
The mechanisms books use to detect arbitrage are mature. They look for: unusual bet sizes (the precise amounts arbitrage requires), bets only on +EV markets, large deposits followed by quick withdrawals, betting on obscure sports where arbs are more common, and patterns of betting one side of a market right after another known arb-prone book moves a line.
A bettor flagged as an arber typically gets one of three responses: a limit reduction (you can now bet $5 max instead of $5,000), a soft ban (account allowed to bet but withdrawals delayed indefinitely), or a hard close. None of these are appealable in practice.
Lines move faster than you can act. A true arbitrage opportunity exists because one book has a stale line. The window before either book corrects is typically 15-90 seconds. To execute, you need both accounts logged in, the stake calculations done, and bets placed on both sides before either book moves. Manual arbitrage at this speed is exhausting and error-prone. Automated arbitrage requires software (Claw Arbs, OddsJam Arbitrage, RebelBetting, etc.) and even then, books can detect speed-of-placement patterns and ban you for that alone.
One-legged arbs. If your first bet places but the second leg's line moves before you can confirm the second bet, you're left with a "naked" one-sided bet at a price that may now be worse than the open market. This is called a one-legged arb. Even with software, this happens often enough that experienced arbers reserve roughly 20-30% of their expected arbitrage profit to cover one-legged losses.
Capital requirements. To make meaningful money arbing, you need significant bankroll spread across many accounts. A typical setup is $5,000-$20,000 spread across 10+ bookmaker accounts. Each account needs to be funded, KYC-verified, and ready to place bets within seconds. The capital is locked up across accounts, not deployable for other purposes, and rebalancing across accounts requires constant deposit/withdrawal activity — which itself is one of the signals books use to detect arbers.
What real arbitrage looks like in 2026
Sharp arbers in 2026 have largely shifted to cross-venue arbitrage, specifically between traditional sportsbooks and prediction markets like Kalshi and Polymarket. Prediction markets price outcomes as 0-100 cent contracts, sportsbooks use American or decimal odds, and the different pricing mechanisms create discrepancies that traditional book-vs-book arbs no longer reliably do.
Even there, arbing has become a tightly-margined professional activity. Solo bettors with manual workflows are unlikely to make meaningful money. Successful arbers operate at scale (large stakes, dozens of accounts, full-time attention) and even then yields are typically 0.5-2% per opportunity with high operational overhead.
The honest picture: arbitrage is a real strategy that exists, but it's a job with real capital requirements, real time commitments, and real downside risks (account closures, locked funds, voided bets). It's not a passive income stream, and it's not a side hustle that compounds while you focus on your day job.
When arbitrage opportunities exist legitimately
A few situations where arbitrage opportunities are more common and worth knowing about:
Major events with promotional pricing. Sportsbooks sometimes offer odds-boosted prices on specific bets as promotions. The boosted price can be high enough that combined with a normal price elsewhere, an arb exists. These are rare and short-lived but real.
Live in-game arbs. Live odds move fast and books update at different speeds. Brief mispricings emerge during games. Catching them requires having multiple live odds feeds open and being able to bet within seconds.
Niche sports and prop markets. Books with thin pricing on table tennis, darts, or specific player props sometimes set lines that don't match other books. The risk: these are exactly the markets books monitor most aggressively for arbing, so finding the opportunities often means finding bans.
Prediction market vs. sportsbook discrepancies. As of 2026, this is the most sustainable arbitrage surface. Kalshi or Polymarket prices on a market often differ from sportsbook prices on the equivalent event. The cross-venue execution requires accounts on both types of platform.
The realistic alternative
For most bettors interested in the idea of profitable, math-driven betting, the honest answer isn't arbitrage. It's the combination of:
Expected value betting at sharp books. Pinnacle, Circa, and BetCRIS don't limit winning accounts. Bettors who find +EV bets at these books can size up to meaningful stakes without getting banned. The vig is lower (often 2-3% on lines vs. 4-5% at retail books), and the action is welcome. The downside: smaller markets, lower limits than U.S. retail books offer for large bettors, and offshore status creates friction for U.S. residents.
Promotional EV from soft books. New-user bonuses, odds boosts, and risk-free bets at retail books offer genuine positive expected value if you can claim them. Once the promos run out, the account is much less useful, but the initial EV captured is real money. We cover this in Promotional EV and Bonus Abuse.
Closing line value as the metric. Focus on consistently beating the closing line at any book that will take your action. CLV is the predictor; arbing is a different game. See Closing Line Value.
The summary
Arbitrage betting is mathematically real but practically narrow. The theory promises guaranteed profit; the reality requires significant capital, mature operational discipline, software, multiple accounts, and acceptance that any successful sportsbook account is on borrowed time.
For solo bettors with limited time and capital, the math suggests focusing on expected value betting — finding bets where your estimated probability is meaningfully better than the sportsbook's implied probability — rather than chasing risk-free arbitrage that the market and sportsbook policy combine to make less risk-free than it sounds.
If you do explore arbitrage, do it at modest scale, expect account closures, and have a plan for moving funds out before books delay withdrawals. The strategy works for some people. It's not the easy money the math makes it sound like.
ParlayX provides analytics tools and educational content, not betting advice. Sports betting involves financial risk and is intended for adults only. If you or someone you know has a gambling problem, call 1-800-GAMBLER for confidential help, 24 hours a day.