Arbitrage Betting Explained: How to Lock in Guaranteed Profit
Arbitrage betting is the closest thing in sports betting to a free lunch. You bet both sides of a market at different sportsbooks at prices that guarantee a profit no matter who wins. No prediction, no edge, no model — just math.
That sounds too good to be true, and there are real catches: account limits, line movement, capital tied up, and the pure operational grind. But arbitrage is real, it works, and you can do it without a model. This guide covers what arbitrage is, how to find arbs, how to size your bets, what can go wrong, and when arbing makes more sense than +EV betting (and when it doesn't).
What is sports arbitrage betting?
An arbitrage bet (an "arb") exists when two sportsbooks disagree enough on a market that their combined implied probability is less than 100%. When that happens, you can split your bankroll across both sides and lock in profit.
Example: Pinnacle has the Lakers -3.5 at +105. DraftKings has the Celtics +3.5 at +105. If you bet $100 on each at +105 (decimal 2.05), one side wins $105 profit, the other loses $100. Net profit: $5 on $200 wagered, regardless of outcome. That is a 2.5% arb.
The math: implied probability of each side at +105 is 100/205 = 48.78%. Combined: 97.56%. Anything below 100% is an arb. The lower the combined probability, the bigger the profit.
How to calculate stake split
The classic arb formula assumes you want equal profit no matter which side wins:
Stake A = Total / (Decimal A) × (Decimal A × Decimal B / (Decimal A + Decimal B))
Or simpler: total bankroll × (Decimal B / (Decimal A + Decimal B)) on side A; total × (Decimal A / (Decimal A + Decimal B)) on side B.
Our free arbitrage calculator handles this for you — paste both prices, set your total bankroll, and it returns the exact stake on each side and the guaranteed profit.
Where do arbs come from?
Arbs exist because sportsbooks update lines on different schedules and react to information at different speeds. Common sources:
- Sharp vs soft disagreement. Pinnacle moves 5 minutes after a starting lineup change. DraftKings doesn't move for 15. The window is the arb.
- Promo-driven mispricing. Books offering boost odds or odds boosts on specific sides create temporary arbs against the unboosted side at another book.
- Player prop disagreements. Three books pricing Anthony Edwards over 6.5 assists — one at -110, one at +120, one at +145 — almost always means one of them is wrong. The +145 vs the no-vig fair price (~+105) is your arb or +EV play.
- Live betting. Books update in-game lines independently. Arbs appear and disappear in seconds. Live arbing is high-yield but high-effort.
Realistic arb returns
Public arbs (the ones that show up on free or low-tier scanners) typically pay 0.5% – 2% per cycle. Private feeds and live arbing can find 3% – 5% arbs more often, but they require capital across many books and very fast execution.
A realistic arb bettor making 10–20 plays per day at $200/play averages $40–$80/day in profit before account limits. That is real money but it is grind work, not passive income.
The 5 things that go wrong arbing
- Account limits. Sportsbooks identify arb bettors by patterns: round-number stakes, betting both sides of markets, taking lines right after a move. They limit or ban accounts. Mitigation: vary stake sizes (round to $7.43 not $200), avoid betting both sides on the same book family, mix in -EV bets to look recreational.
- Line move between bets. You hit submit on side A. By the time you go to book B, the price moved. You either skip the arb (locking in a loss on side A) or chase a worse price (cutting your arb).
- Bonus restrictions. Sign-up bonuses often have minimum odds or rollover requirements that interact badly with arbing. Read the fine print.
- Withdrawal friction. Some books slow-walk withdrawals on suspicious accounts. Plan your bankroll around 1–2 week withdrawal cycles.
- Stale lines / palp errors. Sometimes a book leaves a line up at the wrong price. Hitting it might trigger a palpable error void — the book voids your bet but not the bet on the other side, leaving you exposed. Sportsbook rules vary; check before you bet.
Arbitrage vs +EV betting
Arbing locks in small risk-free profit. +EV betting takes positions that win on average over time. They are different strategies with different risk profiles:
| Arbitrage | +EV | |
|---|---|---|
| Risk per bet | ~0% (assuming both bets land) | Variable |
| Edge per bet | 0.5–2% | 3–10% on a real edge |
| Variance | Very low | Higher (need 200+ bets to converge) |
| Capital required | High (across many books) | Moderate |
| Account-limit risk | High | Lower |
| Skill required | Operational speed | Modeling + line shopping |
Most pros do both. Arbs and bonus hunting cover overhead and grow bankroll fast in year 1. +EV betting is where the real edge compounds in years 2–5 because account limits matter less when the book doesn't immediately recognize you as +EV.
Should you start arbing?
Probably yes if: you have $5K+ deployable across 6+ accounts, you can place bets quickly from your phone, and you are okay with eventual account limits. Probably no if: you have less than $1K bankroll (the friction kills you), you can't watch your phone during games, or your jurisdiction limits sportsbook diversity.
Tools
Use our free arbitrage calculator to verify any arb you find manually — paste the two prices, set your total stake, get the exact split and profit. Pair it with the no-vig calculator to sanity-check whether one of the lines is sharp and one is soft (the typical arb shape) or whether both books are wrong about something.
Related reading
Want picks like these?
Turtle +EV scans thousands of player props every day and surfaces only the +EV opportunities backed by our probability models.
See Today's Picks