Sports Arbitrage Explained: H2H, Spreads, and Totals
Sports arbitrage—sometimes called "sure betting" or "arbing"—is the practice of exploiting pricing discrepancies between different betting venues to lock in a mathematical edge, regardless of the event outcome. In the world of sports betting, these opportunities can pop up across various markets: the basic head-to-head (moneyline or h2h), point spreads, and totals (over/under). In this article, we’ll break down how arbitrage works in each context, using concrete numbers and real-world logic. We’ll also show how automated tools like ArbMonster help surface these fleeting opportunities.
Note: This article is about the math and mechanics, not about profits or guarantees. Actual outcomes depend on venue policies, timing, and execution risk.
What Is Sports Arbitrage?
At its core, sports arbitrage relies on differences in odds between venues. If you can bet all possible outcomes (e.g., Team A wins vs. Team B wins) at prices that sum to less than 100% probability, you can construct a set of bets that pays a positive return, regardless of outcome—at least in theory.
Let’s start with the simplest case: the head-to-head (h2h) or moneyline market.
H2H (Head-to-Head) Arbitrage: The Basics
Suppose you find the following odds for an NBA game:
- Venue 1: Lakers win @ +110 (decimal odds: 2.10)
- Venue 2: Warriors win @ +110 (decimal odds: 2.10)
If you bet $100 on each side, you will lose one bet and win the other:
- If Lakers win: You win $110 at Venue 1 and lose $100 at Venue 2. Net: +$10
- If Warriors win: You win $110 at Venue 2 and lose $100 at Venue 1. Net: +$10
But in practice, odds are rarely this generous. Let’s look at a more realistic scenario:
- Venue 1: Lakers win @ +105 (2.05)
- Venue 2: Warriors win @ +115 (2.15)
You want to know if you can lock in a profit. To check, convert the odds to implied probabilities:
- Lakers: 1/2.05 ≈ 48.78%
- Warriors: 1/2.15 ≈ 46.51%
- Total: 48.78% + 46.51% = 95.29%
If the total is less than 100%, arbitrage is possible.
How much to bet?
Suppose you want to stake a total of $100. The formula for stake sizing:
- Stake on Lakers: $100 × (1/2.15) / (1/2.05 + 1/2.15) ≈ $49.02
- Stake on Warriors: $100 - $49.02 = $50.98
Or, more generally: allocate stakes so that the payout for either outcome is the same.
Check the payout:
- If Lakers win: $49.02 × 2.05 = $100.50
- If Warriors win: $50.98 × 2.15 = $109.56
Subtract your total staked ($100):
- Lakers win: +$0.50
- Warriors win: +$9.56
In this example, arbitrage exists, but the payouts aren’t equal (due to rounding and small market inefficiency). The key is that both are positive.
Venue fees and limits:
In real life, you must also account for:
- Venue fees/commission (e.g., exchange commissions, prediction market fees)
- Betting limits (maximum you can bet at quoted odds)
- Changes in odds between placing the bets
- Jurisdictional or account restrictions
Tools like ArbMonster automate these calculations, scanning 20+ sportsbooks, Kalshi, and Polymarket, adjusting for all known fees, and showing only fillable, net-positive opportunities.
Spread Arbitrage: When the Lines Diverge
Point spreads give the underdog a head start (or the favorite a handicap). Arbitrage here relies on finding different lines, or the same line at different prices, across venues.
Example 1: Same Spread, Different Prices
Suppose for an NFL game:
- Venue 1: Patriots +3.5 (-110)
- Venue 2: Jets -3.5 (+115)
Here, the spread is the same but the prices differ. Calculate implied probabilities:
- Patriots +3.5 (-110): $110 to win $100 ⇒ implied probability = 110/(110+100) = 52.38%
- Jets -3.5 (+115): $100 to win $115 ⇒ implied probability = 100/(100+115) = 46.51%
- Total: 98.89%
Less than 100%: arbitrage is possible.
Stake sizing:
- Bet $110 on Patriots +3.5 at Venue 1
- Bet $100 on Jets -3.5 at Venue 2
If Patriots +3.5 covers: Win $100 at Venue 1, lose $100 at Venue 2 ⇒ Net: $0 If Jets -3.5 covers: Win $115 at Venue 2, lose $110 at Venue 1 ⇒ Net: $5
Here, arbitrage only if Jets cover the spread. Not a "true" arb unless both outcomes net positive. To achieve a true arbitrage, you’d need prices to be slightly better.
Example 2: Middle Opportunities (Line Divergence)
Suppose:
- Venue 1: Bulls -5.0 (-110)
- Venue 2: Bulls +6.0 (-110)
Now, you can bet both sides:
- Bet $110 on Bulls -5.0 at Venue 1
- Bet $110 on Bulls +6.0 at Venue 2
Outcomes:
- If Bulls win by 7+: Both bets win. Win $100 + $100 = $200 profit
- If Bulls win by 6: Bulls -5.0 is a win ($100), Bulls +6.0 is a push (refund). Net: $100
- If Bulls win by 5: Bulls -5.0 is a push (refund), Bulls +6.0 is a win ($100). Net: $100
- If Bulls win by 0-4 or lose: Bulls -5.0 loses ($110), Bulls +6.0 wins ($100). Net: -$10
This is known as a "middle"—not a guaranteed profit, but a chance at a big payout if the final margin lands in the overlap.
Totals Arbitrage: Over/Under
Totals (over/under) markets can also produce arbs, especially when venues differ on the line (the number) and/or the price (the odds).
Example: Over/Under Line Divergence
Suppose:
- Venue 1: Over 214.5 points (-105)
- Venue 2: Under 216.5 points (+105)
Now, you can bet Over 214.5 and Under 216.5. Both bets can win if the final score lands on 215 or 216!
- Bet $105 on Over 214.5 at Venue 1 to win $100
- Bet $100 on Under 216.5 at Venue 2 to win $105
Outcomes:
- If game finishes at 215 or 216 points: Both bets win! Net: $100 + $105 = $205 profit
- If game finishes at 214 or less: Over loses, Under wins. Net: $105 - $105 = $0
- If game finishes at 217 or more: Over wins, Under loses. Net: $100 - $100 = $0
- If game finishes at 215 or 216: Both win. (Middle hits)
Just like with spreads, this is a "middle" opportunity—not risk-free, but mathematically favorable if priced right.
The Role of Prediction Markets (Kalshi, Polymarket)
Traditional sportsbooks aren’t the only venues. Prediction markets like Kalshi and Polymarket offer binary contracts (yes/no bets), often with different fee structures and liquidity. These can be included in arbitrage calculations—sometimes creating opportunities when their prices diverge from sportsbooks.
For example, if the implied probability of "Over 214.5" on Kalshi is 48%, but a sportsbook offers "Under 214.5" at a price implying 52.5%, you could, in principle, hedge both sides for a net-positive return, after accounting for all fees and fills.
ArbMonster automates the scanning of both sportsbooks and prediction markets, net of all fees and at fillable size, helping users spot such opportunities in real time.
Practical Considerations
- Execution risk: Odds can change quickly. The edge may disappear before both bets are placed.
- Limits: Not all venues allow large bets at quoted odds. Always check the maximum fillable size.
- Fee structure: Sportsbooks may have built-in vig (the house edge); exchanges and prediction markets may charge commissions or fees on winnings.
- Venue rules: Settlement, voids, and grading can differ. Always read the rules.
- Account and jurisdiction restrictions: Some venues restrict access by location or account type.
FAQ
Q1: Can you always find arbitrage opportunities?
No. True arbitrage is rare and fleeting in major markets. Most of the time, odds are set so that all outcomes add up to more than 100% probability (the house edge). Tools like ArbMonster help by automating the search across many venues.
Q2: Are middles the same as arbitrage?
No. Middles are situations where both bets can win if the outcome lands in a "middle" range, but you can still lose or break even if not. Arbitrage aims for a guaranteed net-positive regardless of outcome, while middles offer higher upside with some risk.
Q3: How do fees and commissions affect sports arbitrage?
Fees, commissions, and spread wideners all reduce or erase the edge. Always calculate net of all costs—including prediction market fees and exchange commissions—before acting.
ArbMonster is a data service. Nothing here is financial or betting advice; markets carry risk, venues have age and jurisdiction restrictions, and you are responsible for verifying everything before acting.