We built the data set we wished existed when we started comparing sportsbook prices. The AgentBet Vig Index tracks average vig across 16 sportsbooks, broken down by sport and bet type, updated monthly.
Why We Built This
Two groups need sportsbook vig data and neither had a good source:
Sharp bettors who line shop know that Circa and Pinnacle are cheap and Bovada is expensive — but didn’t have a consistent, quantified benchmark to reference. “Circa has lower vig” is less useful than “Circa averages 2.1% vig versus Bovada’s 5.1% — a 3.0 percentage point gap worth $3,000 per 1,000 bets.”
AI betting agents need programmatic vig data to route bets optimally. An agent’s intelligence layer identifies a +EV bet. The routing logic needs to know which book to send it to. Without a vig baseline, the agent is guessing. With the Vig Index, it knows that a standard NFL spread should cost 2.1% at Circa and 4.3% at DraftKings — and routes accordingly.
March 2026 Highlights
The full data lives in the AgentBet Vig Index. Here are the highlights:
Overall rankings: Circa (2.1%), Pinnacle (2.3%), BetAnySports (2.5%), LowVig.ag (2.8%), CRIS (2.9%), DraftKings (4.2%), FanDuel (4.3%), Bet365 (4.5%), BetMGM (4.7%), Caesars (4.8%), Bovada (5.1%), MyBookie (5.5%).
DraftKings vs FanDuel: Within 0.1% on average. DraftKings wins on NBA spreads and player props. FanDuel wins on NFL moneylines. See the full head-to-head.
Offshore vs regulated: Reduced-juice offshore books (BetAnySports, LowVig) charge roughly half the vig of the best regulated books. But standard offshore books (Bovada, MyBookie) are actually worse than DraftKings. See the full comparison.
By sport: NFL and MLB are priced most efficiently. College basketball has the highest vig. See odds rankings by sport.
By bet type: Spreads and totals carry 2-5% vig. Player props carry 5-15%. Futures carry 12-35%. The vig on a same-game parlay is roughly 3-4x the vig on a straight bet.
What’s in the Index
The main Vig Index page contains:
- Overall sportsbook rankings by average vig
- Sport-specific tables for NFL, NBA, MLB, NHL, college football, and college basketball
- Bet type breakdown (spreads, totals, moneylines, props, parlays, futures, live)
- Prediction market fee comparison (Polymarket, Kalshi) alongside sportsbook vig
- Methodology and data sources
- Agent integration guide
Supporting pages:
- How to Calculate Vig — Formula, worked examples, and Python code
- DraftKings vs FanDuel vs BetMGM Odds — Big three head-to-head
- Offshore vs Regulated Vig — Sharp books vs mainstream
- Best Odds by Sport — Sport-specific rankings
- Vig Shopping Strategy — How to use the data
The Agent Angle
The Vig Index integrates into every layer of the agent betting stack:
At the trading layer, agents use vig data to select which sportsbook or prediction market to execute on. A bet routed to Circa (2.1% vig) instead of BetMGM (4.7% vig) saves 2.6% — pure structural edge.
At the intelligence layer, agents factor vig into +EV calculations. A bet with 3% expected edge at a 5% vig book is -EV. The same bet at a 2% vig book is +EV. Vig is not a secondary concern — it determines whether a bet is worth placing.
At the wallet layer, agents managing autonomous wallets across multiple books can allocate funds proportionally to vig rankings — keeping more capital at low-vig books.
What’s Next for the Index
The Vig Index will update monthly. We plan to add:
- Historical vig trends (how has each book’s pricing changed over time?)
- Prop-specific vig tracking by sport and prop type
- Live betting vig comparison
- Prediction market fee tracking for Polymarket and Kalshi as their volumes grow
- An API endpoint for agents to query vig baselines programmatically
If you’re building an agent that bets, the Vig Index should be in its context window. If you’re a bettor shopping lines, bookmark it.