Vig shopping is the only guaranteed edge in sports betting. You don’t need a model, you don’t need inside information. You just need accounts at multiple books and 30 seconds to check prices before every bet. The AgentBet Vig Index data shows this saves $2,500-$3,500 per 1,000 bets.
The Math Behind Vig Shopping
The AgentBet Vig Index tracks 16 sportsbooks. On any given market, the vig gap between the best and worst book ranges from 2.5% to 3.5%. Here’s what that means in practice:
| Scenario | Book Used | Vig | Bets | Stake | Vig Cost |
|---|---|---|---|---|---|
| Worst book | MyBookie | 5.5% | 1,000 | $100 | $5,500 |
| Average book | Caesars | 4.8% | 1,000 | $100 | $4,800 |
| Best regulated | DraftKings | 4.2% | 1,000 | $100 | $4,200 |
| Sharp book | Circa | 2.1% | 1,000 | $100 | $2,100 |
| Vig shopper | Varies | ~2.5% | 1,000 | $100 | ~$2,500 |
A bettor using only MyBookie pays $5,500 in vig over 1,000 bets. A vig shopper with accounts across tiers pays roughly $2,500 — saving $3,000 with zero handicapping effort.
The Vig Shopping Playbook
Account Setup
You need accounts at books spanning different vig tiers. The Vig Index identifies three tiers:
Tier 1 — Sharp book (pick one):
- Circa (2.1% avg vig) — Best if you’re in Nevada
- Pinnacle (2.3%) — Best international option
- BetAnySports (2.5%) — Best US offshore reduced-juice option
- LowVig.ag (2.8%) — Another solid reduced-juice offshore
Tier 2 — Competitive regulated (pick two):
- DraftKings (4.2%) — Best overall among regulated books
- FanDuel (4.3%) — Very close to DraftKings
- Bet365 (4.5%) — Good alternative
Tier 3 — Promo book (optional, pick one):
- BetMGM (4.7%) — Largest welcome bonus, frequent promos
- Caesars (4.8%) — Strong loyalty program
Four accounts is the sweet spot. You capture the sharp-to-competitive spread on every bet and the promotional upside when available.
The Shopping Process
For every bet:
- Identify your target market — e.g., Chiefs -3.5 spread
- Pull odds from all your books — takes 30 seconds manually, milliseconds for an agent
- Calculate the vig at each book — convert odds to implied probability, sum, subtract 100%
- Place the bet at the lowest-vig book — route to whichever book offers the best price for this specific market
- Log the vig paid — track your average over time
The key insight: the best book changes from bet to bet. DraftKings might have the best NFL spread on Chiefs -3.5 but FanDuel has the best total on the same game. Never default to a single book.
Where Shopping Matters Most
Vig differentials aren’t uniform across bet types. The biggest gaps (and biggest savings) come from:
Futures markets (10-15% vig variance): A Super Bowl futures bet might carry 15% vig at BetMGM and 25% at MyBookie. That’s a 10 percentage point gap on a single bet. For futures, always check a sharp book first.
Player props (5-10% vig variance): DraftKings charges roughly 8.5% vig on player props while BetMGM charges 10.2%. On a $100 prop bet, that’s $1.70 per bet — and prop bettors often place 5-10 props per day.
Live/in-play betting (3-8% vig variance): Books price live bets with wider margins because they’re harder to model in real time. FanDuel’s live betting engine tends to offer tighter vig than competitors.
Moneylines on heavy favorites (3-5% vig variance): When one side is -300 or heavier, the vig calculation amplifies small odds differences. A -300/+240 line (4.41% vig) versus -280/+230 (3.93% vig) saves 0.48% on a single bet.
Closing Line Value (CLV)
Vig shopping pairs naturally with CLV tracking. CLV measures whether you got a better price than the closing line — the final odds at game time, which are generally the most efficient.
Books that consistently have the best closing lines are also the best to bet at. Research shows:
| Book | % of Games with Best Closing Line |
|---|---|
| Circa | ~20% |
| CRIS / Bookmaker | ~14% |
| LowVig.ag | ~13% |
| Pinnacle | ~14% (tracked separately) |
| DraftKings | ~5% |
| FanDuel | ~6% |
| BetMGM | ~4% |
| Bovada | ~1% |
Circa is the closing-line king. If you can only have one sharp book, make it Circa (or Pinnacle internationally).
How AI Agents Automate Vig Shopping
For human bettors, vig shopping takes discipline and 30 seconds per bet. For AI agents, it takes milliseconds and zero discipline — it’s just code.
The agent vig shopping loop:
1. Intelligence layer identifies +EV bet
2. Query The Odds API for odds across all books
3. Calculate vig at each book (see /guides/how-to-calculate-vig/)
4. Select book with lowest vig
5. Wallet layer places bet at optimal book
6. Log vig saved vs. baseline
An agent connected to 5-6 odds sources via OddsPapi or The Odds API and routing to the lowest-vig book on every bet achieves 2-3% structural edge before any prediction accuracy. This is the lowest-risk component of the entire agent betting stack.
For prediction market equivalents, agents check the spread between Polymarket’s CLOB (0% maker / 1-2% taker fees) and Kalshi ($0.01-0.03 per contract) on overlapping markets.
Tracking Your Vig
Start logging the vig on every bet you place. The number to track is your average vig paid per bet.
- Above 5% — You’re overpaying. Add a sharp book to your portfolio.
- 4-5% — Average. You’re probably using one regulated book without shopping.
- 3-4% — Good. You’re shopping between regulated books.
- 2-3% — Sharp. You’re routing to sharp books as primary.
- Below 2% — Elite. You’re combining sharp books with promotional bets.
Every 1% reduction in average vig paid translates to approximately $1,000 saved per 1,000 bets at $100 stakes.
What’s Next
- AgentBet Vig Index — Monthly vig data across 16 sportsbooks
- How to Calculate Vig — The formula and Python code
- DraftKings vs FanDuel vs BetMGM Odds — Head-to-head comparison of the big three
- Prediction Market API Reference — Connect agents to odds data
- Agent Betting Stack — Full agent architecture
- Offshore Sportsbook Profiles — Sharp book deep dives
- Prediction Market Odds Explained — Odds conversion fundamentals