Aggregate statistics from 50 books in one call. Median, mean, min, max, and standard deviation for each outcome. See where books agree, where they disagree, and who the outliers are.
50
Books aggregated
Real-time
30s refresh
24
Sports covered
{
"event": "Lakers vs Warriors",
"market": "h2h",
"outcomes": [
{
"name": "Lakers",
"median_price": -142,
"mean_price": -140,
"min_price": -155,
"max_price": -130,
"std_dev": 5.8,
"books_reporting": 47
},
{
"name": "Warriors",
"median_price": +120,
"mean_price": +118,
"min_price": +108,
"max_price": +135,
"std_dev": 6.2,
"books_reporting": 47
}
]
}Median
The middle price across all 50 books. This is the market's best estimate of the true line. If you're getting a price significantly better than median, you're likely on the right side of value.
Std Dev
High standard deviation means books disagree on pricing. Disagreement = opportunity. Low std dev means the market has converged — harder to find edges. Sort by highest std dev to find the most inefficient markets.
Min / Max
The widest spread between min and max tells you where to bet and where to fade. Max price = best available odds (bet here). Min price = worst odds (avoid this book). A 25-point min/max spread on a two-way market is highly unusual — investigate.
Pull consensus and find disagreement.
import requests
response = requests.get(
"https://api.theoddsapi.com/v4/intelligence/consensus",
headers={"x-api-key": "YOUR_KEY"},
params={"sport_key": "basketball_nba"}
)
events = response.json()["data"]
# Sort by highest disagreement (std_dev) — most edge potential
for event in sorted(events, key=lambda e: max(o["std_dev"] for o in e["outcomes"]), reverse=True)[:5]:
print(f"{event['event']}")
for o in event["outcomes"]:
spread = o["max_price"] - o["min_price"]
print(f" {o['name']}: median {o['median_price']}, spread {spread}pt, σ={o['std_dev']}")Use the 50-book consensus as a benchmark. If your model disagrees with consensus by more than 3%, investigate. Either your model sees something the market doesn't, or your model is wrong. Consensus keeps you calibrated.
Sort events by standard deviation. High σ = books disagree = edge opportunities. Focus your energy on events where the market hasn't converged. Low σ events are efficient — skip them.
If you don't have a proprietary model, the 50-book median is your best proxy for true probability. Bet any book offering significantly better than median. Combined with /intelligence/fair-odds (vig-removed), you get two independent fair-price signals.
Business plan includes consensus, fair odds, value detection, edge scanning, and arbitrage across 24 sports.