Disclosure: I'm part of the team behind PinnFeed, the odds API used in the examples. 18+, odds data for software development, not betting advice.
Pinnacle's prices are the usual reference for "what is this outcome really worth?". To compare them with any other bookmaker you need two things: the prices, and the price without the bookmaker's margin (the no-vig or fair price).
This post covers both: the maths in a few lines of Python, then where to get Pinnacle prices without paying.
1. Remove the margin (multiplicative method)
A bookmaker's implied probabilities add up to more than 100%. The extra is the margin. The multiplicative method scales every outcome down by the same factor:
def no_vig(decimal_odds):
implied = [1 / o for o in decimal_odds]
overround = sum(implied) # e.g. 1.035 -> 3.5% margin
probs = [p / overround for p in implied]
fair = [1 / p for p in probs]
return overround - 1, probs, fair
margin, probs, fair = no_vig([1.781, 3.62, 5.05])
print(f"margin {margin:.2%}")
for p, f in zip(probs, fair):
print(f"p={p:.4f} fair={f:.3f}")
If another bookmaker offers more than the fair price, that line is priced above Pinnacle's no-vig estimate. That's the core of a +EV screen.
2. Getting Pinnacle prices
Pinnacle closed public access to its own API in July 2025, so today you go through independent services. Free tiers differ a lot (requests per day, delay, whether fair odds are included). We keep a dated comparison of the free tiers here: Is there a free Pinnacle odds API?
With PinnFeed's free key (no card, 1,000 requests a day, 2 sports, prices delayed 30 seconds) the API already returns fair, fairProb and margin, so you can check your own numbers against it:
import os, requests
r = requests.get(
"https://pinnfeed.io/v1/odds",
params={"sport": "soccer", "market": "moneyline", "main": "true", "limit": 5},
headers={"x-api-key": os.environ["PINNFEED_KEY"]},
timeout=10,
)
r.raise_for_status()
for ev in r.json()["events"]:
for m in ev["markets"]:
prices = [o["decimal"] for o in m["outcomes"]]
margin, probs, fair = no_vig(prices)
print(ev["home"], "v", ev["away"], f"margin {margin:.2%}")
for o, f in zip(m["outcomes"], fair):
print(f" {o['name']:<5} {o['decimal']:>6} mine {f:.3f} api {o['fair']}")
3. Where to go from here
- Which APIs include fair odds on every market: no-vig odds API comparison
- Closing prices for CLV tracking: Pinnacle closing odds sources compared
- Asking Claude or Cursor for odds instead of writing code: sports odds MCP servers compared
PinnFeed is independent and not affiliated with Pinnacle. Questions about the maths or the code are welcome in the comments.
Top comments (0)