> ## Documentation Index > Fetch the complete documentation index at: https://docs.fieldfunded.com/llms.txt > Use this file to discover all available pages before exploring further. # Analyze Sports Data with Python and Pandas > Analyze sports odds data with Python and pandas. Fetch live odds, build DataFrames, calculate implied probabilities, and find value bets. # Analyze Sports Odds Data with Python and Pandas Turn raw odds data into actionable insights using Python and pandas. This tutorial fetches live odds from the FieldFunded API, loads them into DataFrames, and runs analysis — implied probabilities, margin calculation, and value detection. ## Prerequisites ```bash theme={null} pip install requests pandas ``` You also need a free API key from [fieldfunded.com/docs](https://fieldfunded.com/docs) (no credit card required). ## Step 1: Fetch Odds into a DataFrame ```python theme={null} import requests import pandas as pd API_KEY = "your_api_key_here" BASE = "https://api.fieldfunded.com/v1" H = {"X-API-Key": API_KEY} # Get Premier League events events = requests.get( f"{BASE}/events", headers=H, params={"sport": "soccer", "league": "england_epl"} ).json() # Build a flat list of all match winner odds rows = [] for event in events["events"][:10]: odds = requests.get( f"{BASE}/events/{event['id']}/odds", headers=H ).json() for market in odds["markets"]: if "1x2" not in market["name"].lower() and \ "winner" not in market["name"].lower(): continue for sel in market["selections"]: rows.append({ "match": f"{event['home_team']} vs {event['away_team']}", "kickoff": event["commence_time"], "selection": sel["name"], "odds": sel["odds"], "market_id": market["id"], "selection_id": sel["id"], }) df = pd.DataFrame(rows) print(df.head(9)) ``` Output: ``` match kickoff selection odds 0 Manchester United vs Liverpool 2026-05-10T15:00:00Z Man United 3.40 1 Manchester United vs Liverpool 2026-05-10T15:00:00Z Draw 3.60 2 Manchester United vs Liverpool 2026-05-10T15:00:00Z Liverpool 2.10 3 Arsenal vs Aston Villa 2026-05-10T17:30:00Z Arsenal 1.55 4 Arsenal vs Aston Villa 2026-05-10T17:30:00Z Draw 4.20 5 Arsenal vs Aston Villa 2026-05-10T17:30:00Z Aston Villa 5.80 ``` ## Step 2: Calculate Implied Probabilities ```python theme={null} # Implied probability = 1 / decimal_odds df["implied_prob"] = (1 / df["odds"] * 100).round(1) # Group by match to calculate the margin (overround) def calc_margin(group): total_prob = group["implied_prob"].sum() group["margin"] = (total_prob - 100).round(1) return group df = df.groupby("match", group_keys=False).apply(calc_margin) print(df[["match", "selection", "odds", "implied_prob", "margin"]].head(6)) ``` Output: ``` match selection odds implied_prob margin 0 Manchester United vs Liverpool Man United 3.40 29.4 5.9 1 Manchester United vs Liverpool Draw 3.60 27.8 5.9 2 Manchester United vs Liverpool Liverpool 2.10 47.6 5.9 3 Arsenal vs Aston Villa Arsenal 1.55 64.5 5.1 4 Arsenal vs Aston Villa Draw 4.20 23.8 5.1 5 Arsenal vs Aston Villa Aston Villa 5.80 17.2 5.1 ``` The margin (5-6%) represents the bookmaker's edge. Lower margins mean better odds for bettors. ## Step 3: Find Value Bets A "value bet" is when you believe the true probability is higher than the implied probability. Use your own model or estimates: ```python theme={null} # Example: your model's probability estimates your_model = { "Man United": 32.0, # Your model says 32%, book says 29.4% "Draw": 26.0, # Model says 26%, book says 27.8% "Liverpool": 42.0, # Model says 42%, book says 47.6% "Arsenal": 68.0, # Model says 68%, book says 64.5% } df["model_prob"] = df["selection"].map(your_model) df["edge"] = (df["model_prob"] - df["implied_prob"]).round(1) # Value bets: where your model gives higher probability than the book value_bets = df[df["edge"] > 0].sort_values("edge", ascending=False) print(value_bets[["match", "selection", "odds", "implied_prob", "model_prob", "edge"]]) ``` Output: ``` match selection odds implied_prob model_prob edge 3 Arsenal vs Aston Villa Arsenal 1.55 64.5 68.0 3.5 0 Manchester United vs Liverpool Man United 3.40 29.4 32.0 2.6 ``` ## Step 4: Track Odds Over Time Build a historical dataset by polling at intervals: ```python theme={null} import time from datetime import datetime def collect_snapshots(event_id, intervals=10, pause=60): """Collect odds snapshots over time.""" snapshots = [] for i in range(intervals): odds = requests.get( f"{BASE}/events/{event_id}/odds", headers=H ).json() timestamp = datetime.now().isoformat() for market in odds["markets"]: if "winner" not in market["name"].lower(): continue for sel in market["selections"]: snapshots.append({ "timestamp": timestamp, "selection": sel["name"], "odds": sel["odds"], }) print(f"Snapshot {i + 1}/{intervals} collected") if i < intervals - 1: time.sleep(pause) return pd.DataFrame(snapshots) # Collect 10 snapshots, 1 minute apart # history = collect_snapshots("event_abc123", intervals=10, pause=60) ``` ## Step 5: Visualize Odds Movement ```python theme={null} import matplotlib.pyplot as plt def plot_odds_history(history_df): """Plot odds movement over time for each selection.""" fig, ax = plt.subplots(figsize=(10, 5)) for name, group in history_df.groupby("selection"): ax.plot( range(len(group)), group["odds"].values, marker="o", label=name, linewidth=2, ) ax.set_xlabel("Snapshot") ax.set_ylabel("Decimal Odds") ax.set_title("Odds Movement Over Time") ax.legend() ax.grid(True, alpha=0.3) plt.tight_layout() plt.savefig("odds_movement.png", dpi=150) print("Chart saved to odds_movement.png") # plot_odds_history(history) ``` ## Step 6: Export for Further Analysis ```python theme={null} # Save to CSV df.to_csv("epl_odds.csv", index=False) print(f"Saved {len(df)} rows to epl_odds.csv") # Save to Excel with multiple sheets with pd.ExcelWriter("odds_analysis.xlsx") as writer: df.to_excel(writer, sheet_name="All Odds", index=False) value_bets.to_excel(writer, sheet_name="Value Bets", index=False) print("Saved to odds_analysis.xlsx") ``` ## Rate Limit Math | Analysis type | Requests | Frequency | Monthly | Plan | | ---------------------------------------- | -------- | --------- | ------- | -------------- | | One-off EPL analysis (10 matches) | 11 | Once | 11 | Free | | Weekly analysis across 3 leagues | 90 | 4x/month | 360 | Free | | Daily odds tracking (10 events, hourly) | 240 | Daily | 7,200 | Free | | Full pipeline with historical collection | 1,000 | Daily | 30,000 | Starter (\$29) | Data analysis is typically low-frequency — most workflows fit in the [free tier](/compare/free-sports-api). *** ## Related Guides * [Get Live Odds with Python](/blog/get-live-odds-python) — simpler Python tutorial for beginners * [Build a Line Tracker](/guides/build-line-tracker) — track odds movements with a UI * [Build a Prediction Model](/guides/sports-prediction-model) — use odds data for ML predictions * [Web Scraping vs API](/blog/scraping-odds-vs-api) — why an API is better than scraping Start analyzing in 5 minutes — 10,000 free requests/month All plans compared side by side