Sportsbook vs Polymarket Edge Finder | Devig +EV Scanner

Find mispriced Polymarket sports markets in one run. This Actor pulls live DraftKings moneylines (via ESPN's public feed), strips the bookmaker's vig to get the fair win probability, then compares it with the live Polymarket price for the same game. The output is a ranked feed of games where Polymarket is cheaper than the sportsbook's fair price — the classic +EV signal sports bettors and prediction-market traders act on every day.

Works for MLB, NFL, NBA, NHL, College Football and the Premier League. No proxies, no API keys, no logins.

Who is this for?

• Polymarket traders — buy the side the sportsbook says is underpriced. Sportsbooks sharpen their lines with millions in handle; when Polymarket disagrees by 3-5 points, that gap is your edge.
• Sports bettors — use devigged fair probabilities as a single source of truth for line shopping and +EV betting.
• Arbitrage & modelers — a clean, matched cross-platform dataset (sportsbook odds + prediction-market prices per game) that normally takes two scrapers and a matching layer to build.
• Newsletter & Discord operators — auto-publish a daily "top edges" card.

What you get (per matched game)

Field	Description
rank	Position sorted by best edge
sport	League (mlb, nfl, nba, nhl, cfb, epl)
game	Away Team @ Home Team
startTime	Scheduled start (UTC)
sportsbook	Odds provider (DraftKings via ESPN)
moneylineHome / moneylineAway	American moneylines
vigPct	Bookmaker margin removed by the devig
fairProbHome / fairProbAway	No-vig fair probabilities
polymarketPriceHome / polymarketPriceAway	Live Polymarket prices
polymarketVolume24h	Market's 24h volume (USD)
edgeSide / edgeTeam	Which side has the edge
edgePct	Fair probability − Polymarket price, in points
buyPrice / fairPrice	What you pay vs what it's worth
recommendation	Plain-English action, e.g. BUY Texas Rangers on Polymarket @ 0.595 (sportsbook fair 0.6497)
polymarketUrl	Direct link to the market
spread / overUnder	Sportsbook spread & total for context
capturedAt	Snapshot timestamp

Pricing — pay per event

You pay $0.001 per run start and $9.60 per 1,000 edge rows — 20% below comparable sports-odds feeds, and every row here is enriched (devig + cross-platform match), not a raw line. Unmatched games, rows below your filters, and failed runs are never charged.

A typical daily run across MLB + NFL + NBA + NHL returns 20-80 rows — a few cents per snapshot.

How to schedule it

Odds move all day, so this Actor shines on a schedule:

1. Open the Actor page → Schedule tab (or Console → Schedules → Create new).
2. Pick a cadence — every 1-3 hours on game days is the sweet spot for edge hunting; daily at 10:00 ET works for a morning shortlist.
3. Add an integration (webhook, Slack, Google Sheets, email) to push the top rows wherever you work.

Each run is an independent snapshot with capturedAt, so you can also build a time series of edges and closing-line value.

Use with AI agents & MCP

This Actor is exposed through the Apify MCP server, so Claude, ChatGPT, LangChain, CrewAI or any MCP-capable agent can call it as a tool. Ask your agent things like:

"Run the Sportsbook vs Polymarket Edge Finder for today's MLB and NFL games and tell me the three biggest positive edges with at least $1,000 of 24h volume."

The structured dataset (numeric edgePct, recommendation strings, direct URLs) is designed to be agent-readable out of the box.

Input

{
  "sports": ["mlb", "nfl", "nba", "nhl"],
  "minEdgePct": "0",
  "minPolymarketVolume24h": 1000,
  "maxRows": 100
}

• minEdgePct: "0" keeps only positive-EV rows; "-100" (default) outputs every matched game so you see the full board.
• minPolymarketVolume24h filters out thin markets you can't actually get filled in.

Honest limitations

• Devigging a two-way moneyline removes the margin proportionally; sharper multiplicative/power devigs exist, but the proportional method is the standard baseline.
• An edge vs the sportsbook fair price is a signal, not a guarantee — thin Polymarket books can be stale because nobody can fill them. That's what the volume filter is for.
• Pre-game markets only; live/in-play edges are out of scope.
• This is a data tool, not betting advice.

Related Actors

• Prediction Market Movers — Polymarket & Kalshi Catalyst Feed — the biggest odds swings and volume spikes across both platforms.
• Polymarket Kalshi Arbitrage Scanner — matched markets and price gaps between the two prediction exchanges.
• Kalshi Weather Markets Scraper — every Kalshi weather market with live prices.

Changelog

• 0.1.3 (24 Sep 2026) — Reliability: Runs near their time limit now stop scanning cleanly and still deliver every edge found so far, instead of being cut off with nothing saved. No price or field change.
• 0.1.2 (21 Sep 2026) — Explains thin results. Sportsbooks post moneylines only about a day before a game, so an early-week or late-night run can legitimately return a handful of rows. The log and the run's status message now say how many upcoming games in your window have no lines yet, so a small result is not mistaken for a broken Actor. No fields, defaults or prices changed.
• 0.1.1 (19 Sep 2026) — Typo guard on input. Apify accepts input fields an Actor does not know without complaining, so a misspelled option used to produce a successful run with that setting silently inactive. The run log now warns for every unknown field and suggests the closest real one.
• 0.0.3 (11 Sep 2026) — Fix: Polymarket pagination. Gamma now caps limit at 100; game events sit beyond the first page, so NFL matched 1 of 15 games. Every league now pages until exhaustion (NFL 15/15). New input daysAhead (1–14, default 2 = today + tomorrow): weekly leagues (NFL, college football, EPL) play 1–3 days a week — set 7 to always cover the next round.
• 0.0.1 (1 Sep 2026) — Launch: ESPN/DraftKings moneylines devigged vs Polymarket game markets for MLB, NFL, NBA, NHL, CFB, EPL.

Found this useful? Bookmark & review

A quick ⭐ review or bookmark on this Actor helps other people find it and tells us which fields to add next. Have a question, a missing field or a source that changed? Open an issue in the Issues tab — we answer within 24 hours.