CS2 and NFL research · prediction markets, DFS pick'em and sportsbooks
Hunt the edge. Show the math.
A read-only research tool that puts Kalshi, Polymarket, PrizePicks, Underdog, Sleeper and the sportsbooks side by side, models matches and player props with explicit uncertainty, and shows expected value as a range — at the price or line you could actually get.
- Read-only — never places bets or trades
- Markets, DFS apps and sportsbooks in one view
- Every pick recorded and graded
The problem
The number on the screen isn't the price you get
Thin edges in prediction markets are easy to see and hard to capture. Most tools skip the parts that decide whether a bet is actually worth taking.
Displayed odds aren't executable
A headline percentage is usually a last trade or a mid-price. What you can buy is set by the order book — and the best level may only hold a handful of contracts.
Fees and spread eat thin edges
Exchange fees and the bid–ask spread come straight out of expected value. An edge of a few points can disappear once both are counted.
One number hides the uncertainty
A single probability says nothing about how sure the model is. A thin sample — a new roster, a player's first few games — deserves a wide range, not false precision.
One app is half the picture
Kalshi, Polymarket, the DFS apps and the sportsbooks often price the same outcome — or set the same player's line — differently. Without them side by side you can't tell a good price from a merely available one.
Pipeline
How it works
Four steps, each one shown in the product so you can check the work.
- 01data
Match and player data
CS2 results, maps, side splits and per-player round stats; NFL schedules, results and per-player box scores; live Kalshi and Polymarket markets, DFS pick'em lines from PrizePicks, Underdog and Sleeper, and sportsbook prices including Pinnacle.
- 02p ± interval
Model with uncertainty
CS2: Elo with rating uncertainty and map-specific ratings, plus a round-by-round simulation of each map for player kills and headshots. NFL props: recency-weighted player distributions. Every estimate carries an interval, and nothing uses data from after the start.
- 03after fees
The real price on every venue
Walk the live Kalshi and Polymarket order books for your size and apply each venue's fees; take each DFS app's line and payout; remove the sportsbooks' margin to get their fair price.
- 04worst … best
EV as a range
Expected value shown from worst case to best case, so you see when an apparent edge is inside the noise.
DFS & sportsbooks
DFS pick'em and sportsbooks, in the same view
Every app's line for a player, side by side — with the model's pick, its chance as a range, and the gap to break-even. Sportsbook prices, margin removed, as the reference.
Best line, highlighted
When apps disagree on a player's line, the better one for the pick is marked.
Slips with honest math
2- to 5-pick entries priced from each leg's chance, one leg per match, EV shown as a range.
Books as the reference
Sportsbook prices — Pinnacle included — with the margin removed, compared at the exact same number.
Lines that move
See when an app moves a line after it was priced, and which way.
Team A vs Team B
Kills on maps 1–2
- Player X28.528.527.5Less 64%55–73% · +6.3
- Player Y31.530.5—Less 61%51–71% · +3.3
- Player Z24.524.524.5More 52%41–62% · −5.7
Features
What's inside
Built for people who want to see the assumptions behind every estimate.
DFS pick'em lines
CS2 kills and headshots on maps 1–2 and NFL player lines from PrizePicks, Underdog and Sleeper, each with the model's side, its chance with a range, and the gap to break-even.
Best line across apps
One row per player: every app's line side by side, with the better line for the pick highlighted when the apps disagree.
Slip builder
Suggested 2- to 5-pick entries with the chance all legs hit and EV as a range — one leg per match, because players in the same match win or lose together.
Sportsbook reference
Sportsbook prices with the margin removed — Pinnacle included — set against prediction-market prices and DFS lines at the exact same number.
Order-book EV
EV at the executable price: the live Kalshi and Polymarket books are walked for your size, with partial fills, slippage and each venue's fees.
Kalshi vs Polymarket
The same bet on both venues: the cheaper side, de-vigged prices, price gaps and fee-aware arbitrage with settlement-rule warnings.
Map-aware CS2 modeling
Uses map names, the veto, the decider and CT/T round splits. Map ratings are shrunk toward overall strength when samples are small.
Uncertainty up front
Every probability carries an interval, and EV is shown from worst case to best case — so an edge inside the noise looks like one.
Line movement & alerts
See when an app moves a line after it was priced, and get alerts when a pick clears break-even, a line moves, or a price gap opens.
Every pick graded
Each test's rules are written down before its data exists, and every pick it makes is recorded and graded after the game — including the ones that lose.
Verified matching
Every market and line is matched to the right game, map and player with checks; anything ambiguous is left out instead of guessed.
Read-only by design
Never places bets or trades. No account linking and no betting or trading permissions required.
Compare
Why we're different
Different tools serve different markets. We focus on quantitative research for CS2 and the NFL across prediction markets, DFS pick'em apps and sportsbooks — priced at what you can actually get, with the uncertainty shown.
Markets covered
- StatsEdge
- Kalshi and Polymarket (CS2 match / map winner, total maps; NFL game lines and player props), DFS pick'em lines on PrizePicks, Underdog and Sleeper (CS2 kills and headshots, NFL player stats), and sportsbook reference prices
- OddsJam
- Odds across many sportsbooks (publicly described as 100+ books)
- DGFantasy
- Research tools for DFS pick'em and exchange apps (e.g. PrizePicks, Underdog), including an optimizer and a middling tool
Where the “fair price” comes from
- StatsEdge
- Statistical models of match and player data, the other venue's de-vigged price, and sportsbook prices with the margin removed (Pinnacle as the sharp reference)
- OddsJam
- Sharp-book pricing, used by its +EV and arbitrage finders
- DGFantasy
- Not publicly described
Fees & order-book depth in EV
- StatsEdge
- Walks the live Kalshi and Polymarket order books (partial fills, slippage) and applies each venue's fee schedule
- OddsJam
- Not publicly described
- DGFantasy
- Not publicly described
Uncertainty shown with each estimate
- StatsEdge
- Intervals on probabilities; EV and slip EV shown as worst case … best case
- OddsJam
- Not publicly described
- DGFantasy
- Not publicly described
Methodology documented
- StatsEdge
- Rules written down before the data, walk-forward backtests, and every forward pick recorded and graded
- OddsJam
- Not publicly described
- DGFantasy
- Not publicly described
Prediction-market venues compared
- StatsEdge
- Kalshi vs Polymarket, bet for bet: best price per side, price gaps, arbitrage with settlement-rule warnings
- OddsJam
- Sportsbooks (line shopping and arbitrage across books)
- DGFantasy
- Not publicly described
Sports covered
- StatsEdge
- Counter-Strike 2 and the NFL
- OddsJam
- Many sports across sportsbooks
- DGFantasy
- Sports offered by DFS pick'em apps
Comparison based on each company's public product descriptions as of October 2026. OddsJam and DGFantasy are trademarks of their respective owners; no affiliation.
Early access
Join the waitlist
Get an email when early access opens. No spam — just the launch.
StatsEdge is read-only research software. It never places bets or trades and needs no account linking.
Questions
FAQ
Does it place bets or trades?
No. It is read-only. It never places bets or trades and doesn't need account linking or any betting or trading permissions.
Is this betting or financial advice?
No. It is a research tool that shows model estimates, executable prices and their uncertainty. Any decision is yours.
Which markets does it cover?
Prediction markets: CS2 match winner, map winner and total maps, and NFL moneyline, spreads, totals and player props on Kalshi and Polymarket. DFS pick'em: CS2 kills and headshots on maps 1–2 and NFL player stats on PrizePicks, Underdog and Sleeper. Sportsbook prices are used as a reference. More sports are planned.
How do DFS picks account for the payout?
Each pick is shown against the hit rate a standard entry needs to break even — about 57.7% per pick for a 2-pick entry paying 3x. The slip builder multiplies the legs' chances and the app's payout, and shows the result as a range.
Why compare with sportsbooks?
Sharp sportsbooks like Pinnacle are a strong reference for the true chance of an outcome. When a prediction market's price or a DFS app's line strays from what the books imply, that gap is worth a look.
Why compare Kalshi and Polymarket?
They often price the same outcome differently. Seeing both lets you find the better price for each side, and spot when the two venues disagree by more than their fees.
Is profit guaranteed?
No. Every estimate is shown with its uncertainty, and prediction markets, DFS contests and sports betting all involve risk of loss.
Where does the data come from?
CS2 results from PandaScore and map/side data from Liquipedia (CC-BY-SA); NFL schedules and player stats from nflverse; Kalshi and Polymarket public market data; licensed DFS line and sportsbook odds data.