● Beginner

HOW TO USE
PROPBETEDGE

A complete walkthrough of every feature — from the live slate to model odds to the K-Score. Takes 10 minutes. Worth every second.

The Daily Workflow

Most PropBetEdge users follow a simple daily routine. Here's the optimal order:

  1. 1

    10AM CT — Check the Live Slate

    The home view shows today's full game slate with probable pitchers, game times, and current scores. Start here to understand who's playing and in what conditions.

  2. 2

    10AM CT — Review Model Odds Top Plays

    After the 10AM model run, check the Top Plays view. This shows today's highest OVER value plays sorted by expected stat. These are the model's best calls for the day.

  3. 3

    Check Weather

    Any game with 10+ mph wind blowing out? That's a HR multiplier. The weather view shows real-time conditions for every park hosting a game today.

  4. 4

    Check Umpires

    The umpire view shows the home plate umpire for each game and their K-tendency (above/below average). A K-friendly ump is a bonus for K prop plays.

  5. 5

    Compare Model Odds to Book Odds

    Each pick shows both our model odds and the best available DK/FD line. The gap is your edge signal. Book paying more than fair value = +EV.

  6. 6

    Flat Bet Your Selections

    Same dollar amount on every +EV play. No parleying model picks — parlays compound variance, not edge.

Reading the Model Odds Output

Every model odds row shows the same fields. Here's what each one means:

{ "player_name": "Paul Skenes", "team": "Pittsburgh Pirates", "market": "strikeouts", // hr | totalbases | hits | strikeouts "line": 6.5, // the over/under threshold "direction": "OVER", "expected_stat": 6.38, // lambda — expected Ks per game "implied_prob": 0.421, // our model: 42.1% chance of OVER "odds_fmt": "+120", // our fair-value odds "book_odds": "+148", // best available DK/FD line "k_score": 78 // K-Score (strikeout plays only) } // Interpretation: // Model says fair value is +120 (42.1% implied probability) // DK is offering +148 (40.3% implied probability) // Book is paying MORE than fair value → +EV OVER play

The K-Score

For strikeout props, a K-Score appears next to each pitcher. It's a 0–100 composite of stuff quality, swinging strike rate, opposing lineup K rate, and umpire tendency.

K-Score RangeSignalApproach
80–100Elite K upsideStrong OVER candidate — check line relative to model odds
70–79Above averageGood OVER candidate if model odds show value
60–69NeutralRequires better-than-average model edge to bet
50–59Slight lean UNDERConsider UNDER if book odds are favorable
Below 50Clear UNDER leanFade this pitcher's K line if book gives value

Stat Definitions in the App

EV (Exit Velocity)
Average speed off the bat in mph. Displayed on batter cards. 92+ is solid; 96+ is elite power.
Barrel%
Percentage of balls hit with both high exit velocity and ideal launch angle. The single best predictor of HR production. 12%+ is strong; 18%+ is elite.
xSLG
Expected slugging percentage based on contact quality. Higher xSLG = more true power. The primary lambda input in our HR model.
SwStr%
Swinging strike rate for pitchers. The single best K predictor. 15%+ is elite. The largest component of the K-Score.
Park Factor
How much a stadium boosts or suppresses HR production relative to neutral (100). Displayed on weather cards and baked into every model odds calculation.

Frequently Asked Questions

When does the model update? +
6 times daily: 8AM, 10AM, 12PM, 2PM, 5PM, and 8PM CT. Each run fetches updated lineups, weather, and pitcher changes, then reruns all Poisson calculations.
What does "expected stat" mean? +
The expected stat is the Poisson lambda — our model's estimate of the average number of events (HRs, Ks, total bases) the player will produce in today's game. It's the key input to the probability calculation.
Why does the model sometimes show picks I shouldn't bet? +
The model surfaces all plays where our model odds are more favorable than the book's implied probability. That doesn't mean every play has massive edge — the /model/top list is pre-filtered for the highest expected stat plays. Use your judgment on line size and edge magnitude.
Can I see historical model performance? +
Historical picks tracking is on our roadmap. For now, the model's inputs (Statcast, xSLG, barrel%) have well-established predictive value in the academic literature. The model is generating probabilities — not predictions. Betting based on true probability, not outcome prediction, is how long-term edge is built.
Should I parlay model picks? +
No. Parlays compound the house edge, not your analytical edge. Each leg of a parlay is priced at the sportsbook's odds, not our model's fair-value odds. Flat betting straight bets is the mathematically correct approach.

READY TO OPEN THE APP?

$29/mo on MLB if you get access for $29/mo, 2026.

Open PropBetEdge →