📊 How to Read the Recommendations Table

What you’re looking at: each row is a bet the model likes. Use EV% to judge value and Model Prob to judge confidence.

Field Definitions

FieldMeaningHow to Use It
GameMatchup (home team listed second).Scan for all bets tied to a single game.
MarketML (moneyline), Spread, or Total.ML = who wins; Spread = margin; Total = combined points.
SideBet direction.“Home/Away” for ML/Spread, “Over/Under” for Total.
Model ProbModel’s probability that the bet wins.≥60% is typically strong.
Book OddsSportsbook price.Determines payout; shown in American odds.
Implied ProbProbability implied by the book odds. For American odds: +X → 100/(X+100), -X → |X|/(|X|+100)
Edge %Model vs. market gap.Edge = ModelProb − ImpliedProb (percentage points).
EV %Expected return per dollar staked. Convert American odds to decimal d: A>0 → d = 1 + A/100, A<0 → d = 1 + 100/|A|
Then EV% = 100 × (ModelProb × d − 1).
Stake $Suggested bet size (Kelly/fractional).Use as a cap; don’t exceed bankroll limits.
LineSpread or total being bet.Ex: -3.5, +2.5, o46.5, u41.
BookWhich sportsbook provided the price.Where to place the bet to match the edge.

Quick Interpretation

SignalWhat It MeansAction
High EV (≥5%) + High Prob (≥60%)Strong value and confidence.Bet-worthy if lines are current.
Moderate EV (2–4%)Edge exists but smaller.Consider when confidence ≥55%.
Low EV (<2%)Likely neutral after juice.Monitor for better price.
Negative EVMarket priced better than model.Skip or fade.

Worked Example

Inputs
Market: ML  |  Odds: +115
Model Prob: 0.58 (58%)
Implied Prob: 100/(115+100)=0.465 (46.5%)
Outputs
Edge %: 0.58 − 0.465 = +0.115 (+11.5 pts)
Decimal odds: 1 + 115/100 = 2.15
EV %: (0.58 × 2.15 − 1) × 100 = +24.7%
Rule of Thumb: prioritize rows with both higher EV% and strong Model Prob. Use Stake $ as an upper bound, not a mandate.