What matters
Most baseball player outcomes reduce to opportunities × rate. Opportunities are plate appearances (for hitters) or batters faced and pitch count (for pitchers). Rates come from skill, matchup and environment. Lineup position alone can move a hitter's expected plate appearances by close to a full trip.
Important data
| Input | Why it matters |
|---|---|
| Plate appearances / lineup slot | Sets opportunities — confirmed lineups change probabilities |
| Contact quality (exit velocity, launch angle, barrels) | More predictive of future power than past home run totals |
| Pitcher profile (stuff, CSW%, handedness) | Strikeout and contact suppression |
| Park and weather | Modest, real adjustments to carry and run scoring |
| Umpire | Small effect on called strikes and walks |
Common metrics
| Metric | What it measures | How to read it |
|---|---|---|
| Barrel rate | Share of batted balls barreled | Power indicator; stabilises faster than HR totals |
| xwOBA | Expected quality from contact, strikeouts and walks | Compare with actual wOBA for luck |
| CSW% | Called strikes + whiffs / pitches | Strikeout indicator for pitchers |
| Park factor | Park effect vs league average (100) | Context, not a switch |
| Base-out state | Runners and outs | Sets run expectancy for the inning |
Model concepts
Home runs are rare, countable events, so a Poisson-style "at least one" calculation from an expected rate is a sensible starting point — with the caveats in the Poisson lesson. Strikeout totals are counts too, but a starter's exit time caps them, so the distribution is narrower at the top than a pure Poisson suggests.
Live intelligence
Baseball's live state updates every pitch: count, outs and runners. PBEcast shows the diamond, the live pitch, the batter's split against the current pitcher's hand and the pitcher's arsenal from this game.
PropBetEdge MLB tools
HR Targets ↗
Model-ranked home run probabilities for the slate.
MLBPBEcast ↗
Live game view: scoreboard, base-out state, live pitch, batter splits, pitcher arsenal.
MLBK Props ↗
Strikeout prop research.
MLBTrack Record ↗
Graded history of published MLB calls.
MLBPark Factor ↗
Park environment context.
MLBWeather ↗
Ballpark weather context.
Worked example
- A hitter homers on about 3.5% of plate appearances; batting second he projects for about 4.4 PAs. λ ≈ 0.035 × 4.4 = 0.154.
- P(≥1) = 1 − e^(−0.154) = 14.3% → fair about +601.
- Dropped to 8th (about 3.8 PAs): λ = 0.133 → 12.5% → fair +703.
Glossary
Related lessons
Poisson
How do you turn "0.35 expected home runs" into a probability — and when is that math wrong?
HR Targets
A batter is a top HR Target at 22%. What does that mean?
PBEcast & live casts
What am I looking at on a live cast, and what is it telling me?
Pitch · possession · round
What is the right "tick" to think about live, sport by sport?
Sample size
How many results before a record means something?