Every metric we show — in plain language.
These are the advanced baseball metrics behind the SportsEdgeData research tools: what each one measures, and how to read it at a glance.
Skill-based metrics that strip out luck and defense to measure a pitcher’s true talent.
Estimates a pitcher’s ERA from the outcomes he controls (strikeouts, walks, ground balls) and how they interact. The most predictive of the ERA estimators — it sees through lucky or unlucky stretches.
How to read Lower is better. ≤3.50 elite · ~4.20 average · ≥4.50 weak.
Like FIP but normalizes home runs to a league-average rate, so a pitcher isn’t punished (or rewarded) for a flukey HR count. Great for projecting forward, and the backbone of our matchup Edge.
How to read Lower is better. ≤3.50 strong · ≥4.50 vulnerable.
What a pitcher’s ERA "should" be based only on strikeouts, walks, hit-by-pitch and home runs — the events the defense can’t influence. Scaled to look like ERA.
How to read Read like ERA — lower is better.
Share of batters faced that the pitcher strikes out. A cleaner skill signal than K/9 because it isn’t distorted by how many baserunners reach.
How to read Higher is better. ≥27% elite · ≤18% weak.
Share of batters faced that the pitcher walks. Low walk rates keep the bases clean and innings short.
How to read Lower is better. ≤6% excellent · ≥10% shaky.
K% minus BB% in one number — the cleanest, most stable measure of pitcher dominance. Stabilizes faster than ERA over a season.
How to read Higher is better. ≥18% strong · 12–18% solid · ≤8% weak.
Share of batted balls hit on the ground. Ground balls rarely leave the park, so high-GB pitchers suppress home runs and extra-base damage.
How to read Higher suppresses power. ≥50% strong · ≥40% solid.
How often a pitcher gives up the long ball, per nine innings. Especially telling in hitter-friendly parks.
How to read Lower is better. ≤0.8 strong · ≥1.4 homer-prone.
The classic box-score numbers — useful, but more luck-dependent than the advanced metrics above.
Earned runs allowed per nine innings. Familiar, but heavily influenced by defense and sequencing — when ERA and SIERA disagree by a lot, trust SIERA.
How to read Lower is better. Pair it with SIERA to spot luck.
Baserunners allowed per inning. A quick read on how much traffic a pitcher puts on the bases.
How to read Lower is better. ~1.00 excellent · ~1.40 hittable.
Strikeouts and walks expressed per nine innings. Handy rate stats; K% and BB% are the more reliable versions.
How to read High K/9 good · low BB/9 good.
Total innings thrown — also the sample size behind every rate stat. Below ~30 IP, advanced numbers are noisy and should be regressed toward league average.
How to read Context for trust — more IP = more reliable.
Leverage, role and workload — the context that decides who actually pitches the high-stakes innings.
A reliever’s outings graded by win-probability swing: a Shutdown raised his team’s win odds by ≥6%, a Meltdown dropped them by ≥6%. A results-in-context view of clutch.
How to read SD/(SD+MD): ≥65% reliable · <40% volatile.
Where a reliever sits in his manager’s trust hierarchy (rank / total arms used), inferred from when he enters and the score situation. Rank 1–2 = high-leverage arm.
How to read Lower rank = more trusted / higher leverage.
A 1–30 ranking of each team’s bullpen exposure tonight, blending top-arm availability, rested SIERA and shutdown/meltdown depth. Surfaces which pens are stretched thin.
How to read Higher rank = more exposed bullpen.
For a starter, the share of starts in which he allowed no run in the 1st inning. A snapshot of how often he opens cleanly — best paired with the quality of opposing leadoff hitters.
How to read ≥65% strong · 50–65% average · <50% leaky.
A fatigue flag from recent usage — 🟢 Rested, 🟡 Recent, 🔴 Overworked — driven by pitches over the last 3 days and days of rest. Tired arms lose stuff late.
How to read 🟢 fresh · 🔴 use caution (>50 pitches in 3d).
Live tracking of fastball speed vs the pitcher’s baseline as the game goes on. A meaningful drop (≥ ~1.5 mph) signals fatigue and a tiring pitcher who’s about to lose effectiveness.
How to read ≥-1.5 mph = significant decay (red flag).
Live pitch count vs the pitcher’s season average per outing. As he approaches and passes his norm, the manager’s hook — and a bullpen handoff — gets closer.
How to read ≥95% of avg = yellow · ≥110% = red (likely pulled).
How much real offense a hitter produces — park- and league-adjusted where it counts.
The single best all-in-one hitting metric. It captures total offensive value and adjusts for park and league, so 100 = exactly league average and every point above/below is a percent better/worse. A 130 hitter is 30% above average.
How to read 100 = average · ≥130 elite · ≥105 above avg · <90 below.
Like on-base percentage, but each way of reaching base is weighted by its real run value (a homer counts far more than a walk). The input behind wRC+.
How to read Higher is better. ≥0.370 strong · ≤0.300 weak.
How often a hitter reaches base per plate appearance — the cleanest single measure of "doesn’t make outs." Drives rallies and run scoring.
How to read Higher is better. ≥0.360 strong · ≤0.310 weak.
Slugging minus batting average — strips out singles to measure raw extra-base power alone. High ISO = real over-the-fence and gap threat.
How to read Higher is better. ≥0.200 strong power.
Average on balls put in play (excluding homers and strikeouts). Useful as a luck gauge: a number far from ~.300 often regresses, hinting a hitter is running hot or cold.
How to read ~.300 is typical; far from it = luck at play.
How often a hitter strikes out vs draws a walk. Low K% + high BB% = a disciplined bat that controls the strike zone.
How to read K% ≤18% / BB% ≥10% = excellent discipline.
The classic triple: batting average (hits per at-bat), slugging (total bases per at-bat) and OPS (on-base + slugging). Quick and familiar, but wRC+ is the park-adjusted upgrade.
How to read Higher is better; OPS ~.800+ is a good bat.
Where the edge usually lives — how a specific hitter or lineup matches up against tonight’s pitcher.
A hitter’s (or pitcher’s) numbers broken out against left- vs right-handed opponents. Many players are dramatically better one way — the split often matters more than the overall line.
How to read Compare to the opposing hand for tonight’s read.
Which side a hitter bats from: L (left), R (right) or S (switch). Switch hitters bat opposite the pitcher’s hand, so they rarely face a platoon disadvantage.
How to read S avoids the platoon disadvantage by design.
Our matchup score for a hitter against tonight’s pitcher. It combines the batter’s wRC+ versus the pitcher’s hand AND the pitcher’s xFIP versus the batter’s side, on one centered scale. Positive = batter advantage.
How to read ≥+20 strong batter edge · ≤-20 strong pitcher edge.
The starter’s xFIP re-weighted for tonight’s specific opposing lineup and its handedness mix — a forward-looking read on how the matchup tilts.
How to read Lower = pitcher edge · higher = lineup edge.
The lineup’s PA-weighted average wRC+, shown overall and split vs RHP / vs LHP, computed over hitters with a real sample (≥30 PA). Tonight’s opposing-starter hand is highlighted.
How to read >100 = above-average attack; check the right hand.
The environment a game is played in — and the reliability of the numbers behind it.
A 5-year rolling measure of how a ballpark inflates or suppresses runs and home runs vs a neutral park. Above 100 favors offense; below 100 favors pitching.
How to read >100 hitter-friendly · <100 pitcher-friendly.
Live conditions that nudge run environment: wind blowing out and warm, humid air help the ball carry; wind in and cold air knock it down. Roof status can neutralize all of it.
How to read Wind out + warm = more offense; roof = neutral.
An asterisk (*) marks a number built on too few events to trust — under ~30 plate appearances for hitters, or thin innings/batters-faced for pitchers. Extreme small-sample values often regress hard.
How to read See a * ? Treat the number as not-yet-reliable.