Pitching statistics tell you how well a pitcher prevents runs and how much of that success he controls himself. ERA (earned run average) and WHIP (walks plus hits per inning) are the classic starting points. Strikeout rate, walk rate and FIP (fielding independent pitching) go a step further by isolating the outcomes a pitcher owns without help from his defense. Read them together and you get a far clearer picture than any single number can give.
What Pitching Statistics Actually Measure
Every pitching stat answers one of three questions. First, how many runs did the pitcher allow? Second, how often did he put runners on base? Third, which outcomes did he control directly, such as strikeouts, walks and home runs, versus outcomes shared with his fielders, such as ground balls that find a hole or a line drive that an outfielder runs down?
Core pitching stats at a glance
| Stat | Formula | Rough MLB benchmark |
|---|---|---|
| ERA | (Earned runs x 9) / innings pitched | League average usually around 4.00; under 3.00 is excellent |
| WHIP | (Walks + hits) / innings pitched | Average near 1.30; under 1.10 is excellent |
| K/9 | (Strikeouts x 9) / innings pitched | Average near 8.5; 10 or more is elite |
| BB/9 | (Walks x 9) / innings pitched | Average near 3.2; under 2.5 is very good |
| K% | Strikeouts / batters faced | Average around 22 percent; 28 percent or more is elite |
| BB% | Walks / batters faced | Average around 8 to 9 percent; under 6 percent is excellent |
| K/BB | Strikeouts / walks | About 2.5 is average; 4.0 or more is excellent |
| FIP | ((13 x HR) + (3 x (BB + HBP)) – (2 x K)) / IP + constant | Scaled to read like ERA |
| ERA+ | ERA adjusted for park and league, indexed to 100 | 100 is average; 130 or more is excellent |
How Innings Pitched Are Written
Before you calculate anything, you need to read innings correctly. Each out is one third of an inning, and box scores record the thirds after a decimal point. So 6.1 innings means six and one third innings, and 6.2 means six and two thirds. There is no such thing as 6.3, because three outs complete the inning and the count rolls over to 7.0.
This matters because every rate stat divides by innings. If you type 6.1 into a calculator as six point one, you will get the wrong answer. Convert it to 6.333 first.
Run Prevention Stats: ERA, Earned Runs and ERA+
ERA is the number of earned runs a pitcher allows per nine innings. If a pitcher gives up 2 earned runs in 6.1 innings, his ERA for that outing is (2 x 9) / 6.333, or about 2.84.
The key word is earned. Under the official scoring rules, a run is unearned when it scores because of a fielding error or a passed ball that the pitcher should not be charged for. The official scorer reconstructs the inning as if the errors had not happened and charges the pitcher only for runs that would have scored anyway. Once an error extends an inning that should have ended, any run that scores afterward in that inning is unearned against that pitcher.
ERA+ corrects for park and league context. It sets the league average at 100, so an ERA+ of 125 means the pitcher was roughly 25 percent better than average after adjusting for his home park. It is the best quick way to compare pitchers across eras and parks.
Baserunner Stats: WHIP, K/9 and BB/9
WHIP counts how many batters reach by hit or walk per inning. A pitcher who allows 5 hits and 2 walks in 6.1 innings has a WHIP of 7 / 6.333, or about 1.11. It ignores hit batters and errors, but it is a clean measure of traffic. K/9 and BB/9 express strikeouts and walks per nine innings, though they can be inflated when a pitcher faces extra batters.
Advanced Stats: K%, BB%, FIP and BABIP
K% and BB% divide strikeouts and walks by total batters faced. They measure what happens in each plate appearance, so they are not distorted by how many runners reach. K% minus BB% (often written K-BB%) is one of the most predictive single numbers for pitcher quality.
FIP, popularized by analyst Tom Tango, estimates what a pitcher’s ERA would look like based only on home runs, walks, hit batters and strikeouts. It adds a league constant, usually a bit above 3.00, so that league average FIP matches league average ERA. When a pitcher’s ERA is far below his FIP, he may have benefited from strong defense or good fortune on balls in play, and some regression is likely.
BABIP, or batting average on balls in play, shows how often balls put in play fall for hits. League average is usually close to .290 to .300. Pitchers have some influence over BABIP, but a figure far from their career norm often signals luck or defense rather than a real change in skill.
Role Stats: Wins, Saves, Holds and Quality Starts
A starting pitcher earns a win when he pitches at least five innings, his team takes the lead while he is the pitcher of record, and the team never gives that lead back. In shortened games the minimum is lower. Because wins depend heavily on run support, modern analysts treat them as a weak measure of individual performance.
A save goes to the reliever who finishes a win without being the winning pitcher and meets one of three conditions: he enters with a lead of three runs or fewer and pitches at least one inning, he enters with the potential tying run on base, at bat or on deck, or he pitches at least three effective innings. A blown save is charged when a reliever enters in a save situation and lets the tying run score.
A hold is not in the official rulebook, but it is widely tracked. It credits a reliever who enters in a save situation, records at least one out and leaves with the lead intact without finishing the game.
A quality start is at least six innings with three or fewer earned runs.
Rulebook note: Official definitions for earned runs, wins, saves and innings come from Rule 9 of the Official Baseball Rules. Holds, quality starts and all advanced stats are unofficial. Amateur leagues and different stat providers may use slightly different constants and qualifying thresholds.
How to Read a Pitcher’s Stat Line in Five Steps
- Check the sample size. A reliever with 20 innings can post almost any ERA. Starters usually need a full season, and even then one year can mislead.
- Start with K% and BB%. A high strikeout rate and a low walk rate are the clearest signs of skill.
- Compare ERA to FIP. A gap of more than half a run suggests defense, luck or sequencing played a role.
- Look at BABIP and home run rate. An unusual figure in either one often explains an ERA that looks out of line.
- Add context with ERA+. Park and league adjustments show whether the raw numbers are truly special.
Common Mistakes When Reading Pitching Stats
- Treating 6.1 innings as 6.1. It means six and one third. Convert thirds before dividing, or every rate stat will be slightly wrong.
- Judging pitchers by wins. A strong pitcher on a weak offense can go 9 and 12 with a 3.00 ERA. Wins say as much about the lineup as the pitcher.
- Reading too much into small samples. One bad outing can wreck a reliever’s ERA for months. Look at strikeouts, walks and batters faced instead.
- Ignoring the park. A 4.00 ERA in a hitter-friendly park can be better than a 3.60 ERA in a spacious one. ERA+ exists for this reason.
- Assuming FIP is the whole truth. Some pitchers, especially those with elite soft contact skills, consistently beat their FIP. It is an estimate, not a verdict.
Frequently Asked Questions
What is a good ERA in baseball?
In recent MLB seasons, league average ERA has usually landed near 4.00. An ERA under 3.50 is very good for a starter, and under 3.00 is excellent. Relievers often post lower ERAs because they face hitters for shorter stretches. At lower levels, averages vary widely, so compare pitchers within the same league.
What is the difference between ERA and FIP?
ERA counts the earned runs a pitcher actually allowed. FIP estimates what his ERA should have been based only on strikeouts, walks, hit batters and home runs. ERA describes what happened; FIP is usually better at predicting what will happen next.
Which pitching stat is the most useful?
No single stat tells the whole story. For a quick read, K% minus BB% captures core skill, ERA+ gives context-adjusted results, and FIP bridges the two. Using all three together is more reliable than leaning on any one of them.
