What Is a Quality Start? Pitcher Workload Metrics Explained
A quality start is six or more innings with three or fewer earned runs. This guide explains the definition, its origin, its worked arithmetic, its flaws and the other measures of how much a starting pitcher works.
A quality start is a start in which the pitcher goes at least six innings and allows no more than three earned runs. That is the whole definition. A pitcher who goes six and allows three has a quality start, and a pitcher who goes 5� � and allows none does not. The statistic was invented to answer a narrow question, did the starter give his team a fair chance to win, in a way that a pitcher's win-loss record could not.
The idea has stuck for forty years because it is easy to state and easy to count, and because it rewards something managers want: a starter who goes deep without leaving the game in the hands of his bullpen. It has also drawn criticism almost since it was invented. A quality start can have an earned run average of 4.50, which is not good in many seasons, while a start that is better in every way but one inning short does not count. This article covers the definition, the arithmetic, where the statistic came from, what it hides and the other measures that describe how much a starting pitcher works.
The definition, exactly
To count as a quality start, a pitcher must meet two conditions in a single game he starts. He must pitch at least six full innings, which means recording at least 18 outs. And he must allow no more than three earned runs. The two conditions are independent, and both must be met.
Earned runs are the key term. An earned run is one that scores without the help of an error or a passed ball, as determined by the official scorer under a set of rules that rebuild the inning as if the defense had made no mistakes. A pitcher who allows four runs in six innings, of which one is unearned, has allowed three earned runs and has a quality start. A pitcher who allows three runs in 5� � innings has not, no matter how well he pitched.
The statistic is a count, not a rate. A pitcher's total for a season is the number of starts in which he met the standard. You can turn it into a rate by dividing by games started, which gives a quality start percentage. It is tracked on most statistical sites, including those that carry MLB data, though it has never been part of the traditional set of official counting categories in the way that wins, saves and strikeouts are. Treat it as a widely used convention.
Where it came from
The statistic is credited to John Lowe, a sportswriter at the Philadelphia Inquirer, who proposed it in 1985. His starting point was a familiar complaint about the win. A pitcher gets a win if he leaves the game with his team ahead and the lead holds, provided he has gone at least five innings. That makes the win depend heavily on how many runs his team scores and on what the bullpen does after he leaves. A starter can allow one run in eight innings and get a no-decision, and another can allow six runs in five and get a win.
Lowe wanted a measure that described the starter's own part of the game. He chose six innings and three earned runs as a reasonable standard for a game in which the starter had done his job. He chose the thresholds as a simple bar for a start that gives the team a fair chance to win. Whether they were exactly right is arguable. They were reasonable, simple and memorable, and that is what made them last.
Worked examples
The arithmetic is simple, but a table of lines makes the edges visible. Earned run average is 9 times earned runs divided by innings pitched.
| Line | Innings | Earned runs | ERA for the game | Quality start? |
|---|---|---|---|---|
| A | 6.0 | 3 | 4.50 | Yes |
| B | 5� | |||
| � | 2 | 3.18 | No (one out short) | |
| C | 8.0 | 4 | 4.50 | No (four earned runs) |
| D | 7.0 | 3 | 3.86 | Yes |
| E | 9.0 | 3 | 3.00 | Yes |
| F | 6.0 | 0 | 0.00 | Yes |
Lines A and C show the central quirk. Both have an earned run average of exactly 4.50 for the game, and one counts and the other does not. The starter who went eight innings and allowed four runs saved his bullpen two more innings than the one who went six, but he gets no credit. Line B shows the opposite edge: a start of 5� � innings and two earned runs, which is excellent by the rate, misses by one out.
The minimum quality start therefore has an earned run average of 4.50. Whether that is good depends on the setting. In 1919, when major-league teams averaged 3.87 runs per game, a 4.50 start would have been poor. In 1995, with 4.85 runs per team game, it was around the league average. Runs per game is only a rough guide to earned run average, but the comparison shows that a fixed standard means different things in different years.
What the statistic does well
Its main virtue is that it separates the starter from his team. A pitcher on a weak offense can have a poor win-loss record and a long list of quality starts, and the second figure is the better indicator of how he pitched. A pitcher on a strong offense can collect wins with ordinary outings. A quality-start total reduces that noise.
It also rewards durability, in a limited way. Getting through six innings is a minimum standard for an effective starter, and a quality start requires it. A pitcher with high strikeout totals and an unusually short outing every night will accumulate fewer.
It is also easy to explain. A fan can follow the standard without a formula, which is a real advantage in broadcasts, box scores and fantasy leagues. Some fantasy formats use quality starts as a category in place of wins for exactly this reason, because they are less dependent on run support. The fantasy section of this site discusses scoring formats in more detail.
What it hides
The weaknesses are as clear as the strengths. The first is the cliff edge. Because the standard is a pair of thresholds, a start just inside the line counts the same as a start far better. A pitcher who goes six innings and allows three earned runs gets the same credit as one who goes nine and allows none. That flattens the best starts into the average ones and does not reward the pitcher who goes eight or nine, which is the most valuable work a starter can give his bullpen.
The second is that it can be earned cheaply. A start of six innings and three earned runs, repeated, produces a 4.50 earned run average. A pitcher who does it every time would be counted as having a perfect quality-start rate and would pitch at an ordinary level. Writers call this a "cheap" quality start. The label has no official definition, but it describes a real pattern: the statistic can look better than the pitching.
The third is that it ignores everything except innings and earned runs. Strikeouts, walks, home runs, hard-hit balls and the quality of opposing batters do not enter. A pitcher who walks five and escapes with three earned runs in six innings is credited the same as one who strikes out nine.
The fourth is that it depends on the scoring rules in ways that are not in the pitcher's control. The unearned-run rule means an error can erase a run from the count, but a starter who leaves with runners on base is charged with the earned runs those runners score after a reliever enters, which can turn a quality start into a non-quality one after the starter has left the game. It also depends on a manager's decision about how long to leave the pitcher in.
Workload metrics: how much does a starter really work?
The quality start sits in a family of measures that describe workload. The most basic is innings pitched, written with the number of outs as a decimal: 6.1 means six innings and one out, and 6.2 means six and two outs. Innings per start is that figure divided by games started, and it is the simplest way to see how deep a pitcher goes. Batters faced counts every plate appearance a pitcher handles, including those that did not end in an out, and so gives a fuller picture of work than innings alone.
Pitch counts are the modern addition. Before the era of detailed tracking, pitch totals were recorded inconsistently, and the historical record does not carry them. Clubs now track pitches per start, pitches per inning and the number of times a pitcher faces the order a third time. The widely used 100-pitch mark is a convention, not a rule, and managers treat it as a rough guide rather than a limit.
The way the job has changed is visible in the long-run totals. Babe Ruth, who pitched mostly for Boston, started 148 games and completed 107 of them, 72 percent. Greg Maddux, among the most durable starters of the late twentieth century, started 740 games and completed 109, about 15 percent. Complete games have become much rarer since. A six-inning standard that seemed modest in 1985 asks more of a modern starter than it did then, because starters now work fewer innings per start.
The shift also helps explain why quality-start totals have fallen over the long run: with starters leaving earlier, fewer starts reach the sixth inning. Some analysts respond by preferring stricter standards or rate measures instead.
Other ways to grade a single start
Game Score, developed by Bill James, scores a start on a single number using a formula. A pitcher begins at 50 points and adds 1 for each out recorded, 2 more for each inning completed after the fourth, and 1 for each strikeout. He subtracts 2 for each hit, 4 for each earned run, 2 for each unearned run and 1 for each walk. A score of 50 is average, and the best starts reach 90 or higher.
Applied to the lines above, the formula shows what the quality-start standard misses. Take a quality start of six innings with three earned runs, five hits, two walks and five strikeouts: 50 plus 18 for the outs, plus 4 for the two innings after the fourth, plus 5 for the strikeouts, minus 10 for the hits, minus 12 for the earned runs and minus 2 for the walks gives 53. Now take an eight-inning start with four earned runs, seven hits, one walk and eight strikeouts, which is not a quality start: 50 plus 24, plus 8, plus 8, minus 14, minus 16 and minus 1 gives 59. By Game Score the second start was the better one, which matches the intuition that eight innings with eight strikeouts deserves credit.
A third start, nine innings with no earned runs, three hits, no walks and ten strikeouts, scores 91. The simple standard counts all three outings as quality starts or not, and the formula separates them. That is the advantage of a graded measure over a threshold. The disadvantage is that no one can calculate a Game Score in their head during a game.
Other tools measure different things. Earned run average (ERA) divides earned runs by innings and multiplies by nine. WHIP, walks plus hits per inning pitched, measures baserunners allowed. Strikeouts per nine innings and walks per nine describe a pitcher's style. The pages explained in how to read a player's statistics page define each column. For the older history of what box scores recorded, see what old box scores tell us and what they do not.
How the statistic is used today
Broadcasts and box-score sites use quality starts as shorthand for a good outing. Fantasy leagues use them as a category. Coaches use innings and pitch counts to plan workloads, and analysts use rate measures such as strikeouts minus walks per batter faced when they want a better predictor.
The statistic has also entered the language of betting markets. A sportsbook may offer prop bets on a pitcher's strikeouts or outs recorded, which depend on exactly how deep a starter is expected to go. Gambling carries risk, is for adults of legal age where it is legal, and nothing in a workload statistic guarantees a result. If gambling is causing you or someone you know harm, the national helpline is 1-800-GAMBLER, and our guide to responsible gambling and baseball lists limits, warning signs and where to get help.
Many teams think of workload in terms of the third time through the order. Batting results tend to improve for hitters with each look at the same pitcher, and that pattern is one reason managers pull starters earlier than the old standard. The connection between starters, batting order and bullpen is worth studying. The article on lineup construction and the math of the batting order explains how the order affects how often each hitter comes to the plate.
Common questions
Is a quality start the same as a good start? Not necessarily. It is a minimum standard. A six-inning, three-earned-run start counts, and so does a nine-inning shutout.
Does a quality start guarantee a win? No. A starter can allow three earned runs and lose 3-2, or allow none and get a no-decision if the game is tied when he leaves.
Are unearned runs counted? No. Only earned runs count against the three-run limit, though the runs are still real in the score.
Who invented the quality start? John Lowe, a sportswriter at the Philadelphia Inquirer, proposed it in 1985.
Why are quality starts becoming rarer? Because starters throw fewer innings per game than they did a generation ago, and the six-inning requirement filters out more starts than it used to.
Where to go next
The quality start sits alongside other ways of reading a game and a career. Batting average, on-base percentage and slugging explained covers the hitting side, and the three true outcomes shows how strikeouts, walks and home runs have reshaped the work of starters and relievers. Our leaders pages rank pitchers by innings, wins and earned run average across baseball history.