WHIP in Baseball: The Formula and What a Good WHIP Is
WHIP in baseball is walks plus hits per inning pitched. Here is how MLB pitchers are measured by it, what a good WHIP looks like and what the stat leaves out.
By CricketTaken EditorialPublished Baseball Stats & Positions27 min read
- Formula
- (Walks + hits) / innings pitched
- Invented
- 1979, by Daniel Okrent
- Lowest AL or NL season
- Pedro Martinez, 0.737 in 2000
- Lowest career mark
- Addie Joss, 0.968
Seven innings, five hits, two walks. Add the hits to the walks, divide by the innings, and the starter's WHIP for the night is exactly 1.00: one batter reaching base per inning. Do that for six months in Major League Baseball and a pitcher is in the conversation for the Cy Young Award.
WHIP stands for walks plus hits per inning pitched. It is the number of batters a pitcher puts on base by hit or by base on balls in an average inning, and the formula is as short as the name: walks plus hits, divided by innings. Lower is better. Across MLB the average sits close to 1.30. A full season below 1.10 is excellent, below 1.00 is rare, and a pitcher running 1.50 or higher is in trouble. Hit batters and runners who reach on errors are not counted.
What WHIP stands for and what the number measures
The acronym is the definition. W for walks, H for hits, IP for innings pitched, with the P squeezed in to make it pronounceable. Announcers say it as a word, "whip," and it is printed to two decimal places on a scoreboard and to three in the record book.
What it measures is traffic. Every pitcher's first job is to keep hitters off the bases, and WHIP counts how often he fails at it by the two most common routes. A figure of 1.25 says that in a typical inning he allows a runner and a quarter, which in practice means one man on in most innings and two on in some.
Unlike most pitching rates, WHIP is scaled to a single inning. Earned run average is stated per nine innings, and so are strikeout and walk rates. To put WHIP on the same footing, multiply it by nine: a 1.30 becomes 11.7 baserunners over a full game, a 1.00 becomes nine.
That per-inning scale is part of its appeal. The picture is immediate. Three up and three down is 0.00 for the inning. A leadoff walk, erased or not, makes it 1.00. Bases loaded is 3.00, whatever happens next.
A few features set it apart from the older statistics.
It ignores runs entirely. A pitcher can load the bases with nobody out and strike out the side; his WHIP for the inning is 3.00 and no run has scored. The reverse case, a solo home run and three quick outs, is a WHIP of 1.00 and a run on the board.
It needs no judgment about earned and unearned runs, though the scorer's hit-or-error decision still feeds it, since a ball ruled a hit goes in and one ruled an error stays out.
And it is not a rulebook statistic. Rule 9.21 of the Official Baseball Rules sets out how to compute a handful of averages, among them batting average, slugging percentage and earned run average. WHIP is not on the list. MLB publishes it on its stat pages and defines it in its glossary, clubs print it in their game notes, and every broadcaster uses it, but it reached that standing by habit, not by legislation.
How WHIP is calculated, for one start and for a season
WHIP = (walks + hits) ÷ innings pitched
Two additions and a division. The only place to go wrong is the innings.
A single start
A pitcher works six innings, gives up seven hits and walks one. Seven plus one is eight. Eight divided by six is 1.33.
Now a start that ends mid-inning. The line reads 5.2 innings, six hits, three walks. The ".2" is two outs, not two-tenths, so the innings are five and two-thirds, or 5.667. Nine baserunners divided by 5.667 is 1.59. Type 5.2 into the calculator by mistake and the answer is 1.73, which is wrong by enough to matter. The thirds convention is explained at more length on our page about how ERA is worked out; the same trap applies here.
The safest method is to count outs. Multiply the hits and walks by three and divide by outs recorded. Seventeen outs, nine baserunners: 27 divided by 17 is 1.59.
A season
Nothing changes except the size of the numbers. Take a starter with 165 hits and 55 walks in 180.1 innings. That is 220 baserunners in 180 and one-third innings: 220 divided by 180.333 is 1.22.
Pedro Martinez's 2000 season with the Boston Red Sox is the cleanest real example, because his innings came out whole. He allowed 128 hits and 32 walks in 217 innings. One hundred sixty divided by 217 is 0.737.
| Pitching line | Hits | Walks | Innings as a decimal | WHIP |
|---|---|---|---|---|
| 7.0 IP, 5 H, 2 BB | 5 | 2 | 7.000 | 1.00 |
| 6.0 IP, 7 H, 1 BB | 7 | 1 | 6.000 | 1.33 |
| 5.2 IP, 6 H, 3 BB | 6 | 3 | 5.667 | 1.59 |
| 9.0 IP, 2 H, 0 BB | 2 | 0 | 9.000 | 0.22 |
| 0.2 IP, 3 H, 1 BB | 3 | 1 | 0.667 | 6.00 |
| 217.0 IP, 128 H, 32 BB | 128 | 32 | 217.000 | 0.737 |
The fifth row is a reliever's bad night: two outs, four baserunners, a WHIP of 6.00 for the appearance. If he fails to record an out at all, the figure is undefined for that game, since nothing can be divided by zero innings. It becomes a real number again with his next out.
Season WHIP is never the average of game WHIPs. A 0.50 over eight innings and a 6.00 over two-thirds of an inning do not average to 3.25; the pitcher allowed eight baserunners in eight and two-thirds, and his WHIP across the two games is 0.92. Always add up the raw totals first and divide once.
What the count leaves out: hit batters, errors and intentional walks
A batter drilled in the ribs trots to first base, and the pitcher's WHIP does not move. That surprises people more than anything else about the statistic.
The reason is historical, and slightly comic. When Daniel Okrent devised the measure in 1979 he needed something his friends could compute from the Sunday newspaper, and the agate box scores of the day did not list hit batsmen. So they were left out. They have stayed out ever since.
| How the batter reaches base | Counted in WHIP? | Why |
|---|---|---|
| Single, double, triple or home run | Yes | All are hits, and each counts once |
| Unintentional walk | Yes | A base on balls |
| Intentional walk | Yes | Scored as a base on balls like any other |
| Hit by pitch | No | Neither a hit nor a walk |
| Fielding or throwing error | No | The scorer charges the fielder, not the pitcher |
| Fielder's choice | No | The batter is safe but another runner is put out or could have been |
| Catcher's interference | No | The batter is awarded first base with no hit or walk recorded |
| Dropped third strike | No | Scored as a strikeout plus a wild pitch, passed ball or error |
| Automatic runner in extra innings | No | He is placed on second base and never bats |
The intentional walk is the entry that draws complaints. Since 2017 a manager has been able to order one with a signal from the dugout, no pitches thrown, and the pitcher is charged with a walk he had no part in choosing. It goes into his WHIP all the same. The defense of the rule is simple enough: the batter is on base, and the pitcher's team put him there.
The omission of hit batters matters more for some pitchers than others. A starter who works inside and hits 15 batters across 180 innings is carrying 15 runners his WHIP does not admit to. Count them and his figure would be about 0.08 higher. For the mechanics of the award itself, see our guide to the hit by pitch rule; the rarer award for a catcher's mitt touching the bat is covered under catcher's interference.
Errors are a subtler case. A ground ball that the shortstop boots is no hit and no walk, so it never touches WHIP. But the decision between "hit" and "error" belongs to the official scorer, and a generous home scorer who calls a hard smash an error has, in effect, lowered his own pitcher's WHIP. The effect is small. It is not zero.
One more wrinkle concerns pitching changes in the middle of an at-bat. Under Rule 9.16(h), if a reliever comes in with the count already in the batter's favor at 2-0, 2-1, 3-0, 3-1 or 3-2 and the batter goes on to walk, the walk is charged to the pitcher who left. Anything else the batter does, a hit included, is charged to the reliever.
What a good WHIP is in MLB
No rulebook or league office grades WHIP. The scale below is the one in common use for a starting pitcher over a full major league season.
| WHIP | Rating | Baserunners per nine innings |
|---|---|---|
| Below 1.00 | Elite | Fewer than 9.0 |
| 1.00 to 1.10 | Excellent | 9.0 to 9.9 |
| 1.11 to 1.20 | Very good | 10.0 to 10.8 |
| 1.21 to 1.35 | Around average | 10.9 to 12.2 |
| 1.36 to 1.49 | Below average | 12.2 to 13.4 |
| 1.50 and above | Poor | 13.5 or more |
The gaps look small on paper and are large on the field. The difference between 1.10 and 1.30 is about two baserunners a game. Across 30 starts that is something like 35 or 40 extra men on base, a share of whom will score.
Sub-1.00 seasons deserve the word elite. In most years only a few qualified starters across both leagues finish there, and in some years none do. A pitcher who does it once has had the season of his life; the list of those who have done it repeatedly is a short roll of Hall of Famers and near-certain future ones.
At the other end, 1.50 is close to a hard limit. A starter putting three men on base every two innings is pitching from the stretch constantly, throwing more pitches per inning, and rarely seeing the sixth. Teams tolerate it only briefly.
The average moves, though less than ERA's does
League-wide WHIP follows the offensive climate. It was about 1.20 across the majors in 1968, the lowest-scoring season of the modern era, and climbed to nearly 1.50 at the peak of the hitting boom around 2000. In the past decade it has mostly sat between 1.25 and 1.35. A 1.30 is a fair working estimate of average, with the understanding that the true figure shifts a few hundredths from year to year.
Relievers, taken as a group, do not beat starters on this measure the way they do on ERA. They allow fewer hits per inning, working short stints at full effort against hitters who see them once, but they walk more batters, and in recent seasons the league's bullpens have finished a few hundredths above its rotations.
WHIP vs ERA: traffic against damage
The two statistics sit side by side on every graphic and answer different questions. WHIP asks how many men reached. ERA asks how many of them scored. Runners are a cause and runs are an effect, and most of the time the two agree: the pitchers with the lowest WHIP in a league are, with few exceptions, the ones with the lowest ERA.
The interesting cases are the disagreements.
| WHIP | ERA | What is usually going on |
|---|---|---|
| Low | Low | Few runners and few runs; the real thing |
| Low | High | Not many baserunners, but the hits are home runs or come in bunches; hit batters and stranded-runner luck can also be involved |
| High | Low | Constant traffic that has not scored yet: double plays, strikeouts with men on, or good fortune that rarely lasts |
| High | High | Too many runners and too many runs; no mystery |
Low WHIP, high ERA. The usual culprit is the home run. A pitcher who allows one hit an inning looks fine by WHIP, but if a quarter of those hits leave the park he is giving up runs at a rate the traffic count cannot see. The second cause is sequencing. Three hits scattered across three innings may cost nothing; the same three hits in a row cost one or two runs. WHIP is identical in both cases.
High WHIP, low ERA. This is the pitcher announcers describe as working out of trouble. Sometimes it is a skill. A sinkerballer who allows singles and walks but gets ground-ball double plays can live with a 1.35 for years, and a strikeout pitcher can leave runners at third. More often it is a warning. A pitcher allowing 13 baserunners per nine with a 3.00 ERA has been stranding runners at a rate almost nobody sustains, and the ERA usually rises to meet the WHIP, not the other way round.
That last point is the practical use of setting the two side by side. Over a few months WHIP tends to be the steadier figure, because it draws on every batter faced and not only the ones who came around to score. When the pair diverge sharply in June, the better bet for August is on the WHIP.
There are things ERA catches that WHIP misses altogether. A runner who reaches on an error and scores is unearned, so he appears in neither. A hit batter who scores is an earned run with no trace in WHIP. And a wild pitch that moves a runner from first to third changes the likely damage without changing the traffic count at all.
The flaw in the formula: every baserunner weighs the same
A four-pitch walk. A bunt single that dies on the third-base line. A home run into the second deck. WHIP scores all of them as one.
This is the statistic's largest blind spot, and it follows directly from the arithmetic. The numerator counts events. It does not ask how many bases the batter took or whether anyone was already on.
Consider two invented starters, each going seven innings and each allowing nine baserunners.
| Seven-inning line | Pitcher A | Pitcher B |
|---|---|---|
| Singles | 7 | 0 |
| Doubles | 0 | 3 |
| Home runs | 0 | 4 |
| Walks | 2 | 2 |
| Hits plus walks | 9 | 9 |
| Total bases on hits | 7 | 22 |
| WHIP for the game | 1.29 | 1.29 |
Pitcher A has had an ordinary night and probably given up two runs. Pitcher B has been hammered and is likely out of the game with six or seven on the board. The statistic cannot tell them apart.
The home run produces a second oddity. It is counted as a baserunner although the batter never stands on a base for the next hitter to drive in. From the point of view of traffic it clears the bases. From the point of view of the scoreboard it is the worst outcome available.
The flaw runs the other way too. WHIP's denominator is innings, which is to say outs, and it does not care how the outs were made. A pitcher who gives up a leadoff single and then gets a double-play grounder has allowed one baserunner and recorded two outs. A pitcher whose catcher throws out a runner stealing gets a third of an inning for it. Ground-ball pitchers and those with strong-armed catchers therefore post slightly better WHIPs than their rate of letting men on would suggest, because some of their baserunners are converted into outs.
Nor does it distinguish a walk from a hit in the way hitters and managers do. A single with a runner on second scores him. A walk in the same spot does not. On average a hit is worth more to the offense than a walk, and WHIP rates them as equal.
The remedy is to read WHIP next to something that measures the quality of contact: home runs per nine innings, opponents' slugging percentage, or opponents' OPS. A pitcher with a 1.05 WHIP and 35 home runs allowed is a different creature from one with a 1.05 and 12. The hitter's version of the same argument, why reaching base and hitting for power are counted separately, is in our guide to on-base percentage.
How much of a WHIP belongs to the pitcher
Split the numerator in two and the halves behave very differently.
- Hits72%
- Walks28%
An approximation for recent MLB seasons, in which pitchers have allowed a little over eight hits and a little over three walks per nine innings. The exact split varies by a few points from year to year.
Show the numbers
| Item | Value |
|---|---|
| Hits | 72% |
| Walks | 28% |
The walks are his
A base on balls involves the pitcher, the batter, the catcher and the plate umpire, and no fielder at all. Nothing the shortstop does can turn ball four into an out. Walk rate is among the most stable things a pitcher does from one season to the next, and it is the part of WHIP he owns outright, give or take the influence of catcher framing on borderline pitches.
A pitcher who walks fewer than two batters per nine innings starts every season with a head start of 0.15 or so on a pitcher who walks three and a half. He has to give back that much in hits before the two are level.
The hits are shared
Hits are a different matter. Once a ball is put in play, whether it becomes a hit depends on where it lands and who is standing there. Research published around the turn of this century by the analyst Voros McCracken showed that major league pitchers differ far less than anyone had assumed in how often balls in play fall for hits, and that a pitcher's rate bounces around from year to year with little regard for how well he is throwing. Across the league roughly three in ten balls in play become hits, and most pitchers finish within a narrow band around that mark. Our explainer on BABIP goes through the evidence.
So a pitcher's hit total is a mix. Part of it is his: strikeouts remove balls from play altogether, and a pitcher who strikes out ten per nine innings simply gives the defense fewer chances to fail. Home runs are his as well, since no fielder can catch one. The remainder runs through his defense, his ballpark and chance. A fast outfield turns doubles into outs. A cavernous park swallows fly balls. A soft liner finds grass one week and a glove the next.
The practical reading: when a pitcher's WHIP drops by 0.15 from one year to the next, look at which half moved. If the walks fell, he has probably changed something. If the hits fell and the strikeouts did not rise, he may only have been luckier, or been traded to a team with better fielders. Statistics built to remove the defense from the picture, such as the one covered in our piece on FIP, exist because of exactly this problem.
Small samples, relievers and the runners WHIP never sees
Sixty innings is a full season for a setup man. In 60 innings, a WHIP of 1.05 is 63 baserunners and a WHIP of 1.20 is 72. Nine men, across six months. Five bloop singles that might as easily have been caught account for more than half the gap.
A reliever's WHIP is therefore a fragile number. It can be excellent one year and ordinary the next with the pitcher throwing the same stuff to the same spots. The 2020 season made the point at scale. With a 60-game schedule, the qualifying threshold fell to 60 innings, and Kenta Maeda of the Minnesota Twins finished at 0.750 in fewer than 67 innings, a figure that now sits third on the all-time single-season list. It was a superb two months. Whether he would have held it over 200 innings is a question the list does not ask.
Small samples are only part of the trouble with applying WHIP to bullpens. The larger issue is what it cannot see.
Inherited runners. A reliever enters with the bases loaded and two out and gives up a double that clears them. Three runs score. His WHIP for the appearance records one hit; if he gets the next batter, it is 3.00 for a third of an inning, which looks bad, but the three runs themselves are charged to the pitcher he replaced. Had he instead walked in a single run and then escaped, his WHIP would be the same. The statistic has no column for runners who were already there.
Outs that were handed to him. A reliever who comes in with two outs needs only one to finish the inning and collects a third of an inning for it. A specialist who retires one batter in each of three appearances has a full inning pitched and a WHIP of 0.00 without ever having to get through a whole frame. The three-batter minimum introduced in 2020 reduced that kind of usage but did not remove it.
Leverage. WHIP weighs a walk in a ten-run game exactly as it weighs a walk with the tying run on third in the ninth. For starters this mostly evens out. For late-inning relievers, whose entire job is the high-pressure handful of batters, it leaves out the thing that defines them. How those jobs are divided up is the subject of our guide to the bullpen.
For a reliever, then, WHIP is a reasonable first glance at whether he keeps the bases clear. It should never be the last one. Strikeout rate, walk rate and the percentage of inherited runners he allows to score tell more.
The neighbors on the stat line and what each one adds
WHIP is a compression. It folds two rates into one and discards the detail. The numbers printed around it restore some of what was thrown away.
| Statistic | How it is figured | What it adds that WHIP lacks |
|---|---|---|
| Hits per nine (H/9) | Hits × 9 ÷ innings | Isolates the half of WHIP that depends partly on defense and luck |
| Walks per nine (BB/9) | Walks × 9 ÷ innings | Isolates the half that is the pitcher's own control |
| Strikeout-to-walk ratio (K/BB) | Strikeouts ÷ walks | Shows whether he is missing bats, which WHIP ignores completely |
| Home runs per nine (HR/9) | Home runs × 9 ÷ innings | Shows how hard he is being hit when he is hit |
| Opponents' batting average | Hits ÷ at-bats against | Measures hits per chance, not per inning |
| Opponents' on-base percentage | Times on base ÷ plate appearances against | Includes hit batters and counts per batter faced |
The first two rows are WHIP taken apart. Add H/9 to BB/9, divide by nine, and the result is WHIP. Looking at them separately answers the question raised two sections ago: is this pitcher's WHIP built on control or on a low hit total?
Strikeout-to-walk ratio covers the gap WHIP leaves widest. Strikeouts do not appear anywhere in the formula. A pitcher who gets every out on a ball in play and one who strikes out a third of the hitters he faces can post the same WHIP, and the second has far more control over his own results.
Opponents' on-base percentage is the closest relative and, strictly, the better-built statistic. It counts hit batters. It divides by plate appearances, so a pitcher is not flattered by double plays and runners caught stealing. It also sits on the same scale hitters are judged by, which makes comparison easy. Its drawback is familiarity: very few fans know offhand whether a .295 on-base percentage against is good. (It is.)
Opponents' batting average has the opposite problem. Everyone understands it, and it tells only the hits half of the story, ignoring walks in the same way the hitter's version does, as our page on batting average explains.
The honest summary is that WHIP survives because it is easy and roughly right, not because it is the best available measure. For a quick sense of whether a pitcher keeps men off base, it does the job in one number.
Reading WHIP on a broadcast graphic or in a box score
The graphic flashes up as a reliever jogs in from the bullpen: 2.45 ERA, 1.02 WHIP, 31 strikeouts. There are about eight seconds to make sense of it.
- How many inningsUnder about 50, treat it as a sketch. Around 100 it starts to describe the pitcher. A full starter's season of 160 or more is solid ground.
- Starter or relieverA reliever's figure rests on a third of the innings and swings far more. Compare a starter with starters.
- What is beside itCheck the ERA for agreement. A wide gap in either direction means home runs, stranded runners or luck are involved.
- Which half is doing the workGlance at the walks. A low WHIP built on few walks is sturdier than one built on a low hit total.
A reading method, not an official standard. The innings thresholds are rules of thumb for how far a WHIP can be trusted.
In a box score there is no WHIP column for the game itself, only the season figure in the fuller online versions. The raw material is there, though: IP, H and BB sit next to one another in every pitching line, and the day's figure can be worked out at a glance. Six innings, five hits, one walk is a WHIP of 1.00. Anything where hits plus walks is less than the innings is a strong outing by this measure.
Early in the season the number is close to meaningless. A starter who has thrown 18 innings in April with a 0.83 WHIP has allowed 15 baserunners; four more singles would make it 1.06. By the All-Star break, with 100 innings behind him, each additional hit moves the figure by a hundredth. That is roughly the point at which it deserves to be believed.
Career WHIP is the most trustworthy version of all, since thousands of innings smooth out defense, ballparks and fortune. It is also the one least often shown.
Two habits are worth keeping. First, convert it to a game: multiply by nine and ask whether that many baserunners sounds like a pitcher who gets through six innings. Second, never read it alone. Paired with ERA, a strikeout count and the home runs allowed, WHIP earns its place. Alone it can mislead in every way described above. A companion measure that judges a starter outing by outing is described in our guide to the quality start.
The record book: Pedro Martinez, George Walker and Addie Joss
For 118 years the lowest single-season WHIP in major league history belonged to Guy Hecker, who posted 0.769 for Louisville of the American Association in 1882. Walter Johnson came closest in the twentieth century with 0.780 for Washington in 1913.
Then came 2000. Pedro Martinez, pitching for Boston in a league where the average team scored more than five runs a game, allowed 128 hits and 32 walks in 217 innings. Opposing hitters batted .167 against him. His WHIP of 0.737 broke a record older than the pitching distance of 60 feet, 6 inches, and it did so in about the most hostile conditions a pitcher has faced. It remains the lowest in American League or National League history.
It is no longer the lowest on the list. In 2024 MLB incorporated the statistics of seven Negro Leagues from 1920 to 1948 into its official record, and with them came George Walker's 1940 season for the Kansas City Monarchs, listed at 0.735. Walker's mark edges Martinez's at the third decimal place. The figure is less settled than the others on the list: MLB's own database, working from a slightly different set of box scores, credits him with 31 hits and 7 walks in 52 innings, which comes to 0.731. He is ahead of Martinez on either count. Negro Leagues schedules were much shorter than the American and National League seasons of the time, so the innings behind the figure are far fewer, and the record keepers set the qualifying minimum accordingly. Both statements are true at once: Walker is first on the list, and Martinez holds the lowest mark over a full 162-game season. The story of those leagues and how their numbers were recovered is told in our history of the Negro Leagues.
The career list
The career record is Addie Joss's 0.968, compiled over 2,327 innings for Cleveland between 1902 and 1910. The list requires 1,000 innings, and Joss is the only retired pitcher on it to finish clearly below one baserunner an inning.
Behind him the margins are almost invisible. Ed Walsh of the Chicago White Sox and Mariano Rivera of the New York Yankees both round to 1.000. Walsh's figure is 0.9996 and Rivera's 1.0003: one fell a fraction short of a baserunner an inning across nearly 3,000 innings, the other a fraction over across nearly 1,300. Carry the division to four places and Walsh is second, Rivera third.
| Pitcher | Career | Career WHIP | Note |
|---|---|---|---|
| Addie Joss | 1902 to 1910 | 0.968 | The record; every inning in the dead-ball era |
| Ed Walsh | 1904 to 1917 | 1.000 | 0.9996 before rounding |
| Mariano Rivera | 1995 to 2013 | 1.000 | 1.0003 before rounding; a reliever throughout his prime |
| Jacob deGrom | 2014 onward | 1.000 through 2026 | 1.0004 before rounding; still active, so the figure is not final |
Active pitchers move on and off this list. Jacob deGrom stood at 0.994 through the 2024 season, which would have placed him second had it held. It did not. Through the 2026 regular season he had allowed 1,312 hits and 391 walks in 1,702 and a third innings, a WHIP of 1.0004, one ten-thousandth behind Rivera. A career figure is not final until the career is.
The usual caution about eras applies. Joss and Walsh worked when the ball was dead and hits were scarce, and Rivera pitched an inning at a time. Martinez's 2000 stands out precisely because none of those advantages applied.
Where WHIP came from: a Rotisserie league in 1979
Most baseball statistics were handed down by league secretaries and newspaper editors of the nineteenth century. WHIP was invented by a writer who wanted to win a game against his friends.
Daniel Okrent, a writer and book editor, drew up the rules for what he called the Rotisserie League in the winter of 1979 to 1980. The name came from La Rotisserie Française, the Manhattan restaurant where the founders met for lunch. Each of them would draft a roster of real National League players and be ranked on those players' actual statistics over the 1980 season. It was the first fantasy baseball league in the form that later spread around the country.
Okrent needed scoring categories, and he settled on eight: batting average, home runs, runs batted in and stolen bases for hitters; wins, saves, earned run average and one more for pitchers. The eighth was his own creation. He wanted a measure of how well a pitcher kept men off base, something that would reward a different skill from ERA, and it had to be computable from the statistics printed in the newspaper. Hits, walks and innings were all there. He added the first two, divided by the third and called the result innings pitched ratio, shortened in the league's paperwork to IPRAT.
The league's founders were writers and editors, and they publicized their own invention. A rulebook published in 1984 carried the game, and its odd eighth category, to offices and dormitories nationwide. By the end of the decade the game had players all over the country, every one of them checking a statistic that no major league club had asked for.
At some point in that spread the clumsy original name was replaced by the acronym. The formula never changed.
The path from there to the scoreboard was indirect. Fantasy players wanted the number, so newspapers and annual guides began to print it. Broadcasters picked it up because their viewers already knew it. By the 2000s it was on the pitching graphic of every national telecast and in the stat tables of MLB's own website, and front offices that had better tools of their own still quoted it because agents, writers and arbitrators did.
Okrent went on to a long career in journalism and has spoken with some ambivalence about being remembered for a fantasy game. The statistic's limits, set out above, are real. But it answered a question nobody had bothered to put a number on before 1979, and it did it with arithmetic a ten-year-old can check. That is why it lasted.
For more on the numbers that surround it, the baseball pages of our United States guides cover the rest of the pitching line.