Defensive Runs Saved Explained: How DRS Rates Fielders

Andrelton Simmons played shortstop for the Atlanta Braves in 2013 and was credited with 41 Defensive Runs Saved, the highest single-season total the statistic has produced since its data begins in 2003. Forty-one runs is about four wins. A shortstop who hits like an average shortstop and fields like that is one of the most valuable players in the game, and almost none of that value shows up in errors or fielding percentage.

Defensive Runs Saved explained in one sentence: DRS estimates how many runs a fielder saved or cost his team compared with an average major leaguer at the same position, so zero is average, a positive number is better and a negative number is worse. Sports Info Solutions, the data company founded by John Dewan, produces it by having video scouts chart the location, type and speed of every ball in play in Major League Baseball, then measuring each fielder against how often an average fielder turns the same kind of ball into an out. Plus 5 is an above-average season, plus 10 very good and plus 15 or more is the level of the best glove at the position. The rest of this page explains the engine, the pieces and the arguments.

What DRS says: runs above an average fielder at his position

Every part of the definition matters. The unit is runs, the same currency a hitter's production is measured in, so a fielder's plus 12 and a hitter's 12 runs above average can be set beside each other. The baseline is an average fielder, not a replacement-level one and not a perfect one. And the comparison is within position: a shortstop is judged against shortstops, a right fielder against right fielders.

That last point is easy to overlook. A first baseman with plus 4 DRS and a shortstop with plus 4 DRS are equally far above their peers, but the shortstop's position is far harder, and a team values that separately. DRS does not add a bonus for difficulty. The positional adjustment that recognizes a shortstop's harder job lives elsewhere, in wins above replacement, where it is added on top. Our explainer on what WAR actually measures covers how that adjustment and the fielding runs combine.

Zero is not a criticism

A fielder at zero did exactly what an average major league fielder would have done with the balls hit near him. That is the standard of a big league regular, not of a casual player, so zero is a perfectly respectable season for most everyday players. A shortstop at zero is a capable major league shortstop. Fans sometimes read a figure near zero as a knock. It is not.

A negative figure is a real cost, though. A center fielder at minus 10 cost his team about 10 runs, or roughly a win, against an average replacement at the position. Whether he keeps his job depends on how much his bat earns back.

What it is not

DRS is not a count of plays. It is not a ranking. It is a runs estimate built from probabilities, and the probabilities come from human charting. Each of those words carries uncertainty. A well-built DRS figure over a full season is a reasonable estimate with a wide margin; over a month it is close to noise.

The DRS scale, from plus 15 down to minus 15

A rough guide for reading a full-season figure travels with the statistic, printed in the FanGraphs library among other places, and it is symmetric around zero. Fans and broadcasters have adopted it largely as written.

Season DRS	How it is read
+15 or more	Gold Glove caliber, the best fielder at the position
+10	Great
+5	Above average
0	Average
−5	Below average
−10	Poor
−15 or worse	Awful, a defender actively costing his team games

• 15runsGold Glove caliber
• 10runsGreat
• 5runsAbove average
• 0runsAverage

The commonly published rule-of-thumb guide for a full season at one position, a convention rather than an official threshold. The same steps apply below zero, so minus 10 is poor and minus 15 is awful.

Why the steps are about five runs

Ten runs is, roughly, one win in most modern run environments. So the rungs mark half-win steps. A fielder at plus 15 is worth about a win and a half over an average fielder at his position from his glove alone. That is a meaningful figure for a front office: a player who hits at league average but fields at plus 15 is an above-average regular before his bat is considered against his position.

Playing time changes the reading

The scale assumes a full season at one position, something like 1,200 to 1,400 innings for a regular. A utility player who gets 300 innings at shortstop and posts plus 5 has had an excellent short stint, and a figure scaled to a full season would be much larger. Stat sites often publish a rate version per 1,200 innings for this reason. Reading the raw number without checking the innings is the most common mistake fans make with DRS.

The best seasons go well beyond the top rung. Simmons's 41 in 2013 is the high-water mark. Seasons in the twenties occur most years somewhere in the majors, usually from a shortstop, a center fielder or a third baseman, the positions where range has the most balls to work with.

Who produces DRS: Sports Info Solutions and its video scouts

DRS is a proprietary statistic. Unlike batting average or OPS, nobody can calculate it from a box score, because it depends on data that is not in a box score: where each ball was hit, how hard and of what kind.

Sports Info Solutions, originally Baseball Info Solutions, was founded by John Dewan, who had earlier been an executive at STATS Inc. and a figure in the early statistical analysis community. The company built its business on charting. Video scouts watch every major league game and record, for each ball in play, where it landed or was fielded on a grid of the field, whether it was a ground ball, line drive, fly ball, pop-up or bunt, and how hard it was hit, in broad categories such as soft, medium and hard. They also record events a box score misses entirely: a diving stop that held a hitter to a single, a throw that was cut off, a ball that bounced off a glove for a hit.

Why human charting

When DRS was built, nothing else could produce that data at scale. Before Statcast, no public system tracked every batted ball's trajectory and every fielder's position. Human scouts watching video were the only practical source. The charting is checked, and SIS has refined its categories over the years, but it remains partly a judgment call. Two scouts may grade a ball's speed differently. That is a real source of error, and SIS has been open about it.

The Fielding Bible

Dewan presented the method publicly in The Fielding Bible in 2006, which introduced the plus/minus system that sits at the heart of DRS. The Fielding Bible, Volume II, published in 2009, built Defensive Runs Saved on top of it, converting plus/minus into runs and adding the components that range alone misses. Later volumes refined it further. The name stuck, and so did the company's annual Fielding Bible Awards.

DRS is published on Baseball-Reference, FanGraphs and SIS's own outlets. Teams buy more detailed SIS data than the public sees.

The engine underneath: the plus/minus system

Plus/minus is the part that does most of the work, and it is simple once the charting is done.

Take every ball in play over a season or several seasons. Group them by location, type and speed: a medium-hit ground ball to a particular zone in the hole between shortstop and third, for example. For each group, work out how often the fielder at a given position turned that kind of ball into an out. That rate becomes the expected value for any future ball in the same group.

Now credit each play against the expectation. Suppose shortstops convert a certain hard-hit ground ball in the hole into an out 30 percent of the time. A shortstop who makes that play gains 0.70 of a play, because he got an out where an average fielder would get one 30 percent of the time. A shortstop who fails to make it loses 0.30, because an average fielder would sometimes have turned it. Routine plays barely move the needle: on a ball converted 98 percent of the time, a made play is worth plus 0.02 and a miss is worth minus 0.98.

1. 1Chart the ballA video scout records where the ball went, whether it was on the ground or in the air and how hard it was hit.
2. 2Find the matching groupThe ball is grouped with every similar ball across a large sample of seasons.
3. 3Set the expectationThe share of those similar balls that fielders at the position turned into outs becomes the expected rate.
4. 4Credit the resultA made play earns one minus the expected rate. A missed play loses the expected rate.
5. 5Add up the seasonThe credits are totaled, giving the fielder's plus/minus in plays above or below average.

A simplified outline of the method Sports Info Solutions describes. The real system uses finer zones, velocity and trajectory categories, and adjusts for the fielder's starting position in its revised form.

Why this beats counting errors

The design rewards range. A shortstop who gets to a ball an average fielder would not reach, and throws the runner out, gets big credit. A shortstop who does not reach it at all is debited for the share of fielders who would have. Under the old error statistic, that shortstop who could not reach the ball was charged with nothing, and his fielding percentage looked perfect. Plus/minus penalizes the ball that got by as well as the ball that was booted.

What it cannot see by itself

Plus/minus counts plays. It does not say what each play was worth, whether a ball that got through became a single or a double, and it does not cover arms, double plays or the other things a fielder does. Turning it into runs, and adding those other components, is what makes DRS a fuller statistic than plus/minus.

From plays to runs: why stopping a double is worth more

A play saved in the outfield and a play saved in the infield are not worth the same. An infielder who stops a ground ball in the hole usually turns a single into an out. An outfielder who runs down a ball in the gap often turns a double or a triple into an out. Those are different swings in run expectancy, and DRS converts each into runs to keep the comparison honest.

The conversion rests on run values from play-by-play data, the same family of numbers that underlies weighted on-base average for hitters. An out instead of a single is worth roughly three-quarters of a run to the defense, on average. An out instead of a double is worth more, about a run, and an out instead of a triple more still. Those values shift slightly with the run environment of each era, so the conversion is calibrated to the period.

Bases as well as outs

For outfielders the system also tracks bases. A corner outfielder who cuts a ball off before it reaches the wall, holding a hitter to a single when an average fielder would have allowed a double, saves a base without recording an out. That saved base has a run value of its own, smaller than the value of an out but not trivial over a season. Plus/minus in its early form was expressed in bases saved for outfielders for this reason.

Why runs and not plays

Runs put fielding on the same footing as hitting. A shortstop who is plus 15 in DRS and plus 10 with the bat, measured in runs above average, is plus 25 overall before the positional adjustment. Nobody can add 15 plays to 10 runs, but 15 runs and 10 runs add cleanly. That is why every serious fielding metric today ends in runs, even when, as with Statcast's system, it starts by counting outs.

The same principle drives how batting is measured; our page on OPS in baseball shows how hitting statistics evolved toward run values in the same period.

The components of DRS beyond range

Range, the plus/minus part, is the largest piece for most fielders, but DRS adds several more. Which ones apply depends on the position.

Component	Positions	What it credits
Range and positioning	All fielders	Plays made or missed against the plus/minus expectation, converted to runs
Outfield arm	Outfielders	Runners held to fewer bases, and runners thrown out, against average
Double plays	Infielders	Double plays turned against the expected rate on chances
Bunts	Corner infielders, pitchers, catchers	Handling bunts compared with an average fielder
Good plays and misplays	All fielders	Defensive events not captured elsewhere, from a diving stop to a missed cutoff
Home run robbery	Outfielders	Balls taken back over the fence
Stolen base runs	Catchers, pitchers	Stolen bases and caught stealing prevented, plus pickoffs
Strike zone runs	Catchers	Extra strikes gained by framing pitches at the edge of the zone

Arms

Outfield arm runs credit outfielders for runners who did not take the extra base. An average right fielder allows a runner to go from first to third on a single to right a certain share of the time. A right fielder whose throws keep runners at second more often than that earns arm credit, and so does one who throws out runners trying to advance. Much of an elite arm's value is invisible: runners simply do not try him. DRS catches that through the hold rate.

Double plays

Infielders get credit for turning double plays on ground balls that an average pair would turn less often, and lose credit for missing them. The second baseman's pivot is the big piece here.

Home-run robberies

Robbing a home run is worth a lot in one play: taking away a run or more and adding an out. SIS tracks them separately because they are rare and the run value is too large to bury in range.

Catchers

The catcher's DRS is mostly made of things that have nothing to do with range. Stolen base runs credit or debit him for the running game. Strike zone runs credit him for framing, the craft of making borderline pitches look like strikes, which SIS added to DRS once pitch-tracking data allowed it. Our explainer on catcher framing covers how that skill is measured and why it is under pressure from automated ball-strike systems.

Good fielding plays and misplays

SIS scouts also log specific events, such as a fielder catching a ball at the wall or a fielder letting a throw get away, and assign them run values when they are not already captured elsewhere. This is where the charting is most subjective, and the component is small.

Positioning and the PART breakdown

The shift broke the original plus/minus. Through the 2010s, teams increasingly moved infielders to where a given hitter tended to hit the ball, putting a third baseman in shallow right field or a shortstop directly behind second base. A fielder standing in the right spot made plays an average fielder in the standard spot would not, and plus/minus, which compared against where fielders typically stood, credited him for a coaching decision.

In 2020, SIS revised DRS to deal with this. The revised system separates a fielder's results into pieces under the label PART: Positioning, Air balls, Range and Throwing.

Positioning is the value of where the fielder was standing when the ball was hit, compared with where an average fielder would have stood for that hitter and situation.

Air balls covers balls in the air, from line drives to pop-ups, where the fielder has to get to the spot in time.

Range covers ground balls the fielder reached, judged from where he actually started.

Throwing covers whether the fielder converted the ball he reached into an out with an accurate, timely throw.

Why the split matters

The point of pulling positioning out was to stop crediting or blaming a fielder for a decision that coaches and analysts usually make. A shortstop who makes a play only because the front office's spray chart put him in the right spot did not show range on that play. Separating positioning lets DRS judge the fielder from where he actually was. The breakdown also makes the number more honest: two shortstops with the same total may have arrived there differently, one through quick first steps and the other through a strong arm that turned more reached balls into outs.

The 2023 shift limits

MLB's 2023 rules require two infielders on each side of second base and all four on the dirt when the pitch is thrown. That pushed infield positioning back toward a narrower range of options, though teams still adjust within those limits. Our explainer on the baseball shift ban covers what the rule actually says and what it left alone. For fielding metrics, the change meant infielders' range numbers became more comparable with the pre-shift era again, though not entirely.

DRS against OAA: charting against tracking

Statcast's Outs Above Average starts from a different kind of data. Where DRS relies on scouts to chart a ball's location, type and speed, OAA uses tracking cameras that measure exactly where each fielder was standing, how far he had to go to reach the ball and how much time he had. MLB introduced OAA for outfielders first and extended it to infielders in the late 2010s.

The question OAA asks is simple: given the distance and time, how likely was an out? The outfield version is built on catch probability; the infield version also considers the distance to the base and the runner's speed. A fielder who makes a play with a 25 percent probability earns 0.75 of an out, just as in plus/minus, and loses 0.25 for missing a likely one.

	Defensive Runs Saved	Outs Above Average
Producer	Sports Info Solutions	MLB's Statcast
Data source	Video scouts charting each ball	Tracking cameras measuring each fielder and ball
Starting position	Accounted for in the revised positioning piece	Measured directly for every play
First unit	Runs	Outs, converted to runs separately
Arms and double plays	Included	Mostly outside the core figure
Catchers	Included, with framing and the running game	Measured with separate Statcast tools
Seasons covered	From 2003	From the mid-2010s, when Statcast tracking began

Why the two disagree

The same center fielder can be plus 10 in one system and close to zero in the other. Most disagreements trace to the inputs. OAA knows exactly where the fielder started and how fast the ball came off the bat. DRS relies on a scout's grading of the ball's speed and the location of the play. A ball graded as medium-hit by a scout might, in the tracking data, have given the fielder less time than the category implies, and the credit differs.

The scope differs too. DRS adds arm, double-play and bunt runs on top of range; OAA in its basic form counts range-and-conversion outs. A right fielder with an exceptional arm and average range will look better in DRS than in OAA for that reason alone.

Neither is the truth. Analysts who need a fielding number usually look at both, plus any others available, and trust a player most when they agree.

DRS against UZR: the same idea, different builders

Ultimate Zone Rating, developed by Mitchel Lichtman in the early 2000s, is DRS's closest relative. It follows the same logic: divide the field into zones, work out how often an average fielder at a position converts balls hit to each zone, and credit or debit each fielder against that, then convert to runs. UZR also adds components for arms, double plays and errors.

For years the two drew on the same company's data. UZR used the batted-ball charting from Baseball Info Solutions, the company now called Sports Info Solutions. The differences were in method: how the zones were drawn, how the expectations were adjusted for park and batter handedness, how the arm and error pieces were built. Those choices were enough to make the two systems diverge regularly on individual players, sometimes by 10 runs or more in a season.

What became of UZR

UZR was FanGraphs' fielding statistic for most of the 2010s and fed into its version of WAR. As Statcast's tracking data matured, FanGraphs moved toward Outs Above Average for recent seasons, and UZR is quoted far less than it was. It remains useful for the seasons before tracking data existed, and its history is part of why analysts learned to distrust a single fielding number: when two careful systems built from the same raw data split on a player by a win in a season, the honest conclusion is that fielding is hard to measure.

Why one season of DRS is not enough

A regular hitter gets around 600 to 700 plate appearances in a season. A regular fielder at most positions gets a few hundred balls hit in his area, and most of those are routine. The plays that separate good fielders from average ones, the balls in the hole, the line drives in the gap, the bloop in short right, come along only a handful of times a week.

That small sample makes single-season fielding numbers jumpy. A fielder can be plus 12 one year and plus 2 the next with no real change in ability, simply because the balls hit near him in one year happened to be more reachable. Studies of year-to-year correlation in fielding metrics have found them less stable than most hitting rates over the same playing time, which is why analysts commonly say a fielding number needs two or three seasons to be trusted, roughly the sample a hitter's batting average needs to settle.

The practical reading

Average a fielder's last three seasons, weighting the most recent somewhat more, and the result is far more reliable than any single season. If a player's three seasons run plus 14, plus 9 and plus 16, he is very likely an excellent fielder. If they run plus 14, minus 3 and plus 6, the honest conclusion is that he is probably above average, by an amount nobody can pin down.

Age

Fielding peaks early. Range depends on speed and first-step quickness, which decline from the late twenties. A shortstop's DRS typically drifts down as he ages, and the familiar career path, from shortstop to third base or second base and then to first base or designated hitter, follows that decline. Our guide to baseball positions by number shows the order of the field and the usual direction of travel along it.

Where the number goes: WAR, the Fielding Bible Awards and Gold Gloves

DRS matters beyond the stat pages because other measures and awards use it.

Wins above replacement. Baseball-Reference built the fielding part of its WAR on DRS for seasons from 2003, the first year of SIS data, and used Sean Smith's Total Zone for earlier seasons. The site has revised its inputs over time, so its methodology notes are the place to check which fielding source a given season uses. A swing of 15 DRS moves a player's WAR by around a win and a half, which is why fielding numbers decide so many MVP arguments about shortstops and center fielders.

The Fielding Bible Awards. Sports Info Solutions runs its own annual awards, started in 2006, with a panel of analysts and writers voting on the best fielder at each position across both leagues combined. The panel is not bound by DRS, but the data shapes the discussion. A single winner per position, rather than one per league, is the main structural difference from the Gold Glove.

The Gold Glove. Rawlings has given its Gold Glove Awards since 1957, voted by managers and coaches. Since 2013 a statistical component, the SABR Defensive Index, has counted for roughly a quarter of the selection, alongside the votes. The index blends several fielding measures, and DRS has been among its inputs. Before then, the voting was criticized for rewarding reputation and offense: a player who won once tended to keep winning, and good hitters won more often than their gloves deserved.

1. 1Season chartedSIS video scouts chart every ball in play and compute each fielder's DRS by position.
2. 2PublishedThe figures appear on Baseball-Reference, FanGraphs and SIS outlets during and after the season.
3. 3Into WARBaseball-Reference's version has used DRS for its fielding runs in seasons from 2003.
4. 4Fielding Bible panelThe SIS panel votes one winner per position across both leagues.
5. 5Gold Glove indexThe SABR Defensive Index, which has drawn on DRS among other measures, makes up about a quarter of the Gold Glove selection.

A simplified sequence. Each destination weighs DRS differently, and the Gold Glove and Fielding Bible voters are free to disagree with the data.

What DRS replaced: errors, fielding percentage, range factor and Total Zone

The old fielding statistics measured what a fielder did with the balls he reached. They could not see the balls he never reached, and that is most of the story.

Metric	Formula or source	What it misses
Errors	The official scorer's judgment on a misplay	Every ball the fielder was too slow to reach
Fielding percentage	Putouts plus assists, divided by putouts plus assists plus errors	Range entirely, and it punishes fielders who reach more balls
Range factor	Putouts plus assists per nine innings, from Bill James	Who hit the balls, where, and how many came to the position
Zone rating	Share of balls in a fielder's zone turned into outs	Balls outside the zone and the difficulty within it
Total Zone	Play-by-play records of where balls were fielded	Exact location and speed, which the play-by-play data does not record

Errors and fielding percentage

An error is charged only when the scorer judges a fielder should have made a play with ordinary effort. A slow shortstop who never gets near a ball in the hole is charged nothing. A rangy one who dives, stops it and throws wild is charged an error. Fielding percentage, built on errors, therefore rewards the fielder who reaches fewer balls. It is still printed beside every player's line, and it still tells you something about sure hands, but it is the clearest example in baseball of a statistic that measures the wrong thing well. Our guide to how to keep score explains how the error call is made in the first place.

Range factor

Bill James's range factor in the 1970s took the first step: count chances per game rather than the rate of clean plays. A shortstop who makes many more plays per nine innings than his peers is covering more ground. Its weakness is that a fielder can only field what is hit near him. A shortstop playing behind a staff of ground-ball pitchers gets far more chances than one behind strikeout pitchers.

Total Zone

Sean Smith's Total Zone, built from Retrosheet's play-by-play files, applied the zone idea to decades of seasons before video charting existed, using the fielder recorded on each play. It is the best available estimate for players like Brooks Robinson and Ozzie Smith, whose careers ended before charting existed, and it is the reason a WAR figure exists for them at all. It cannot know exactly where each ball went or how hard it was hit, so it is coarser than DRS.

Each step in that sequence fixed one blind spot of the last. DRS was the first widely published metric to see where every ball went and how hard it was hit. It has its own blind spots, which is why Statcast's tracked numbers sit beside it now, but it changed the public argument about fielding from reputation to evidence. Two decades later, the argument is about which evidence to trust.

For more of our guides to the numbers on the field, see the baseball section of the site, and for the full player profile these numbers feed into, our page on the five-tool player.