+ "How many runs did this player save (or cost) their team with their glove?" +
++ Defense was baseball's "dark matter" for over a centuryβeveryone knew it mattered, but no one could measure it properly. + Errors and fielding percentage captured only a tiny fraction of defensive value, ignoring the most important skill: range. +
+The Traditional Era: Errors and Fielding Percentage
+ ++ For most of baseball history, defense was evaluated almost exclusively through two statistics: +
+ +The Fatal Flaw: Range Doesn't Count
++ Errors and fielding percentage share a catastrophic blindspot: they only count plays the fielder actually touches. + If a ball is hit 10 feet to a shortstop's left and they can't get to it, nothing gets recorded. No error. No penalty. +
++ This creates a perverse incentive. The easiest way to have a high fielding percentage is to have limited range. + If you only attempt easy plays, you'll make fewer errors. But you'll also let dozens of hittable balls become singles. +
++ Meanwhile, an aggressive defender with exceptional range might commit more errors (because they attempt difficult plays), + but they'll get to 50+ balls per season that an average defender wouldn't reach. Those extra outs are worth far more than + avoiding a few errors. +
+Example: The Range Paradox
+The Verdict
++ Shortstop A has the better fielding percentage (.985 vs .970), but Shortstop B made 120 more outs (440 vs 320). + Those extra 120 outs prevented roughly 60 runs over the season. Traditional stats couldn't see this. +
+The Revolution: Measuring What Matters
+ ++ Starting in the early 2000s, researchers developed metrics that finally measured defensive value properly: + range, arm strength, + and positioning. These metrics answer the core question: "How many runs did this defender save?" +
+ +DRS: Defensive Runs Saved
+ ++ Defensive Runs Saved (DRS), developed by Sports Info Solutions (SIS), is the most comprehensive defensive metric. + It combines multiple components to estimate total runs saved above/below average. +
+ +DRS Components
+RANGE RUNS
++ Measures plays made compared to expected based on ball location, speed, and trajectory. Divides field into zones. + Calculates: "What percentage of fielders historically make this play?" Rewards getting to difficult balls. +
+ERROR RUNS
++ Accounts for errors, but weighted by difficulty. Bobbling a routine grounder costs more runs than dropping a + diving catch. Not just error countβconsiders context of each play. +
+ARM RUNS (OUTFIELDERS)
++ Measures throwing effectivenessβpreventing extra bases, getting runners out. Considers arm strength and accuracy. + Strong, accurate arms prevent 5-10 runs per season by holding runners. +
+DOUBLE PLAY RUNS (INFIELDERS)
++ Measures ability to turn double plays compared to average. Factors in speed of runners, ball location, game situation. + Elite middle infielders turn 10-15 more DPs per season than average. +
++ DRS = Range Runs + Error Runs + Arm Runs + DP Runs +
++ All components measured in runs above/below average. Zero is league average. +
+| DRS | +Rating | +
|---|---|
| +20 or better | +Historic, generational defender | +
| +15 to +20 | +Gold Glove caliber, elite | +
| +10 to +15 | +Excellent defender | +
| +5 to +10 | +Above average | +
| -5 to +5 | +Average range | +
| -10 to -5 | +Below average | +
| -15 to -10 | +Poor defender, significant liability | +
| -20 or worse | +Historically bad, DH candidate | +
Calculating Range Runs
+ ++ The field is divided into precise zones. For every batted ball, video scouts record: +
+ +-
+
- Location: Exact zone where ball was hit +
- Type: Groundball, line drive, fly ball, popup +
- Velocity: Soft, medium, hard +
- Result: Did fielder make the play? +
EXAMPLE CALCULATION
++ A ball is hit to shallow right field, medium velocity, line drive trajectory. +
+Historical conversion rate: 40% of right fielders make this play
+Expected outs: 0.40
+Actual result: Fielder makes the catch = 1.00 out
+Plays above average: 1.00 - 0.40 = 0.60 plays
+Run value (out vs. single): ~0.80 runs
+Runs saved on this play = 0.60 Γ 0.80 = +0.48 runs
++ Over a full season of 1,400+ innings, hundreds of these plays accumulate into the final DRS total. +
+UZR: Ultimate Zone Rating
+ ++ Ultimate Zone Rating (UZR), published by FanGraphs, uses a similar methodology to DRS but with + different data sources and zone definitions. Both measure runs saved above/below average. +
+ +DRS vs. UZR: What's the Difference?
+-
+
- + β’ + Different data providers: DRS uses Sports Info Solutions data; UZR uses Baseball Info Solutions + +
- + β’ + Zone definitions vary slightly: Different ways of dividing the field + +
- + β’ + High correlation (~0.85-0.90): They usually agree on who's good/bad + +
- + β’ + Both are valid: Use DRS if available (more widely cited), UZR is fine too + +
UZR/150: Normalized Scale
++ UZR is often presented as UZR/150, which normalizes to 150 games (roughly a full season). + This allows fair comparison between players with different playing time. +
++ Scale is identical to DRS: 0 = average, +15 = elite, -15 = poor +
+OAA: The Statcast Revolution
+ ++ Outs Above Average (OAA) represents the next evolution in defensive metrics. Instead of relying on + human video scouts to classify plays, OAA uses measured Statcast data: exit velocity, launch angle, + direction, and sprint speed. +
+ ++ For every batted ball, Statcast knows the exact exit velocity (e.g., 95.3 mph), + launch angle (e.g., 28Β°), and direction. + Using millions of tracked balls, MLB calculates: "What's the historical catch probability for this exact batted ball profile?" +
+How OAA Works
+CATCH PROBABILITY
++ Every batted ball gets assigned a catch probability from 0% to 100% based on historical data: +
+β’ 90%+ catch probability: Routine play, almost everyone makes it
+β’ 50-90% catch probability: Moderate difficulty
+β’ 25-50% catch probability: Difficult play, requires good range
+β’ Under 25%: Very difficult, spectacular plays
+β’ Under 10%: 4-5 star catches, highlight reel
+EXAMPLE
++ Ball hit at 98 mph, 32Β° launch angle, to center field +
+Statcast catch probability: 35% (difficult play)
+Expected outs: 0.35
+Fielder makes the catch: 1.00 out
+Outs above average: 1.00 - 0.35 = +0.65 outs
++ Accumulate these over a season to get total OAA +
+Why OAA is Revolutionary
+-
+
- + β’ + Objective measurement: No human classification bias, uses actual tracked ball flight + +
- + β’ + Accounts for difficulty: Makes spectacular catch = credited properly + +
- + β’ + Publicly available: Updated daily on Baseball Savant + +
- + β’ + Correlates well with DRS/UZR: All three usually agree (~0.75-0.85 correlation) + +
Statcast: Supporting Metrics
+ ++ Statcast provides additional metrics that help explain why a player has good or bad defensive value: +
+ +Sprint Speed
++ Measures maximum speed in feet/second. Faster players cover more ground, leading to better range. +
+β’ Elite: 30+ ft/sec (Byron Buxton, Trea Turner)
+β’ Above avg: 28-30 ft/sec
+β’ Average: 27 ft/sec
+β’ Below avg: Under 27 ft/sec
+Reaction Time
++ Time from bat-ball contact until first movement. Quicker reactions = better jumps on balls. +
+β’ Elite: Under 0.70 seconds
+β’ Good: 0.70-0.75 seconds
+β’ Average: 0.75-0.80 seconds
+β’ Slow: Above 0.80 seconds
+The Connection
++ A center fielder with 30 ft/sec sprint speed and 0.65 sec reaction time will cover + significantly more ground than an average defender (27 ft/sec, 0.75 sec reaction). This translates directly to + higher DRS/UZR/OAA valuesβthey simply get to more balls. +
+Positional Context Matters
+ ++ Not all defensive positions are equally difficult. A +10 DRS center fielder is more valuable than a +10 DRS + first baseman because center field is harder to play and affects more balls. +
+ +Defensive Spectrum
++ Positions ranked by difficulty (hardest to easiest): +
+How MLB Teams Use Defensive Metrics
+ +Player Acquisition
++ Undervalued defenders with +15 DRS can be acquired for less than their true value contributes. + Defense saves runs just like offense creates them. +
+Positional Decisions
++ Move declining defenders from premium positions (SS, CF) to easier spots (1B, LF, DH) to minimize + defensive liability while keeping bat in lineup. +
+Defensive Shifts
++ Use spray chart data and defender range profiles to position fielders optimally for each batter. + Can save 10-20 runs per season. +
+Contract Valuation
++ A +20 DRS defender saves ~2 WAR on defense alone. That's worth $16M+ on the free agent market. + Defense can command premium salary. +
+Limitations and Caveats
+ +What to Watch For
+-
+
- + β’ + Sample size: Defense metrics stabilize slower than offense. Need ~3 years of data for reliable assessment. + +
- + β’ + Year-to-year variance: DRS/UZR can swing Β±10 runs between seasons for the same player. Average multiple years. + +
- + β’ + Positioning matters: Team defensive shifts affect individual metrics. A player in optimal position looks better. + +
- + β’ + Doesn't capture everything: Pitcher familiarity, situational awareness, leadership aren't measured. + +