Mariano Rivera retired with a 2.21 ERA. Greg Maddux finished at 3.16. Rivera's ERA is almost a full run better.
Does that make Rivera the better pitcher? No.
Rivera threw 1,283 career innings, all in relief. Maddux threw 5,008 innings as a starter. They're doing different jobs. Comparing their ERAs directly is like comparing a sprinter's 100-meter time to a marathoner's average pace per 100 meters.
Relief pitchers have lower ERAs than starters. Always have, probably always will. Let's talk about why.
The Numbers: How Big Is the Gap?
Research from the Society for American Baseball Research found that relief pitchers have an ERA advantage of 0.15 to 0.25 runs compared to starters of similar ability.
More recent data (2012-2021) shows relief pitchers averaged a 3.97 ERA while starters averaged around 4.20-4.50 ERA. That's about a 0.25-0.50 run difference.
| Category | Starting Pitchers | Relief Pitchers |
|---|---|---|
| Average MLB ERA (2012-2021) | 4.20-4.50 | 3.97 |
| Good ERA | Under 3.50 | Under 3.00 |
| Elite ERA | Under 2.50 | Under 2.00 |
| Innings per appearance | 5-7 innings | 0.5-2 innings |
| Times through lineup | 2-3 times | Usually once |
Rule of thumb: Subtract 0.50 runs from typical starter ERA benchmarks when evaluating relievers. A 3.50 ERA is good for a starter, a 3.00 ERA is good for a reliever.
Why Relievers Have Lower ERAs
1. They Enter with Outs Already Recorded
This is the big one.
Here's an example: Three pitchers each pitch one inning. Each gives up two walks, then gets an out.
- Pitcher 1 (starter): Starts the inning, gives up 2 walks, gets 1 out, leaves with bases loaded and 1 out. Gets pulled. Next pitcher gives up a hit — all 2 runs score. Pitcher 1 is charged with 2 runs.
- Pitcher 2 (middle reliever): Enters with 1 out already recorded, gives up 2 walks, gets 1 out (2 outs total), leaves with bases loaded. Next pitcher gives up a hit — all 2 runs score. Pitcher 2 is charged with 2 runs.
- Pitcher 3 (closer): Enters with 2 outs already recorded, gives up 2 walks, gets the final out. Inning over, runners stranded. Pitcher 3 is charged with 0 runs.
Same performance (2 walks, 1 out). Different results because of when they entered.
Relievers often come in with 1 or 2 outs already recorded. Their mistakes are less likely to score because there's less time in the inning for things to go wrong.
2. They Only Face Batters Once
Starters face the lineup 2-3 times per game. Batters adjust. The more times through the order, the worse a starter's ERA gets:
- 1st time through: .250 batting average against
- 2nd time through: .265 batting average against
- 3rd time through: .280+ batting average against
Relievers usually only face batters once. Hitters don't get a second look to adjust.
3. They Can Go All-Out Every Pitch
A starter throwing 100 pitches needs to pace himself. A closer throwing 15 pitches can max out every throw.
As of 2019, reliever average fastball velocity was 94.1 mph. Starters averaged 93.3 mph. That 0.8 mph gap has been shrinking, but relievers still throw harder on average because they pitch fewer innings.
4. They Don't Have to Save Energy for Late Innings
A starter in the 2nd inning knows he might pitch 6 more innings. He can't blow his best stuff early.
A reliever comes in knowing exactly how long he needs to go. One inning? He can empty the tank.
Calculate Your ERA
Whether you're a starter or reliever, track your performance.
Use ERA Calculator →What's a Good ERA for Each Role?
Starting Pitchers
Elite: Under 2.50
Excellent: 2.50-3.00
Above Average: 3.00-3.50
Average: 3.50-4.50
Below Average: 4.50-5.00
Poor: Over 5.00
Closers
Elite: Under 1.50
Excellent: 1.50-2.50
Above Average: 2.50-3.00
Average: 3.00-3.50
Below Average: 3.50-4.00
Poor: Over 4.00
Middle Relievers / Setup Men
Elite: Under 2.00
Excellent: 2.00-3.00
Above Average: 3.00-3.50
Average: 3.50-4.00
Below Average: 4.00-4.50
Poor: Over 4.50
Best Relief Pitcher ERAs All-Time
Here are the greatest relievers ever and their career ERAs:
| Pitcher | Career ERA | Saves | Years Active |
|---|---|---|---|
| Mariano Rivera | 2.21 | 652 | 1995-2013 |
| Trevor Hoffman | 2.87 | 601 | 1993-2010 |
| Dennis Eckersley | 3.50 (overall) 2.85 (as reliever) |
390 | 1975-1998 |
| Rollie Fingers | 2.90 | 341 | 1968-1985 |
| Billy Wagner | 2.31 | 422 | 1995-2010 |
Mariano Rivera's 2.21 ERA is the lowest in baseball history for any pitcher with 1,000+ innings pitched. His ERA+ of 205 means he was 105% better than league average — the best mark ever.
The Inherited Runners Problem
ERA gets messy for relievers because of inherited runners.
Situation:
- Starter leaves with bases loaded, 2 outs in the 6th inning
- Middle reliever enters, gives up a grand slam
- All 4 runs score
Who gets charged?
- Starter: 3 runs (his baserunners scored)
- Reliever: 1 run (only the batter he faced)
The reliever gave up a grand slam but only 1 run counts against his ERA. The starter's ERA takes the bigger hit even though he wasn't on the mound when the runs scored.
This is why "inherited runners scored" is tracked as a separate stat for relievers. A reliever who constantly allows inherited runners to score is bad, even if his ERA looks okay.
Why ERA Isn't Great for Relievers
ERA was designed for starting pitchers who throw 6-9 innings per game. For relievers who pitch 0.5-2 innings at a time, ERA has problems:
1. Small Sample Size Issues
A reliever who pitches 60 innings all season has a tiny sample. Allowing 3 runs in one bad outing can spike his ERA by 0.45.
For a starter with 200 innings, those same 3 runs only raise his ERA by 0.13.
2. One Bad Inning Ruins Everything
A closer can have 30 scoreless outings, then give up 4 runs in the 31st. His ERA jumps from 0.00 to 4.00 instantly.
3. Inherited Runners Distort Responsibility
If a reliever enters with the bases loaded and no outs, his job is incredibly hard. If he gets out of it allowing 1 run, that's actually great work. But his ERA takes the hit.
Better Stats for Relievers
Because ERA has limitations for relievers, scouts and analysts use other stats:
Scoreless Outing Percentage: How often did the reliever finish his appearance without allowing a run?
K/BB Ratio: Strikeouts divided by walks. Great relievers have ratios above 3.0. Mariano Rivera: 4.10 career K/BB.
K% and BB%: Percentage of batters faced who struck out or walked. Removes the "per 9 innings" problem for pitchers who don't throw 9 innings.
Inherited Runners Scored %: What percentage of runners already on base when the reliever entered actually scored? Good relievers keep this under 30%.
FIP (Fielding Independent Pitching): Measures strikeouts, walks, and home runs allowed. Better predictor than ERA for future performance.
The Changing Gap
Interestingly, the ERA gap between starters and relievers has been shrinking.
In 2012, relievers had ERAs half a run better than starters. By 2019, that gap shrank to about 0.25 runs.
Why?
- Starting pitchers throw harder now: The velocity gap is closing. Starters in 2019 averaged 93.3 mph, relievers 94.1 mph — only 0.8 mph difference.
- Starters don't go deep anymore: Most starters get pulled after 5-6 innings now, before the third time through the order crushes them.
- Too many relievers: Teams carry 13-14 pitchers now. Not all of them are good. More mediocre relievers = higher reliever ERAs overall.
The role might be getting overused, which dilutes quality.
The Bottom Line
Relief pitchers have ERAs 0.20-0.50 runs lower than starters because:
- They often enter with outs already recorded
- They only face batters once (no lineup turnover)
- They can throw max effort every pitch
- They pitch fewer innings per appearance
This doesn't make them "better" pitchers. It makes them different.
A 3.50 ERA is good for a starter. A 3.00 ERA is good for a closer. Adjust your expectations by about half a run when comparing the two.
And if you really want to evaluate a reliever, look at strikeout rate, walk rate, inherited runners scored, and scoreless outing percentage. ERA tells part of the story, but not all of it.