A pitcher finishes the season with a 2.80 ERA and a 4.20 FIP. Good season or lucky season?
Another pitcher: 4.50 ERA, 3.00 FIP. Bad season or unlucky?
The gap between ERA and FIP tells a story. ERA measures what happened. FIP measures what the pitcher controlled. When they diverge by a run or more, something interesting is going on.
What Each Stat Measures
ERA
Earned Run Average — runs allowed per 9 innings (excluding unearned runs)
Includes: Everything that happens on the field — hits, walks, homers, defense, luck, sequencing
Best for: Measuring past performance
FIP
Fielding Independent Pitching — what ERA would be with league-average defense and luck
Includes: Only strikeouts, walks, hit batters, and home runs
Best for: Predicting future performance
According to MLB's definition, FIP "focuses solely on the events a pitcher has the most control over" while removing everything that happens when a ball is put in play.
Why FIP Exists
In the late 1990s, analyst Voros McCracken discovered something weird: pitchers had very little control over whether balls in play became hits.
A grounder 10 feet to the right of the shortstop becomes an out if Francisco Lindor is playing short. Same exact ball becomes a hit if a slow-footed defender is there instead. The pitcher threw the same pitch. The result changed because of who was playing defense.
McCracken called this Defense Independent Pitching Theory (DIPS). The idea: strip out everything the pitcher doesn't control and see what's left.
What pitchers DO control:
- Strikeouts (no ball in play)
- Walks (pitcher's command)
- Hit batters (pitcher's wildness)
- Home runs (hard to defend, pitcher's mistake)
FIP only counts those four things. Everything else — whether a line drive finds a gap or gets caught, whether a grounder gets through or gets fielded — FIP ignores.
The Research: FIP Predicts Future ERA Better Than ERA Does
Multiple studies have tested this. The methodology: take a pitcher's stats from one year and see which stat better predicts their ERA the following year.
Results from a 2018 study looking at pitchers who threw 162+ innings in back-to-back seasons:
- 2014 ERA predicting 2015 ERA: Correlation of 0.382
- 2014 FIP predicting 2015 ERA: Correlation of 0.462
FIP wins. It's not close.
A 2020 Cardinals-focused study found that the year-to-year correlation for "beating your FIP" (having a lower ERA than FIP) was only 1.9%. Translation: if a pitcher has a much better ERA than FIP this year, don't expect that to continue next year.
What this means: If you're trying to predict how a pitcher will perform next season, their FIP tells you more than their ERA does. Past ERA includes luck, defense, and sequencing that probably won't repeat. FIP strips that out.
Real Example: James Paxton 2016
James Paxton pitched for the Mariners in 2016. His season:
- ERA: 3.79
- FIP: 2.81
- Gap: 0.98 runs
ERA said he was mediocre. FIP said he was one of the best pitchers in the rotation. What happened?
Paxton got unlucky. Balls in play found holes more often than normal. His defense made some mistakes at bad times. Runners scored when they were on base more than league average.
But his strikeouts, walks, and home runs allowed (what he actually controlled) were excellent. FIP saw that. ERA didn't.
Prediction: Paxton should improve the next year if his peripherals stay similar.
What actually happened: 2017 ERA dropped to 2.98. FIP was right.
When ERA and FIP Disagree
High ERA, Low FIP = Unlucky Pitcher
Probably got hit hard on balls in play, had a bad defense behind them, or gave up hits with runners on base at the worst times. Expect improvement.
What to look for:
- High strikeout rate but lots of runs scored
- Balls in play became hits more than normal
- Defense made errors or didn't make plays
Prediction: ERA should drop closer to FIP next season
Low ERA, High FIP = Lucky Pitcher
Got bailed out by great defense, had balls in play become outs at an above-average rate, or stranded a ton of runners. Expect regression.
What to look for:
- Low strikeout rate but low ERA
- Lots of weak contact (but FIP doesn't measure contact quality)
- Great defense turning balls into outs
- Runners left on base at high rates
Prediction: ERA should rise closer to FIP next season
Calculate Your ERA
Track your earned runs and see how your ERA compares to expectations.
Use ERA Calculator →FIP's Blind Spots
FIP is great, but it's not perfect. Here's what it misses:
1. Contact Quality
A soft grounder to second and a line drive rocket to second both count the same in FIP — neither is a strikeout, walk, or homer. But one is way more dangerous.
Pitchers like Greg Maddux or Kyle Hendricks induced weak contact for years. Their FIP was higher than their ERA because FIP doesn't measure contact quality. They weren't getting lucky — they were skilled at forcing weak swings.
2. Sequencing
FIP doesn't care when things happen. A solo homer and a 3-run homer both count as one home run in the FIP formula, even though one scores three times as many runs.
Pitchers who excel at stranding runners or pitching out of jams can consistently beat their FIP. This might be skill, not luck.
3. Defense Quality
FIP assumes league-average defense. If you pitch for a team with Gold Glove defenders at every position, your ERA will beat your FIP every year — and that's fine. Your defense is actually that good.
4. Knuckleballers and Soft-Tossers
R.A. Dickey (knuckleballer) had a career 3.84 ERA and 4.16 FIP. His knuckleball induced weak contact that became outs more often than FIP expected. FIP thought he was lucky. He just threw a weird pitch that broke the model.
When to Use Each Stat
Use ERA when:
- Evaluating past performance ("How good was this pitcher last year?")
- Comparing who actually helped their team win more games
- Determining awards like Cy Young based on results
- Talking about what happened on the field
Use FIP when:
- Predicting future performance ("How good will this pitcher be next year?")
- Evaluating pitchers with bad defense behind them
- Identifying buy-low or sell-high candidates (fantasy, trades)
- Isolating what the pitcher actually controlled
Use both when:
- Getting the complete picture
- Understanding why a pitcher over- or underperformed
- Deciding if regression (good or bad) is coming
The Formula Breakdown
FIP looks scary but it's just weighted averages:
FIP = ((13 × HR) + (3 × (BB + HBP)) − (2 × K)) ÷ IP + constant
Why those numbers?
- 13 for home runs: A homer is really bad. Guaranteed run, nobody can stop it.
- 3 for walks and HBP: Free baserunner, pitcher's fault.
- -2 for strikeouts: Best possible outcome for a pitcher.
- Constant (usually around 3.10): Makes FIP match the ERA scale so you can compare them directly.
That's it. No balls in play, no hits allowed, no sacrifice flies. Just the four things pitchers control most.
Common Misconceptions
"FIP is a projection"
Wrong. FIP measures what already happened (strikeouts, walks, homers). It just ignores balls in play. It's predictive because what pitchers control tends to repeat. But it's not projecting the future — it's measuring a different version of the past.
"A big ERA-FIP gap means the pitcher will regress to their FIP"
Not exactly. If a pitcher has a 2.50 ERA and 3.50 FIP over a full season, expect some regression — but not all the way to 3.50. Maybe they're actually good at inducing weak contact. Maybe their defense is legitimately elite.
FIP is a yellow flag, not a crystal ball.
"Career FIP tells you if a pitcher was lucky their whole career"
No. Over 10-15 years, luck and defense even out. If a pitcher has a career ERA well below their FIP, they probably had a legitimate skill FIP doesn't capture (weak contact, strand rate) or played their whole career behind great defenses.
FIP works best for single seasons or short stretches where luck matters more.
Which Stat Is "Better"?
This is the wrong question. They measure different things.
ERA is better at: Telling you what happened. It's the runs that actually scored. That's what wins games.
FIP is better at: Telling you what the pitcher controlled and predicting what comes next.
You wouldn't say a thermometer is "better" than a barometer. They measure different weather conditions. Same with ERA and FIP.
For most purposes: Use ERA to evaluate the past. Use FIP to predict the future. Use both to understand what's really happening.
Quick Reference
ERA-FIP gap under 0.50: Pitcher performed close to expected. Both stats agree.
ERA-FIP gap 0.50-1.00: Moderate difference. Some luck involved, might see some regression.
ERA-FIP gap over 1.00: Big difference. Either very lucky, very unlucky, or the pitcher has a skill FIP doesn't measure. Investigate deeper.
ERA much lower than FIP: Expect ERA to rise (regression).
ERA much higher than FIP: Expect ERA to drop (positive regression).
The Bottom Line
FIP isn't trying to replace ERA. It's asking a different question.
ERA asks: "How many runs did this pitcher allow?"
FIP asks: "How many runs should this pitcher have allowed based on what they controlled?"
For evaluating overall performance, use ERA. For projecting forward, use FIP. For making smart decisions about pitcher value, use both.
Just don't treat a 0.80 gap between them as random noise. That gap is data. Sometimes it's luck evening out. Sometimes it's skill FIP doesn't measure. Your job is figuring out which.