FIP vs ERA: What Is a Good FIP in Baseball?

A good FIP is anything under about 3.80, and 3.20 or lower is excellent, on the scale FanGraphs uses. FIP stands for fielding independent pitching. It takes the results a pitcher controls on their own (strikeouts, walks, hit batters and home runs) and puts them on the ERA scale, so a 3.50 FIP reads like a 3.50 ERA. Because it leaves out what the defense and luck did, it's a better guide than ERA to how that pitcher will do next.

A good FIP, by the numbers

FanGraphs publishes a rule-of-thumb scale for FIP. Here it is next to a 2026 qualified starter who finished near each mark on the FanGraphs 2026 leaderboard.

Read it like ERA: lower is better, and "below average" sits at 4.40. FanGraphs built the scale on the 2016 run environment, and 2026 landed almost on its "average" mark. League-average FIP is set to equal league-average ERA every season, and in 2026 both came in at 4.18.

FIP counts only what the pitcher controls

The formula is short: take 13 times home runs, add 3 times walks plus hit batters, subtract 2 times strikeouts, divide by innings, then add a constant of about 3.10 that moves the result onto the ERA scale. Home runs carry the most weight because they do the most damage. Strikeouts pull the number down, and strikeouts start with missed bats, which is why swinging strike rate is the first thing to check on a pitcher's FIP.

What the formula ignores is every ball in play. In the early 2000s, research by Voros McCracken found that the share of balls in play falling for hits against a pitcher doesn't carry over well from one season to the next, per FanGraphs. Whether a grounder finds a hole has a lot to do with the fielders and plenty to do with luck.

The order of events matters too, and there's no evidence pitchers control it. FanGraphs gives this example: with two outs, two singles and then a home run cost three runs, while a home run followed by two singles costs one. ERA charges those two innings differently, and FIP scores them the same.

FIP vs ERA: the gap shows where ERA is headed

In FanGraphs' words, FIP "does a better job of predicting the future than measuring the present." Most pitchers end up with an ERA close to their FIP over a career. In a single season the two can split wide, and when they do, next year's ERA drifts toward the FIP more often than not.

Here are four of the widest 2025 gaps on FanGraphs, next to what those pitchers did in 2026.

Those are the clean cases. For the fuller picture, take the 10 qualified pitchers whose 2025 ERA ran furthest below their FIP and the 10 whose ERA ran furthest above it. In 2026, 13 of those 20 saw their ERA move the way FIP pointed and seven didn't, per Baseball Savant. FIP makes the better guess, and it still misses.

Looking ahead to 2027, the widest 2026 gaps among qualified pitchers belonged to Robbie Ray (3.59 ERA, 4.69 FIP) and Jacob deGrom (4.20 ERA, 3.11 FIP), per FanGraphs. FIP says Ray's ERA flattered him and deGrom's undersold him.

Some pitchers beat their FIP for real

FanGraphs names FIP's blind spots itself. Pitchers who shut down the running game, or who give up fly balls in place of line drives and grounders, are more likely to beat their FIP than the average arm. If a pitcher's ERA has sat below their FIP for three seasons running, some of that is probably skill.

Home runs are the other weak spot, because the share of fly balls that leave the park jumps around from year to year. xFIP handles that by swapping in the home runs a pitcher would have allowed at the league rate per fly ball, which was 11.6% in 2026. Aaron Nola's 2026 season is a good example: 18.3% of his fly balls went for home runs, his FIP was 4.87 and his xFIP was 3.86. If you think those homers were bad luck, xFIP is your number.

FIP also isn't adjusted for ballparks, so a pitcher who works in a hitter's park carries a higher one. FanGraphs' FIP- adjusts for park and league, and the park factors post covers what the park does to a single game.

A new role makes the old numbers useless

Everything above assumes a pitcher keeps the same job. A reliever pitches an inning or two and almost never sees a hitter twice: in 2026, 94% of plate appearances against relievers were the hitter's first look at that pitcher in the game, per Baseball-Reference. A starter has to get through the order two or three times, and hitters improve each trip. Against starters in 2026 they posted a .701 OPS the first time through, .735 the second and .750 the third.

Griffin Jax is the 2026 case. Pitching in relief in 2025, he struck out 35.4% of the hitters he faced, per MLB.com, and his four-seam fastball averaged 97.0 mph on Baseball Savant. Tampa Bay moved him into the rotation in 2026, and his four-seamer averaged 95.9 that season. In his 23 starts he struck out 25.5% and got 13.5 outs a start. Project his starts off that relief rate and you'd have expected about 10 more strikeouts per 100 batters than he delivered.

So when a pitcher changes roles, start their sample over. Watch how many pitches they're allowed in the new job, and give the new strikeout rate time to build (the small samples post has the numbers). It shows up fastest on an outs pick. Per FanGraphs, starters averaged about 15 outs a start in 2026, and Jax averaged 13.5. If the card puts YES at 55% on "Outs 15+" for a converted reliever, NO is the 45% side: +5.5 if they're pulled before their 15th out, −4.5 if they get there.

How FIP changes an earned-runs pick in GAGE

Say the GAGE menu has a starter at "Earned runs 3+" and the card gives YES a 40% chance. Their ERA is 2.70 and their FIP is 3.64, the gap Peralta carried through 2025. YES scores +6.0 if they allow three and costs 4.0 if they don't.

On ERA alone, three earned runs looks like a stretch. The 3.64 FIP says luck and defense have been keeping the runs down, and that's your case for YES. The scoring page has the full rules, and GAGE is on the App Store when you want to make the pick.