What Is ERA+?

Comparing two pitchers from different seasons or ballparks using raw ERA gives you a misleading answer. A 3.50 ERA in 1968 is excellent; the same number in 2023 is merely average. Park factors add another layer of distortion. A pitcher at Coors Field faces a park that inflates ERA by roughly 0.50-1.00 runs compared to a neutral park. ERA+ solves this by normalising a pitcher's raw ERA to a league-average baseline of 100, adjusting for both the league average of their season and the park factor of their home ballpark.

ERA+ belongs to the same family of adjusted stats as FIP and SIERA, but it answers a different question. While FIP estimates what a pitcher's ERA would be if they had average defence behind them, ERA+ focuses on the context of era and venue. A score of 100 means the pitcher is exactly league-average for their season and ballpark. A score of 130 means they are 30 percent better than average. That lets you compare a 2023 starter at Coors Field against a 1968 starter at Dodger Stadium on a single scale.

Baseball-Reference defines ERA+ as follows: ERA+ = (League Average ERA / Pitcher's ERA) × (100 / Park Factor). The league average ERA comes from the season the pitcher played. For the 2024 MLB season that figure was 4.38, according to Baseball-Reference's annual table. The pitcher's own ERA is their earned runs per nine innings. The park factor is a multiplier that captures how much a given ballpark inflates or deflates runs allowed relative to a neutral venue; a park that inflates ERA gets a factor greater than 1.0, and that factor divides the score down.

What You Need From Baseball-Reference

To calculate ERA+ yourself, you need three numbers from Baseball-Reference: the league-average ERA for the pitcher's season, the pitcher's raw ERA, and the park factor for their home ballpark that season. Baseball-Reference publishes the league-average ERA table going back to 1905. The 1905 average was 2.87, the 1968 average was 3.49, the 2023 average was 4.26. Park factors are updated annually by FanGraphs, which also publishes the ERA- counterpart.

Where Partial Innings Matter

Get the pitcher's ERA right first. The formula is (Earned Runs / Innings Pitched) × 9. Innings pitched are recorded in thirds. 5.2 IP means 5 and 2/3, which is 5.667, not 5.2. Treating 5.2 as 5.2 instead of 5.667 is the single most common failure mode in ERA calculation. It shifts the result by roughly 0.10-0.30 runs, enough to move a pitcher from average to slightly above average on the ERA+ scale.

Take a 2023 pitcher who allowed 42 earned runs over 90 innings pitched. Their raw ERA is (42 / 90) × 9 = 4.20. The 2023 MLB league average was 4.26. Their home park has an estimated park factor of 1.08 (a mild inflation park, roughly in line with Coors Field's effect). Plug those numbers into the formula:

ERA+ = (4.26 / 4.20) × (100 / 1.08) = 1.014 × 92.59 = 93.9.

This pitcher is slightly below the 100 baseline, about 6 percent worse than the league average after park adjustment. Their raw ERA of 4.20 looked a hair better than the league average of 4.26, but the park factor cancels that out because their home venue suppressed run scoring a bit. A raw ERA of 4.20 without park adjustment would have produced an ERA+ of 101.4, a misleadingly good number.

Now compare a 1968 pitcher with the same raw ERA of 4.20. The 1968 league average was 3.49, and their home park factor is 0.95 (a mild suppression park). Their ERA+ = (3.49 / 4.20) × (100 / 0.95) = 0.831 × 105.26 = 87.5. That same raw 4.20 ERA that was nearly average in 2023 is 12.5 percent below average in 1968, a massive difference that raw ERA alone cannot show.

FanGraphs publishes both ERA+ and ERA-. The two are inverses of each other. ERA+ sets 100 as average and higher is better. ERA- also sets 100 as average but lower is better. A 90 on ERA- means the pitcher is 10 percent worse than average. The formulas mirror each other: ERA- uses the same inputs but flips the division. FanGraphs' library also includes FIP, xFIP, and SIERA, each adjusts for a different confound. FIP removes defence from the equation entirely by using only strikeouts, walks, and home runs. xFIP replaces the pitcher's actual home-run rate with the league-average home-run rate for a given count of fly balls, making it more predictive for small samples. SIERA goes further by accounting for batted-ball types, ground balls, fly balls, line drives, and is considered the most predictive of future ERA for sample sizes over 50 innings, per FanGraphs.

If you are comparing pitchers across different seasons and ballparks, use ERA+ (or ERA-). If you are trying to project future performance from a small sample, choose SIERA. The FanGraphs glossary notes that for relievers, ERA is not reliable below 30 innings, and for starters, below 50 innings. At those sample sizes, adjusted stats like xFIP or SIERA become more useful than ERA+.

Raw ERA dropped steadily from the Dead Ball Era through the 1910s. The 1910 league average was 3.09, and it fell further to 2.87 by 1905. It climbed through the 1920s and 1930s, peaked at 4.95 in 1935, then settled in the 3.50-4.00 range through the 1940s to 1970s. Since then it has risen again. The 2024 average was 4.38. A pitcher with a 3.50 ERA in 1968 (league average 3.49) is essentially average. That same 3.50 ERA in 2023 (league average 4.26) is about 18 percent better than average. Without ERA+, that distinction is invisible.

Park factors compound the effect. A 2023 pitcher at Coors Field already has a raw ERA inflated by roughly 0.50-1.00 runs compared to a neutral-park peer. ERA+ divides by that park factor, giving the Coors Field pitcher a boost that the neutral-park pitcher does not get.50 ERA at a neutral park in 1968 produces an ERA+ around 100. The Coors Field pitcher is actually performing slightly better relative to their context.

When ERA+ Does Not Apply

ERA+ is not useful for small samples. A 15-inning reliever's ERA+ is meaningless because the raw ERA itself is unreliable. Use FIP or SIERA for those cases. It also does not adjust for defence quality, which is what FIP handles. If you need to compare pitchers across levels, high school to MLB, ERA+ cannot bridge the rulebook gap, because NFHS and MLB have different earned-run judgment rules. The ERA formula is identical, but what counts as an earned run differs, and those differences produce gaps of 0.50-2.00 runs. Stick to league-average ERA tables by season and level for that comparison.

Use ERA+ if you compare pitchers across different seasons or ballparks, project value for fantasy baseball, explain performance across eras, or evaluate a pitcher who moved from a suppression park to an inflation park. It also suits scorekeepers verifying their software's ERA output, because the same formula applies to any level once you have the correct league average and park factor.

Skip ERA+ if you need a single 'good ERA' chart for all levels. The league-average table from Baseball-Reference exists for exactly that. Skip it if you want to calculate WHIP, K/9, or FIP. Those are separate stats with their own calculators. Skip it if you are evaluating a reliever over 15 innings. At that sample size the raw ERA itself is not reliable, and no adjustment will fix that. The most common failure with ERA+ is applying it to a sample too small to be meaningful, which gives a precise-looking number that means nothing.