Fun With RE-RA9

Right off the bat, I have to tell you that I don’t love the name I gave the statistic I created last week. RE-RA9 doesn’t exactly roll off the tongue. The concept – adjusting run-scoring statistics to account for inherited runners – is easy to get your head around, and I think it’s clearly interesting. But while I had fun writing that article, I wasn’t quite happy with where I left off, either on the name front or on the analysis front.

The name thing probably can’t be fixed. I’m not a great namer of things, historically, and I don’t think that’s going to change today. But while I can’t do anything about that, I quickly expanded my coverage from 2026 to, well, as much of baseball as I could. If this statistic is interesting, it’s interesting as much for its application throughout history as for who’s good and bad at it this year. So with the help of the FanGraphs play-by-play database, which stretches back to 1974, I built RE-RA9 for the vast majority of the era where there were enough relief appearances for this statistic to even make sense. Forget Grant Anderson and Chase Silseth, the two poster boys from my first article. Let’s get some famous guys and seasons in here.

For example, here are the 10 pitchers who have done the most to prevent inherited runners from scoring (RE-RA9 lower than actual RA9), minimum 1,000 innings pitched:

Now we’re talking! Jesse Orosco had a long and productive career highlighted by a dominant mid-80s stretch with the Mets, and didn’t retire until 2003. In his later years, he was more of a journeyman middle reliever, but throughout his career, he was downright elite at not allowing the runners he inherited to score. Trevor Hoffman, one of the greatest closers of all time, was nearly as good. I was a bit surprised to see Hoffman at the top, actually; I thought he would have entered in clean innings so often that he’d struggle to stand out by this metric. Mariano Rivera, for example, had a career 2.38 RA9 and 2.31 RE-RA9. Amazing run prevention, yes, but not run prevention that was overlooked by traditional ERA.

Of note, Rollie Fingers is on the list even though he debuted before 1974; the numbers on here are from 1974 on, so he may even be underrated if he was particularly great at this skill early in his career. Also of note, firemen from the pre-closer era, when bullpen roles were less hierarchical, dominate the top of this list. That makes great sense to me – back in the day, elite relievers entered in strange situations more frequently, and it’s no surprise that these great pitchers allowed fewer runs than a run-expectancy table would imply when entering with runners on. Another way of putting this: The impact of a fireman reliever is greater than can be seen from ERA and WAR.

The flip side of the list is interesting as well:

There are no relievers who did as poorly by this metric as Orosco and Hoffman did well. I think that’s a selection issue; if you’re the kind of reliever who allows that many inherited runners to score, you’re probably not going to stick around long enough to pitch 1,000 innings. The relievers here were pretty good, or at least had some stretches of being good. I recognize most of the names without being a huge baseball history buff. I also love seeing the ageless Jesse Chavez on there; he’s finally retired, but in my head, he’s always a few days away from rejoining the Braves on a one-year deal.

There are starters on this list, too. That wasn’t the case in the other direction, but I think there’s a clean reason for that: Since starters don’t exert any control over what happens with their runners after they leave the game, this is mostly a noise statistic for them. Instead of showing us skill, it shows us which starters received the most good fortune after departing the game. That number will naturally be smaller in magnitude than the relievers who stand out when it comes to preventing inherited runners from scoring. Indeed, there are plenty of starters who are better by RE-RA9 than by RA9; they just didn’t make the top 10 because those top relievers are really good at this skill.

There’s nothing special about McCatty, Bauer, and Kelly that made them record better RA9 marks than you’d expect based on the runners they left on base. They just happened to get some good relief work after departing, which flattered their runs-allowed numbers. (As I mentioned in my introduction piece last week, I skipped ERA for score-parsing reasons, but I think that RA9 is still capturing what we want here.)

If you’re the kind of person who wants to see this entire career list, I have two things to say to you. First, haha, what a nerd! Second, you’re our kind of nerd — it’s great to have you here. Here’s the career list, with no innings minimum set. Feel free to slice up, restack, and otherwise peruse to your heart’s content.

Those career records are great, but by their very nature, the effect sizes are small. That’s what happens when you look at a rate statistic over an entire career’s worth of playing time. Pitchers have some good years, some bad years, and the total line reflects some of that. So in addition to looking at which pitchers this statistic evaluates most differently overall, I also looked at single-season numbers. For example, Denny Neagle had a bummer of a 1994 season, with a 5.12 ERA. But some of that came down to the fact that the relievers who came in after him were absolutely flammable. Basically everyone he left on base scored. He departed with a total of 16 runners left on, and mostly not in particularly dangerous situations. Run expectancy would forecast roughly five of those to score. Instead, 14 did. That’s a 0.58 gap between RA9 and RE-RA9, one of the largest in history.

Here’s another fun one for you. Pedro Martínez’s 2001 season wasn’t his best – thanks to injury, he only made 18 starts. But he was still quite dominant, with a 2.39 ERA and peripherals that were even better. His relievers let him down – he left only seven runners on base all season, and six of those scored after he left. Any statistic that makes Pedro look even better is welcome in my book.

An amusing counterpoint: Do you remember Andrew Heaney’s resurgent 2023 season with the Rangers, where he made 28 starts in the regular season and then three in the playoffs as the team won the World Series? It was a charmed year in more ways than one. Most notably from our perspective, Heaney left a whopping 30 runners on base when he departed. That’s quite profligate of him. But Rangers relievers had his back. Only one (!!) of those runners scored. That’s how you end up with a 4.52 RA9 and a 5.12 RE-RA9, the second-largest single-season gap in history (starters, minimum 100 IP).

Among very good seasons, Rick Rhoden’s 1983 campaign stands out to me. It was one of the best years of his sterling career. It was also aided by a wonderful bullpen. He put up a 3.09 ERA and a 3.50 RA9 over 35 starts and 244 innings. His relievers did yeoman’s work to bail him out, though. Runners on third with less than two outs, tough jams – you name it, they escaped it. If you simply took his run expectancy when he left the game, his RA9 would be 3.93, but the Pirates after him were much better than that.

If you switch over to relievers, you can get some truly phenomenal numbers. Just last year, two different notable relievers posted gaps of 1.4 or higher between their RE-RA9 and their actual RA9. First, you’ve got Seranthony Domínguez. On the surface, his 4.31 RA9 isn’t all that inspiring. Under the hood, though, he was outrageous. Dominguez came in with 34 runners on base, a run expectancy near 11 runs. He allowed only four of them to score. On the other side, he departed with 17 runners on base. Nine of those scored. That’s some trick – twice as many runs scored in half as many opportunities with him out of the game as with him entering.

Another surprising name: Devin Williams. Betrayed by the relievers who came after him, Williams left 12 runners on base, and seven of them scored. He also came into the game in extra innings frequently, but didn’t allow that runner to score very often. His rate statistics were all quite poor – 4.79 ERA, 5.37 RA9 – but his 3.92 RE-RA9 wasn’t nearly as bad as you’d expect. In other words, he did all the things that don’t go on the scoreboard very well, and the guys who pitched after him fell apart.

Let’s dive a bit further into history for some more exciting names. The greatest relief season of all time might be 2016 Zack Britton – the year that Buck Showalter failed to use him in an 11-inning Wild Card elimination game. Britton posted a 0.54 ERA and a 0.94 RA9, the best marks in history for each statistic. But he was even better than that! He faced 18 inherited runners, with a run expectancy of five runs, and allowed only two to score. He left only two runners on base all year, and those runners carried a mere 0.31 run expectancy – on first base, late in the inning, basically. Both scored! Account for that, and his RE-RA9 is a ludicrous 0.33. No one’s ever been better.

What about the flip side, relievers who look meaningfully worse after considering all the runs they allow to score that aren’t credited to them? There’s bad news if you’re looking for current relievers with ugly numbers here – there aren’t any. There’s some selection bias going on here. Teams might not be looking at this statistic exactly, but if they have a reliever that’s performing poorly on a per-batter basis, that reliever probably won’t get much rope, shiny ERA or no. Exactly one reliever in the last 10 years has thrown 60 innings with a gap of 1.3 runs or higher. (The 25 worst seasons since 1974 eclipse that mark.) That’d be 2017 Heath Hembree, who managed a 3.63 ERA in 51 innings for the Red Sox, but posted an RE-RA9 of 5.95. He allowed six more inherited runners to score than run expectancy would suggest, and the relievers who replaced him did an excellent job stranding the runners he left them.

Don’t let modern teams’ boring focus on “efficiency” and “correct thinking” get you down, though: There are still plenty of amusing seasons throughout history to point at and chuckle. One of my favorites is Nate Jones’ 2012 campaign. By all accounts, he was pretty freaking good. He had a 2.39 ERA and also a 2.39 RA9 – elite numbers. The White Sox saw those numbers and used him in higher and higher leverage spots as the season went on; despite his starting the season as a rookie mop-up man, he was one of the team’s two most-trusted relievers by year’s end.

There was just one problem – it was all an illusion. Jones’ numbers were flattered by the spectacular relievers who entered after him. He left 24 runners on base when he couldn’t finish innings that year – only one scored. He inherited 54 runners, 25 scored. His RE-RA9 was 3.99, a full 1.6 runs higher than the headline number. Oof — that one is gonna sting.

Perhaps even more shocking is the season with the highest gap between RE-RA9 and RA9. That belongs to Mike DeJean in 2004, and boy did he have a rough one. DeJean pitched 61 innings and even featured in a deadline trade – the contending Mets acquired him for the stretch run, and he played pretty well in limited time there. But his somewhat shiny ERA (a gaudy 1.69 with the Mets, 4.57 on the season) obscured some problems. And by problems, I mean that he was an absolute sieve when it came to allowing runners to reach base. He inherited 31 runners, and left 32 to the relievers who followed him. Of the ones he inherited, 19 scored; only seven of the ones he passed on did. That 12-run gap was a transfer payment from starters’ ERAs (he was flammable, but they took the crooked numbers) to his own (he deserved to have plenty of runs scored after he left, but none did). DeJean’s RE-RA9 was a ghastly 6.91, which is honestly hard to wrap my head around.

As was the case with the career numbers, here’s a leaderboard that you can play around with. There are a lot of names in here, of course. Making your own copy and then filtering is probably the way to go. But I enjoyed my own meandering walk through the data enough that I’m sure others also will. It’s a great way to both remember some guys and learn something new about your own perception of those guys. Dennis Eckersley in 1990? Absolute RE-RA9 beast. Dennis Eckersley in 1996? Got by with a little help from his friends. That’s just another thing that caught my eye, but you get the idea. If you’re looking to more equitably allocate runs allowed, this is a good way to do it – and it’s fun, too.