Paying Relievers Is Dumb and the Braves Should Stop Doing It

The one where I rant about bullpen volatility and propose a better way forward

One of the common refrains from Atlanta Braves fans this winter is about spending money on a “proven” commodity in the bullpen.

Look, I get it. Last year’s bullpen was comprised of some of the usual suspects - Raisel Iglesias, Pierce Johnson, Dylan Lee - plus a whole platoon of minor league free agents signed over the winter, most of whom didn’t make it out of spring training and none of whom made it to the end of the season. The relievers finished with a cumulative 4.19 ERA, the 19th-best mark in baseball, and converted a below-average 62% of their save opportunities (34 for 55).

However, with the news that the Braves have brought back closer Raisel Iglesias and signed September waiver claim Joel Payamps to a $2.25M deal for 2026, Atlanta could be close to done with their bullpen for next season. This is clearly predicated on both the health of Joe Jiménez, who had a second knee procedure earlier this offseason, and the availability of Grant Holmes, who is rehabbing a UCL tear and could be pushed out of Atlanta’s rotation. Even if one of those two is unavailable, adding a single reliever to the returning names might be as far as the Braves go in their hunt for a solid backend to the pitching staff.

And in that spirit, I’m here to ask for the Braves to stop paying large contracts to veteran relievers based on their past performance. Let’s talk about why.

Reliever volatility is real

And it’s because of how they’re used.

Raisel Iglesias is a great example of this. In his four-plus-season Atlanta Braves career, he’s faced 856 batters, striking out roughly 28% of them and allowing a home run rate of 2.5%.

Chris Sale, by contrast, faced 702 batters in 2024 alone, striking out 32.1% of them and allowing homers about 1.3% of the time. In just one season, Sale faced 82% of the number of batters that Iglesias has faced in his entire Braves career.

And funnily, home run rate is one of the slowest pitching statistics to stabilize, taking roughly 1320 batters faced (BF) to give a number that truly correlates to skill.¹ While certain power metrics can get towards a predictive point earlier, like homers per fly ball (400 BF) or ISO (630 AB), it’s still not a very fast process.

Even some of the fastest-stabilizing statistics, strikeout rate (70 batters faced) and walk rate (170 batters faced), take a surprisingly long time to become stable enough to draw a meaningful conclusion from. Using Iglesias again as an example, his 70th batter faced came on May 12th, his 17th game of the season, while his 170th batter was on July 26th.

Using research from Russell Carleton of FanGraphs, here’s the common stabilization point for several popular pitching statistics using either Batters Faced, At-Bats, or Balls In Play.

• 70 BF: Strikeout rate

• 170 BF: Walk rate

• 640 BF: HBP rate

• 670 BF: Single rate

• 1450 BF: XBH rate

• 1320 BF: HR rate

• 630 BF: AVG

• 540 BF: OBP

• 550 AB: SLG

• 630 AB: ISO

• 70 BIP: GB rate

• 70 BIP: FB rate

• 650 BIP: LD rate

• 400 FB: HR per FB

• 2000 BIP: BABIP

Let’s be clear, some relievers are just good

Back to Iglesias again.

If he’s not faced enough batters in his Braves career for every single statistic to normalize yet, how do we know he’s good? Think about him as compared to his peers.

Since Iglesias debuted for the Braves in August of 2022, his 2.35 ERA is the lowest of any National League reliever with as many appearances in the same time period. Despite the rough start to his season, he finished with a 1.25 ERA over his last 45 games and converted each of his final 18 save opportunities, showing he had put the troubles of early 2025 behind him.

And to be clear, Iglesias’s first half was rough. Prior to losing his hold on the closing role in mid-June, he had a 6.75 ERA and had blown four saves, taking five losses.

The flipside here: that same time period saw him face a total of 108 batters, two less than Chris Sale faced in his first five starts.

Can we really take anything away from Iggy’s seven homers allowed or his 18 runs allowed? It’s all small-sample-size theater at that point. Five of those home runs came on sliders, a pitch that had given up a total of three homers in his Braves career prior to 2025. Of those eighteen runs, one-sixth of them came in a single game, when he allowed three runs to St. Louis in a 7-6 Braves win on April 21st.²

Iglesias would give up earned runs in 13 of those 25 first-half games, but Atlanta would go on to win seven of those 13 games anyway.

And then Iglesias locked in down the stretch, allowing a second-half ERA of 1.76 with a single homer allowed.³ In 31 games, he allowed six total runs, and four of those were in one game versus the Yankees right out of the break. From his first game after that New York series through the end of the season, Iglesias allowed a grand total of two earned runs in 28.2 innings, a 0.63 ERA.

It was a stretch of 103 batters faced.

I’m saying that conventional statistics like ERA, WHIP, etc, are even less useful when looking at relievers than when talking about starters. While individually effective pitchers should roughly remain effective pitchers, one significantly bad outing or a run of bad luck can skew these numbers and the small samples make the entire thing suspect.

So, a new way of thinking about relievers

Let’s talk about inputs.

It feels like an ideal way to set up your bullpen is to take the closest thing you can have to a known quantity, a veteran who has a history of outperforming their peers (in this case, Raisel Iglesias), and surround them with as many filthy arms as you can find.

Let’s look at Stuff+, an arsenal grade keyed off of an ‘average’ pitch being graded a 100. The FanGraphs version, sorting for all pitchers who threw at least 20 innings for the Braves last year, looks like this:

Aaron Bummer, Tyler Kinley = 111
Spencer Schwellenbach = 110
Daysbel Hernández, Chris Sale = 107
Dylan Lee = 104
Hurston Waldrep = 103
Dylan Dodd, Pierce Johnson = 100

When you remove the two “one-trick ponies” in Kinley (slider of 115) and Johnson (curveball of 107), you’re left with the team’s two best starters, two lefties that throw four pitches in Bummer and Dodd, a three-pitch lefty in Lee, and a flamethrowing righty in Hernández.

Not a bad start to a bullpen, right?

Let’s expand out a bit, covering the last few seasons, and poof - there’s Joe Jiménez at a 106 overall with three total pitches, two of which are close to or over one standard deviation above 100 in his fastball (109) and slider (106). Raisel Iglesias also comes in at a 101, carried by his previously filthy slider (128).

But when you look at the projected bullpen depth chart for Atlanta and cross-reference with their previous performance, Joel Payamps (102 Stuff+, four pitches) pops up, with a “not throwing his cutter as much” version of Joey Wentz not far behind (96 Stuff+, including a 107 on the cutter).

Rather than shelling out large guaranteed dollars towards other impressive names that are likely to be subject to the same randomness and fluky sample size shenanigans that doomed Iglesias in the first half , why not look at raw stuff and the ability to locate it (also quantified by FanGraphs as Location+)?

Hunter Harvey is likely available for cheap, owing to his injury-riddled season with the Kansas City Royals (10.2 total innings). Across the last four seasons, he put up a 106 Stuff+ with his four-pitch arsenal, including a 123 on his slider, a 113 on his splitter, and an above-average fastball at 103.

On the trade market, Los Angeles Angels righty Ryan Zeferjahn has a top 20 Stuff+ mark of 120, thanks to three phenomenal pitches: a slider (130), a cutter (111), and a four-seam fastball (116), all supported by a changeup that he knows not to throw more than a few times an outing as a tendency breaker.

While it’s true that several of Atlanta’s best relievers in the last few seasons from a Stuff+ perspective all departed in free agency, like A.J. Minter (117 overall, with three pitches all at 109 or better) and Kenley Jansen (131 overall, with three pitches all 113 or better), several of Atlanta’s best weapons of the last four years are currently under team control. And most of them are cheap, too, with only Iglesias ($16M) and the combo of Bummer and Jiménez ($9.5M) making more than a few million dollars.

Lean into that. Find the bargains by searching for the stuff, not paying for the previous variance-influenced statistics. Some organizations build these guys out of spare parts - thinking specifically about the Seattle Mariners here - but either way, you can search the available player pool for this exact skill.

Let’s Moneyball this thing.

I re-read this book every February and then re-watch the movie, because why not?

I understand the risk in this approach - we just watched a season where a substandard Braves pen blew several leads and cost the team wins on the scoreboard. And that was with the game’s fourth-most expensive bullpen at $26.63M. The only pricier armbarns were the Dodgers ($32.4M), Mets ($32.89M), and the Reds ($36.89M, largely because they were using Nick Martinez and his $21.05M Qualifiyng Offer salary in relief.)

But most of those bullpens weren’t very good, either. LA’s bullpen ERA was 21st in baseball and they mostly stuck to using starters in relief during the postseason. New York’s bullpen was roughly average at 15th in MLB and the Reds came in just above them, in 14th place at a 3.89, but with the 6th-fewest innings of any bullpen in MLB.

If spending precious free agency dollars doesn’t guarantee success in the bullpen, don’t spend it there. Go find undervalued guys with good stuff and spend the money on everyday position players, like a shortstop and another rotational bat for the outfield.

1. An R-squared value of .49, for you statistics nerds

2. Two of the three runs came via a Willson Contreras homer, and yes, it was on a slider

3. Yes, on a slider