The Most Important Red Sox Might Be the Middle Relievers
On the surface, the Red Sox and Indians series is somewhat evenly matched. The Indians won 94 games, the Red Sox won 93, and that one-game difference gives Cleveland home-field advantage for the ALDS. But if you look at our Playoff Odds page, our forecasts give the Red Sox a 60% chance of winning this series, because the Red Sox are quietly a monster in waiting.
They had the best offense in the AL this year, and by a laughable margin.
With an offense that dominant, the pitching staff just needs to be okay, and that’s mostly what the Red Sox staff was this year. They weren’t great, but they were solidly above average, and that’s why the Red Sox outscored their opponents by 184 runs, the second-best total in baseball. The Sox pitching staff was strong up front but weak at the back end, as their collection of No. 4 and No. 5 starters all struggled, but that’s the kind of weakness that is downplayed in the postseason. And with more emphasis on bullpen usage in the playoffs, a couple of those struggling starters could turn out to be incredibly valuable for the Red Sox this October.
Joe Kelly, who has had a rough couple of years as a starter, was finally converted into a reliever this year. See if you can spot where it happened in his monthly splits.
| Month | TBF | H | BB | SO | BA | OBP | SLG | wOBA |
| Mar/Apr | 50 | 14 | 10 | 11 | 0.359 | 0.500 | 0.632 | 0.472 |
| May | 52 | 10 | 6 | 15 | 0.222 | 0.327 | 0.432 | 0.327 |
| Jun | 17 | 7 | 3 | 1 | 0.500 | 0.588 | 0.833 | 0.548 |
| Jul | 16 | 4 | 2 | 1 | 0.286 | 0.375 | 0.500 | 0.372 |
| Sept/Oct | 53 | 9 | 3 | 20 | 0.180 | 0.226 | 0.240 | 0.210 |
After getting beat like he stole something for the first four months of the year, the Red Sox sent Kelly to the minors in August, let him begin to work as a reliever, and then brought him back in September when rosters expanded. And after getting back to Boston, he was one of the best relievers in baseball during the final month of the season.
Unlike most starter-to-reliever conversions, the big jump in success didn’t come from a spike in velocity; Kelly already threw really hard, and he didn’t throw appreciably harder out of the bullpen than he did in the rotation. Mostly, he just started hitting better spots. Look at Kelly’s heat map of pitch selections for his career, and then look at September.
If you want to know why a guy with good stuff hasn’t been a good pitcher yet, that first graph should tell you all you need to know: he lived in the center of the strike zone, where hitters could square up his fastball and drive it to the moon. But look at what he did in September. He got his pitches up and started living at the top of the zone, and if there’s one sure way to get strikeouts, it’s to throw hard fastballs up. As a starter, Kelly was always a pitch-to-contact ground-ball guy, but in September, he was a guy who missed bats by living up in the zone.
This version of Joe Kelly could be an important part of the Red Sox bullpen, giving the team a power right-handed arm to bridge the gap between the starters and the late-inning relievers. Especially with David Price and Eduardo Rodriguez penciled in to start Games Two and Three, Kelly could be useful following the left-handers in Boston’s rotation. Don’t be surprised if the Indians end up seeing a good amount of Joe Kelly in this series — and if he pitches the way he did the final month of the season, they will be sick of him by the end of the ALDS.
Of course, the Red Sox already have a number of good right-handed relievers, so Kelly just gives them extra depth there. What they don’t have a lot of is good left-handed relievers. But after Drew Pomeranz proved sufficiently healthy to stick in the team’s rotation, the team used him out of the bullpen right before the season ended. Here’s a velocity graph from Brooks Baseball, which makes it clear where the relief appearance happened.
In that one outing, Pomeranz topped out at 96, after having mostly maxed out at 93-94 as a starter. He struck out two of the five batters he faced, and looked like a legitimate weapon for John Farrell in October. And we’ve already seen Pomeranz work as a pretty nasty lefty reliever before.
Last year, in Oakland, the A’s converted Pomeranz to the bullpen, and watched him turn into one of the best left-handed bullpen guys in baseball. As a reliever in 2015, Pomeranz allowed opponents to hit just .199/.279/.308, and was especially nasty against LHBs, holding them to a .139/.236/.258 mark. In his career, he’s faced 111 left-handed batters while pitching in relief, and those batters have managed a .212 wOBA against him.
With a big knockout curveball and a fastball maybe now sitting around 95, Pomeranz is a rough matchup for any left-handed hitter. Along with Kelly, he could give the team two starter-conversion relievers who could be highly effective in the middle innings, and allow the Red Sox to not have to lean too heavily on their starting pitchers. And with the Red Sox offense likely to be a wrecking ball in the postseason, the addition of two extra quality bullpen arms could make Boston a very tough October opponent.
The Indians are a also good team, of course, and in the postseason, anything can happen. But with Kelly and Pomeranz looking like potential weapons out of the bullpen, in addition to all the other good players they already had, the Red Sox look like the kind of team built for a strong postseason run.
Dave is the Managing Editor of FanGraphs.




Another big problem Kelly had was that he got very predicable which you can’t get away with when you’re going through the lineup 2-3 times but you can get away with it when you only face a couple of batters. That and he seemed way too fond of his 2-seamer which had almost too much movement and it would leak back into the middle much too often.
Sox fan here: biggest concern is Kimbrel who looked just awful in September and has spoken about his mechanics being out of whack/coming along in recent bullpen sessions. If he comes out and walks the bases loaded like he did twice during the stretch run before getting pulled in each outing the sox could be in trouble. This would be especially so if Kelly is the playoffs 8th inning guy. He was the one called upon to clean up those two Kimbrel debacles mentioned.
So FG actually has the Dodgers with the best odds of going all the way, followed by the Red Sox. That’s an indication of the respect given the Giants, who are given a better chance of eliminating the Cubs than the Nats beating the Dodgers or Cleveland over Boston. It looks as though either winner of Giants-Cubs may be favorites to win the WS.
I’m pretty sure it’s an indication that the FG odds really like Kershaw and Hill and assign them most of the Dodgers’ innings.
FG Playoff odds has LAD winning the LCS 56% of the time they reach it. Cubs are at 51%. SF Giants are at 45%. Nats are at 43%.
You’re right, my bad. The odds will be revised when the two teams in the LCS are actually known, but it looks like Dodgers will be slight faves over Cubs, and certainly over the Giants. Still surprised that the Cubs aren’t bigger favorites over the Giants.
Big assumption here Dave. I think it’s clear their strength is their offense and Gold Glover Mookie Bets. If the kid doesn’t win the MVP this year as the clear engine of the World Series Champs, something will be severely wrong with the voting process and I’d call the entire institution into question.
If you’re going to switch it up from posting whiny comments about the Red Sox to pretending to be an obnoxious Red Sox fan, you should probably change your old handle.
This guy’s got the chops for oppo research. Has the Hillary campaign been in contact yet?
Thumbed down for a political reference on Fangraphs. Don’t bring politics here…please.
“Clear engine of the World Series Champs”? Betts (sic) wasn’t even on the team in 2013. Oh, you mean this year… well, don’t start polishing that hardware just yet, sport. There’s a lot of baseball still to play. And even if the Sox do win it all this year, it’s completely irrelevant to the MVP voting process, which is based on regular season performance and has already happened (even if the results are not released until after the postseason).
I realize I’m just feeding the troll here, but even reasonable adults seem to forget that the MVP (and the other awards) have nothing to do with postseason performance.
He’s no Kevin Kiermaier
You know what I’m talking about. We’ve got to get out there and watch the voters in certain areas.
Quasi-related: I feel like the title hurts my brain. I get that it’s grammatically correct, but ‘Sox’ (either the Red or White variety) seem like awkward descriptors for ballplayers.
‘Mookie Betts is a Red Sox’ sounds ridiculous. So does ‘Mookie Betts is a Red Sock.’ Nothing sounds good other than ‘Mookie Betts is on the Red Sox’ or ‘is a member of the Red Sox.’
I doubt this was the article’s intended purpose, but I can’t help but be reminded how dumb the ‘x’ is in ‘Sox.’
The Sox struggle against good pitching like anybody else, or against ordinary pitching having a good day. They lead baseball in garbage runs-and I’m a Sox fan. It surprises me that Fangraphs does not have a run scoring index filtering out non-competitive games. These games will be decided by pitching, and quite possibly the incompetence of John Farrell, whose incompetence specializes in pitching.
Every team that ever leads the league in hitting also leads in “garbage runs.” No team can time their big-run games to line up with when their pitching fails.
Is that true, though? I remember reading Jayson Stark’s article the other day that had some data on this:
They hit good pitching. Or to put this in the words of a rival GM: “Boston’s offense is the singular best strength of all the teams.” Now those are words that get my attention this time of year. When we try to break down which clubs profile best to survive the four-week October marathon, isn’t that what we look for — a team so dominant that even the best teams in the sport can’t stop it from doing what it does best? And that’s the Red Sox. Against pitchers who average a strikeout per inning or better, the Sox lead the major leagues in average, on-base percentage, slugging and OPS. Against pitchers defined as “power pitchers,” Boston’s OPS (.798) is 65 points higher than the next-best team. But the Red Sox are also first in OPS against “finesse” pitchers (.848). And they went into the weekend as the only team in the past five years with an OPS of .800 or better against both left-handed and right-handed pitching. So good luck finding a matchup that looks like this team’s kryptonite.
This doesn’t mention – and I don’t know where to find – data vs “good pitching”, as you put it. So you might be right after all. But as far as I can tell, the offense doesn’t struggle very often.
Because of their pitching, I wouldn’t pick them for the World Series or even necessarily the ALCS. But their hitting is inarguably excellent.
I’m sorry for my continued skepticism, but “the sox were a dominant offense this year” still does not equate to “the sox are a dominant offense going forward”.
Fangraphs has once again forgot about dre i mean babip. The Sox had an insane first 2 months with a babip near .350 which has skewed their offense ever since. If we just look at a simple 2nd half split we see a normal .302 babip resulting in a solid 105wrc+ and just over 5 runs per game. This normal-babip team is who they are – the insane babip team of the first 2 months was an illusion. Still a good offensive team, but nowhere close to dominant.
It seems obvious when you actually just look at it – no team starting all of sandy leon, brock holt, and a 21yr old fresh out of AA, and 2 guys in bogaerts and bradley who haven’t hit at all since babip turned on them sometime in June, is that dominant an offense.
They are a good offensive team, no more, no less.
107 Off Runs vs 60 for the Cubs (the 2nd best team) sure sounds like a lot to put solely on two months of the season.
And of course past performance doesn’t mean future performance will be the same…that applies to everything (including the idea that they will be a .302 BABIP team because that’s what they did in the 2nd half).
Even if .302 BABIP, 5 r/g is who they are…what says they can’t exceed that in the playoffs? If they can do it for 2 months, they could also very well do it for 3 weeks or so of the playoffs.