Almost 30 percent of ESPN's baseball experts picked Cleveland to win the World Series this year, and for an easy-to-explain reason: The Indians came within one run of winning it last year without the injured Danny Salazar and Carlos Carrasco in their postseason rotation. This year, the thinking went, those two are healthy -- the Indians' playoff rotation will be nasty.
Carrasco is doing his part to make the predictions look smart. In four starts entering Friday night, he had allowed only five runs. Last Saturday he went eight scoreless innings against the White Sox. By the advanced pitching metrics at Baseball Prospectus, he has been the third-most valuable pitcher in baseball this year.
But the day after Carrasco's near-shutout, Salazar allowed four hits and two walks in the first inning against Chicago. He took the loss, his ERA swelling to 4.37.
What might surprise you, then, is that Salazar appears right alongside Carrasco near the top of those advanced leaderboards. At BP, Salazar's deserved run average -- an ERA-like metric that takes into account everything from the quality of his opponents to the defense behind him to the weather -- is a minuscule 1.09, the eighth-lowest in baseball this young season.
DRA isn't alone. At Baseball-Reference, Salazar's 2.02 FIP would be a career best by more than a run -- and ranks eighth among all starters this year. He's 10th on FanGraphs' wins above replacement leaderboard. He's leading the majors in strikeout rate, almost a full strikeout per nine innings better than the all-time. And 38 percent of swings against him have come up empty, the highest whiff rate of any starter in baseball.
Put it all together, and we are watching the long-awaited breakout by one of the game's most powerful arms, the midcareer blossoming of a true ace. Or else we're watching the continuing frustration of an often-unhittable pitcher who perpetually seems one adjustment away from actually fulfilling his promise.
Salazar packs an extraordinary amount of gunpowder into his right arm. He's listed at 6 feet, though is a shade under. He survived Tommy John surgery in 2010. He dealt with shoulder fatigue, elbow inflammation and a forearm strain in 2016. And he's 27, an age by which most flamethrowers have already lost some velocity (or moved to the bullpen). But he has maintained his mid-90s fastball. Only 16 starters have thrown harder this year.
Some days that's enough, but Salazar has always been just shy of a headliner. He has, for instance, the eighth-best strikeout rate in baseball -- just behind Stephen Strasburg and Noah Syndergaard -- since debuting in 2013, yet he has never struck out even a dozen in a game. He has averaged only 5.7 innings per start -- just behind Tom Koehler and Edinson Volquez -- and has completed only one game.
But this year he has made adjustments. Three, primarily, and all were in play in the loss Sunday -- when he followed up the bad first inning by striking out eight of the next 17 batters, allowing just two more hits.
1. He is throwing more changeups than ever.
Last June, ESPN's Mark Simon made the case that Salazar's changeup -- a pitch he learned from "a random guy" at a Little League game when Salazar was 8 -- was the most valuable pitch in baseball. One way to try to make it more valuable would be to throw it even more often, and he has.
In his first start this year, Salazar threw changeups on 38 percent of his pitches, the second-highest rate in any start in his career. The next time out, he threw a career-high 42 percent. In the past two, he's thrown 26 percent changeups, still well above his 18.5 rate last year. On Sunday, he threw at least one changeup in every count except for 3-0 and 3-2.
The extra frequency hasn't cost the pitch effectiveness, as he throws it for a strike three-quarters of the time and gets whiffs on 37 percent of swings against it. But more importantly, and perhaps relatedly, is that every other pitch he throws has become more effective. Here's the percentage of swings against each of his pitches that have come up empty, by year:
Before getting too excited: These are early numbers based on small numbers of pitches. Batters have swung at exactly two sliders, for instance, and all of 26 sinkers. But it makes sense that the more Salazar can put his changeup at the front of hitters' minds, the less prepared they'll be for anything else. Only one of his nine Ks Sunday came on a changeup.
2. He is throwing with more movement than ever.
Salazar throws three nonbreaking pitches -- four-seam fastball, two-seam fastball and changeup -- and all three are moving more this year than they ever have before, by about an inch per pitch more than last year and two inches more than in previous years. This doesn't guarantee a good outcome -- it might make the pitches harder to command, or bring his pitches back into the middle of the strike zone -- but the average exit velocity against Salazar has dropped from 91 mph last year (and 89 mph in 2015) to 87 mph this year.
3. He has lowered his release point, and his release points for each type of pitch are more tightly clustered.
See here what that looks like:
A lower arm slot generally leads to more horizontal movement. Tightly clustered release points are generally thought to help pitchers disguise their pitches better.
So there we go. Three adjustments, all of which seem like they should help Salazar take the next step. And lots of advanced metrics and peripheral stats, all suggesting he has taken the next step. Why isn't his excellence as apparent as Carrasco's?
The obvious answer is that he's been burned by bad luck, with the .440 batting average on balls in play that has floated his high ERA. But beyond that, each of Salazar's three adjustments comes with a bit of ambiguity. It's great that he has been able to throw his best secondary pitch more frequently; on the other hand, he might be throwing it too frequently. For instance, after blowing two fastballs past White Sox leadoff hitter Tim Anderson -- who entered the game hitting .159/.172/.222 -- Salazar threw an 0-2 changeup. Anderson could actually catch up to that one, and he doubled to lead off the game.
Or, consider Salazar's matchups against Avisail Garcia. In the third inning, Garcia got a first-pitch changeup in the zone. He grounded into a double play. The next time up, Salazar threw the same pitch twice more, in similar locations. On the second, Garcia doubled off the top of the wall to bring a run in.
The extra movement on his pitches seems great, but maybe it's also why he's grooving (slightly) more changeups and four-seamers this year, as pitches he starts on the edge of the plate tail into the middle of the zone. And if the movement is due to his lower release point, it might be an accident. Salazar's release points for different pitches have been tightly clustered this year, but from start to start those clusters have varied quite a bit. From an optimistic angle, it looks like Salazar made a smart adjustment. But from another -- the fool-me-twice angle -- Salazar looks like a pitcher with a history of wildness struggling to stay consistent with his release point.
When Mark Simon wrote about Salazar last June, Salazar had a 2.23 ERA. Simon quoted a scout talking about Salazar's excellent first half: "You could see this coming last year," the scout said. But Salazar walked five that night. His ERA during the rest of the season was 6.19, his home run rate doubled and his sub-6-foot body with the gunpowder arm broke down. That's always been the thing with Salazar, ever since he hit 99 in his major league debut: You could always see it coming, and sometimes you were even sure you'd seen it arrive, but he never really got it there. There sure is a lot of evidence saying this year is different.