Heartburn Hardball: Starting Over

It’s that time of year again: spring, the season of growth and change when life returns to the world, in baseball as in nature. For baseball fans and writers alike, it is a season dominated by prospects blooming into Opening Day starters, free agent closers descending on the bullpen like choking ivy, and Rich Harden blossoming onto the Disabled List (poor Rich has been done for the season since early February after aggravating the same shoulder capsule for the fifth straight year).

It is also the time of year when that young reliever with filthy stuff your team brought up for late-inning leverage gets his annual look as a starter. Sticking a young starter from the minors into the major-league pen for half a year to get him some seasoning and then plugging him into the rotation the next spring is a time-honored tradition; the Baltimore Orioles under Earl Weaver and company used this process to develop most of the guys who would start for them during that team’s heyday in the late Sixties and Seventies, including Hall of Famer Jim Palmer, who remains a vocal advocate of the process to this day, but they were hardly alone in doing so. However, since Dennis Eckersley’s fateful union with manager Tony La Russa in Oakland in 1988, the pendulum swing towards the closer over the past two decades and change led to some of those promising young minor-league starters never leaving the bullpen, lingering there to give their skippers late-inning certainty while collecting the big paychecks that come with that peace of mind.

It would be a mistake to give the introduction of sabermetric thought to baseball’s front offices too much credit for that pendulum swinging back the other way, but it certainly didn’t hurt. Most good baseball men and women, whatever their background, know that generally speaking, a good starter is more valuable to a team than a great reliever. Shaun Marcum, for instance, was the second starter on the Milwaukee Brewers last year and threw 200 innings of 3.54 ERA ball in one of the lighter-hitting divisions in baseball during a great year for pitching. At 2.9 WARP, he was more valuable than any closer in the league last season, and after Craig Kimbrel (2.2) and Jonathan Papelbon (2.0), it wasn’t particularly close.

The guy who led the league in saves, Detroit’s Jose Valverde, was worth a measly 0.7 WARP. His near-mediocrity as a pitcher was actually part of his charm: he was a two-pitch pitcher who struck out twice as many batters as he walked, but neither of his pitches broke too well, so when guys swung, they’d connect. Unfortunately for them, they’d usually connect right into a Tiger defender’s mitt, and even when they did get on base, it was rare that they’d be able to score two or three runs before Valverde got three of them out. And once he did, the fun began; Detroit’s closer has a post-save celebration routine that puts K-Rod’s famous Anaheim-era fist-pumping to shame. Is it appropriate? Probably not, but retiring three major-league hitters isn’t easy, and it’s hard to begrudge him his fun after the magic he worked on the Yankees in the last year’s ALDS.

But guys like Jose Valverde have been kicking around the league for years, and they generally don’t stay in one place for too long: once their team control expires, they get very expensive very quickly. There are more than enough teams out there that will pay $8-9 million a year for a guy like Valverde because they think he’s that final piece they need to get into the playoffs—and he’s just a journeyman with two decent pitches who doesn’t listen when people say it’s harder to pitch in the ninth inning. When the likes of Papelbon or Francisco Rodriguez hit free agency, teams start throwing $12-13 million at them for 70 innings or so of work every year, because that’s the premium great closers command.  

In today’s market, it’s only marginally more expensive to acquire a more dependable front-line, non-“ace” starter. C.J. Wilson’s new deal will pay him $15.4m annually; Mark Buehrle’s has a $14.5m AAV. Given that MLB free agent contracts are constantly growing, something that will continue under the new CBA, the “heavier” final years of any long-term deal a young starter signs today will look much better down the line as long as he doesn’t completely fall apart. Zack Greinke signed a four-year extension with Kansas City in the offseason before his Cy Young Award-winning 2009; he’ll make $13.5m in Milwaukee this season, which is just about right for him. If he returns to his 2009 form, that price tag becomes a steal.

With all of that in mind, here are profiles of three relievers who are trying to make the jump from the bullpen to the rotation this season—and one who isn’t.

Neftali Feliz, Texas Rangers
Feliz is in many ways the platonic ideal of this starter/reliever quandary. Feliz, a native of the Dominican Republic, was signed by the Atlanta Braves in 2005 at age 17 and remained in their system until they dealt him to Texas as part of their deadline acquisition of Mark Teixeira. During his stay in the Atlanta organization, the Braves gave him both starting and relieving opportunities in rookie ball and Low-A, but once he was dealt to the Rangers he joined the rotation full time, starting all 27 of his appearances in A-ball and Double-A as he moved through the Texas organization in 2008. The next season, his last in the minors, he was promoted to Triple-A Oklahoma City. Feliz started about half of his appearances there and spent the other half finishing games, though not in save situations—he averaged over three innings per appearance in 25 games with 13 games started, six games finished, zero saves, and a 3.49 ERA in the notoriously hitter-friendly Pacific Coast League. His K/9 dropped a bit moving up from Double-A, but his BB/9 stayed about the same, and more importantly scouts still loved him; Kevin Goldstein had him as the top prospect in the Rangers organization in January 2010, and he was not alone.

Going from starting in the launching pads of the PCL to closing in the AL West, which outside of Texas is surprisingly pitcher-friendly, Feliz was able to put a few more MPH on an already-electric fastball and turn his 3.49 Triple-A ERA into a 1.74 major-league ERA. The trick, of course, is that those numbers can’t be compared apples to apples, since bush-league stats can’t always be trusted to translate to the bigs. For an extreme example, see John Axford’s minor-league BB/9 or Chris Davis’s minor-league OPS. In a general sense, it “should” be the case that a player performs better in the minors than he does in the majors due to weaker competition, but maybe his role changes in MLB, or he figures out how to command his fastball like Axford did in 2010, or—like Davis—he’s a creation of the PCL who struggles to lay off superior breaking balls.

Whatever the reason for his post-promotion improvement, Feliz came up as a closer and dominated for Texas in 2009. There was talk that he might move into the rotation, if only because the Rangers desperately needed another starting pitcher. However, over the winter before the 2010 season, the Rangers changed ownership, and incumbent General Manager Jon Daniels suddenly had an actual payroll to work with. Feliz remained in the closer role while Daniels got pitching help elsewhere in the form of Cliff Lee, dealing Justin Smoak to Seattle for the lef-hander. While Lee was a bit of a “disappointment” for the Rangers considering that he’d maintained a wizardly 14.8 K/BB as a Mariner, he was still worth the pickup, and with C.J. Wilson, Colby Lewis, and an outstanding second half from Tommy Hunter, Texas was set with starting pitching going into the 2010 postseason.

The Rangers would lose the World Series, Lee would leave for Philadelphia, and again Neftali Feliz was considered for a starting job. Instead, Daniels and the Texas staff promoted Feliz’s fellow reliever Alexei Ogando to the rotation. Why? Daniels could be confident that between Matt Harrison and Derek Holland, he had at least one rotation slot; Wilson would be the ace, Lewis the number-two, and Hunter the fourth or fifth starter. That left an opening for one guy. At the time, Ogando and Feliz were both essentially two-pitch fastball/slider pitchers who technically had changeups available, though neither was very good with it. But Feliz was the closer, Ogando wasn’t, and Ron Washington loves his closer, so Ogando went into the rotation.

In 2011, Alexi Ogando was able to pitch to a mid-threes ERA in 167 innings with essentially two pitches. He threw a changeup occasionally, and according to Pitch F/X, he had a horrific “sinker” that was either a ball or put in play—over 85 percent of the time hitters swung at that pitch they at least made contact, and one out of every five of them was put into play. Assuming Ogando wasn’t trying to develop a sinking fastball during regulation play, those are probably changeups he didn’t grip quite right. Even given all of that, his fastball and his slider were still good enough to make him a 2.7 WARP starter, which is more than acceptable considering both Matt Harrison and Derek Holland panned out, and Colby Lewis and C.J. Wilson turned in quality seasons too.

But now Wilson is in Anaheim, Tommy Hunter was shipped off to Baltimore because he was always secretly a swingman in disguise, and the Rangers once again need another starter—and the Angels just turned Texas’ ace into a number-three. So Feliz is getting another look in the rotation, and considering the Rangers gave up two pretty decent prospects for the Padres’ Mike Adams last year, they’ve likely got a ninth-inning succession plan already in place.

First and foremost, Feliz is going to lose some movement and speed from the fastball Texas fans are accustomed to seeing from him. He didn’t have the best 2011 in terms of his peripheral stats, and even traditionalists will note that he blew twice as many saves last year as the year before. A lot of that had to do with his fastball being a bit more hittable than it used to be. It’s going to continue to be like that, because unless Feliz plans on leaving his 2012 starts in the third inning, he can’t throw his heater as hard as he has in the past.

Feliz has also been working on a changeup. To date, his experiments with it at the major-league level have gone about as well as Ogando’s—that is to say he, too, has been known to throw a suspiciously terrible “sinker” every once in a while. However, last season he was at least able to induce groundballs off of the thing, even if hitters weren’t missing it all that often.

Luckily, the Rangers have Mike Maddux, who is probably the best pitching coach in the league now that Dave Duncan in St. Louis has followed Tony La Russa into retirement; if there’s anyone who can turn a two-pitch reliever into a major-league starter, it’s him. Feliz and Maddux are reportedly focusing on getting his changeup where it needs to be for the 2012 season; Feliz probably has a better chance than Ogando, given that it was an MLB-ready third pitch before he more or less dropped it when he came up to close. Yu Darvish is a fine acquisition, but there aren’t any more Cliff Lees out there to defer this decision another year; the Rangers are thinking World Series, and Feliz will have to either sink or swim as a starter. PECOTA thinks he’ll pitch 144 innings with a 3.47 ERA and a whole bunch of strikeouts; I tend to agree with that assessment.

Chris Sale, Chicago White Sox
When the White Sox took Chris Sale with the 13th pick of the 2010 Rule IV draft, he was a college starter from Florida Gulf Coast University generally considered to be one of the most MLB-ready pitchers in the class. The White Sox agreed: they signed him to a deal that gave him a bonus at slot ($1.656m) but that guaranteed him a fast-track to the MLB bullpen by starting his service time clock as early as possible.

The Sox did exactly what they promised; Sale signed with Chicago in June of 2010 and was in the majors by August. It helped that Sale was exactly what his college career promised: a left-handed strikeout artist who didn’t walk a lot of guys and could embarrass right-handed hitters just as badly as lefties. But he didn’t close—in 2010, that was Bobby Jenks’s job, and in 2011 it went to Sergio Santos. Considering both those men have since departed, Jenks in free agency and Santos in the deal that brought pitching prospect Nestor Molina over from Toronto, it was assumed in some corners that Sale would compete with Matt Thornton and Addison Reed for that role; instead, it looks like Sale will be going back into the rotation.

Sale isn’t following an entirely unusual path for a college starter; his is just highly accelerated. Both David Price, who came out of Vanderbilt at the top of the 2007 draft, and Stephen Strasburg, who came out of San Diego State University as the first-overall pick two years later, spent more time in the minors than Sale did; it’s hard to say Sale is “better” than either of those two at anything (except staying healthy, in Strasburg’s case), but he had a remarkable turn-around time from the draft to his first contributions to the team. That helps the White Sox, sure, but it also helps him: he’ll become a free agent before just about everyone in his class except Cincinnati’s Mike Leake, who didn’t even spend Sale’s 10 innings in the minors: his contract put him directly on the Reds’ 25-man active roster.

Sale is probably one of the “safer” picks to go from starter to reliever back to starter because he has two well-developed off-speed pitches to complement his fastball—he uses his changeup twice as often as Feliz does in relief, about a third of the time batters swing at it they miss, and two-thirds of the time they put it in play it’s a ground ball. Sale’s best pitch is still his slider, but the changeup won’t embarrass him, and if he develops it further, it could turn into a great out-pitch.

PECOTA has him at 120 IP with 143K and a 2.81 ERA this year; if I had to guess, I’d put the IP a bit higher, the strikeouts a bit lower, and the ERA around 3.40. To be fair to PECOTA, it is but a computer, and Sale is drastically changing the way he contributes to the team. However, if Chris Sale can throw 120 IP of 2.81 IP ERA ball while striking out two dozen more guys than innings pitched but can’t go any longer than that, Chicago should slot him into a Goose Gossage-like super-reliever role and leave the starting to Peavy, Floyd, Danks, and maybe Nestor Molina, if he turns out to be the starter the White Sox hope he is. We won’t know until next offseason (if then), but Sale is definitely a guy to watch.

Daniel Bard, Boston Red Sox
The Red Sox are in a curious position this spring, stemming specifically from the perception that a team that won 90 games last year has something to “come back” from. Despite the somewhat-embarrassing circular firing squad in Boston this winter, the Red Sox are a team with two top-flight starters in the American League, three everyday players who were in the MVP conversation as late as last September, and the depth and budget to keep all of their core guys while getting younger and better over the next few years. New manager Bobby Valentine has walked into the perfect job: as long as he keeps pretending the Red Sox are actually underdogs, he’ll look like a mythical hero when they win the division.

Which brings us to Daniel Bard. Bard, the setup man for departed closer Jonathan Papelbon last season, is another young, two-pitch guy with an alleged changeup trying to move from late-inning relief to a starting role. Unlike Chris Sale, he won’t reap the benefits of facing AL Central lineups for most of the year, and unlike Neftali Feliz, he doesn’t have a fastball that hitters miss. Feliz gets a whiff rate around 30% percent on his fastball; Bard’s is below 20 percent. That would be fine—or at least better—if he could control the thing. However, only 15 percent of his fastballs are called strikes. If he were missing bats, that would be enough, but he’s not. On a general control level, out the three guys profiled here, Bard has the lowest percentage of first strikes to start an at-bat—50.4 percent compared to Sale’s 54.5 percent and Feliz’s 51.7 percent. Is it possible to come back from 1-0 to record an out? Absolutely. But neither Sale nor Feliz is a control pitcher, and both have problems with falling behind in the count—Bard just has more of them.

The Red Sox have a lot of confidence in Bard’s changeup, specifically because it induces groundballs—about two-thirds of the balls put in play off his change are on the ground, and only five percent of them are line drives, which is even better. However, Bard has a mystery “sinker” as well, and it’s somewhat concerning that like Sale and Feliz, this “sinker” has essentially the same batted ball profile as his fastball—swing and whiff percentages almost the same, GB percentage similar, and so on—except it gets jacked. Almost 40 percent of these sinkers were put in play as flyballs, compared to 22 percent of Bard’s regular four-seam fastball, and none of them were pop-ups. More concerning, Bard threw only 83 changeups last year, but 46 “sinkers.” Either Bard is trying out a pitch that combines the terrible command and inability to miss bats of his fastball with a dangerous proclivity to go yard, or—as I suspect is the case with both Sale and Feliz—these “sinkers” are changeups that don’t change. They lose a couple MPH like real sinking fastballs, but the break charts show they don’t actually sink—they’re probably better classified as mistake pitches.

It’s curious, too, how Bard started throwing the “sinker” only in 2011—and the percentage of his changeups put in play as line drives decreased from 23 percent in 2010 to five percent. About 40 percent of the time Bard throws his changeup, he makes a mistake with it. That’s not a third pitch; that’s a disaster waiting to happen.

Sale has a sinker, too but his actually sinks. He gets twice as many ground balls off of it than off his normal fastball, and it actually comes out of his hand slightly faster than his normal four-seam offering. When it arrives, it arrives at the knees or slightly below. That’s a sinker. Bard’s ends up in the wheelhouse. Bard threw 1109 pitches last year; those 46 “sinkers” resulted in 10 percent of his flyballs. Bard is going to need to fix his fastball and changeup if he’s going to be successful next year as a starter. PECOTA has him at 132 IP and a 3.52 ERA; both of those numbers are low, in my opinion. Absent him coming out of the gate with a fastball he can locate and a solid changeup, I think a mid-fours ERA is his best-case scenario, and while the Sox won’t be happy about it, they’ve really left themselves no other choice but to grin and bear it. But I’ve been wrong before. And really, a 4.50 ERA pitcher on that team should win 12-14 games, so Boston will probably be more than happy with him unless the conversion (reversion?) goes completely sideways.

Jim Johnson, Baltimore Orioles
Unlike the three gentlemen profiled above, Orioles reliever Jim Johnson does throw a sinker. He throws it more than his fastball; it’s the mainstay of his arsenal. Also unlike the three men profiled above, Johnson will not be joining his team’s rotation.

Johnson joined the Baltimore organization in 2001 at the age of 18 and started 122 of the 133 games he played in the minors between then and his major-league debut in 2008. When someone puts his sinker in play, about 60 percent of the time it’s a groundball. Only a tad over 16 percent of the time is it a line drive; the rest of the time it’s a flyball. It’s a pretty good pitch.  He was brought into an abysmal bullpen in 2008 and essentially told to sink or swim, and since then he’s been the only consistent reliever the Orioles have had in the late innings as impotent free agents have come and gone.

But as bad as the Orioles bullpen has been over the last five years, their starters have been worse. Jeremy Guthrie was the only thing close to a stabilizing presence the rotation had for a good five years, and he’s now starting for the Colorado Rockies. Over the last few years, the bullpen has pitched about half of the innings thrown by all Orioles pitchers, and Johnson himself recorded a career-high 91 IP in relief in 2011.

Going into this season, the Orioles don’t have a single pitcher who projects anywhere better than a number-three starter, and that’s Wei-Yin Chen, one of their acquisitions from the NPB. Zach Britton could bounce back from a five-plus ERA, 1.0-WARP season, just like Jason Hammel could figure out how to strike people out again and Brian Matusz could rediscover how his fastball works. But in Jim Johnson, the Orioles have a guy who can actually get people out at a reasonable rate, whom they didn’t have to buy low on, and who, given his minor-league resume, should have more than a grand total of one MLB start (it was in 2006 during bizarre July one-game callup, and it ended in disaster because he was 23 and pitching at Double-A at the time). Near the end of last year, there was a lot of talk of him getting his second start as the season wound down, but it never happened.

As the Orioles camp solidifies down in Sarasota, there’s every indication that instead of starting, Johnson will be competing with Matt Lindstrom, the other guy who came over in the Guthrie trade, for the closer role—something that would limit the number of innings he’d pitch. Considering he was the workhorse of the bullpen and its best pitcher last season, that seems like a bad idea. Other serious candidates for the rotation include guys with career ERAs over five and Dana Eveland, whom the Orioles traded for because they were impressed by his performance at Triple-A last year, Johnson should’ve been a guy who got a second look.

Unfortunately, he came to camp with back problems and is just now starting to throw off the mound. The Orioles will be taking it easy with him, as they should, nixing any outside chance of his joining the rotation. That’s unfortunate, but considering how much of a lost season this already is for Baltimore, it might be for the best that Lindstrom win the closing job to improve his trade value while the Orioles use Johnson as a fireman. PECOTA puts him at 79.1 IP and a 4.06 ERA; even with the caveat that ERA is less than useful in evaluating relievers, I still think he’ll beat that number. That or his back problems are worse than feared; as gloomy as next season looks for the O’s, they cannot afford to be without Jim Johnson for any extended period of time. He’s the only thing left standing between sanity and 80+ innings of Kevin Gregg.