Sunday, March 16, 2008

Reason #9: The Closer.

Joakim Soria was not the greatest Rule 5 pick of all time. He may not have been the best Rule 5 pick of last season, depending on whether or not Josh Hamilton can continue to tear baseballs and demons apart. But man, you’d be hard pressed to find a Rule 5 pick who, from the moment he first stepped on the mound in spring training, had people saying, “how the hell was he available?”

He was available only because he had spent essentially his entire career in the hinterlands of baseball: Mexico. The relative lack of Mexican representation in the major leagues is an interesting story, and I don’t know the reason why – consider that the population of Mexico in 2005 was estimated at 103 million; by comparison, the Dominican Republic had about 9 million. Mexico isn’t as baseball-crazy as the Dominican is – no country is – but it’s a popular sport there. Yet only 16 Mexicans appeared in a big-league game last season. Canada, with about 32 million people (and for whom baseball is a game you play during the three months a year that the ice melts), had 17 players in the majors last year.

I wonder if part of the reason is the fact that Mexico’s population supports its own summer league, the only independent summer league in any country in the western hemisphere. You would think that having their own league would help Mexico develop major league-caliber players, but so many of the teams play at high elevation that it warps the game. It’s very difficult for hitters to develop proper habits when they never see quality breaking pitches because of the atmosphere. Only five of the 16 Mexicans in the majors last year were hitters, and none of them are good hitters: Geronimo Gil, Humberto Cota, Alfredo Amezaga, Oscar Robles, and Juan Castro. On the other hand, if you can survive the high elevation as a pitcher, adjusting to sea level in the US must be a piece of cake – and the pitchers include Yovani Gallardo, Oliver Perez, Oscar Villarreal, Luis Ayala, Dennys Reyes, and Esteban Loaiza, all guys who have been successful in the major leagues. And Soria.

Anyway, with Soria you didn’t have to wait until spring training to know that the Royals had pulled off a huge coup. The Royals almost didn’t get him in the first place; there was a growing buzz about him at the winter meetings even before the draft, and there was some talk the A’s (who had worked out a deal with Tampa Bay to select the first pick) would take him. Instead they took Ryan Goleski. Thanks, Billy – you owed us one. Forty-eight hours later Soria threw a perfect game in the Mexican winter league, and the hype had begun.

Not to toot my own horn – oh, who am I kidding, of course I’m tooting my own horn – this is what I wrote about Soria in last year’s Baseball Prospectus: “When contemplating Mexican League statistics, it’s important to remember three things: 1) While ostensibly a Triple-A league, the level of competition actually falls between Double-A and High-A; 2) The league, as a whole, is a tremendous hitters league, even more so than the old PCL; 3) There is tremendous variation in altitude between teams, which makes park effects extremely relevant. The first two points make it virtually impossible for hitters to cross the Rio Grande, but the third makes the adjustment for some pitchers surprisingly easy. Soria played in Mexico City, which at 7,300 feet is 2,000 feet higher than Denver. As a result, his translated ERAs are actually better than his actual ERAs. He was drafted based on a scouting impression, but his statistical impression is just as good.

Soria was available because Padres GM Kevin Towers made a huge mistake by not listening to Randy Smith. (This may have been a first, someone making a mistake by not listening to Randy Smith.) Smith, the Padres’ Director of International Scouting, was (in Towers’ own words) “begging me to keep him” before the Rule 5 draft. As he said later, “I guess I should have listened to him.”

Soria made his debut on April 4th, in the Royals’ second game, with two men on base and the Royals down by two. He walked the first batter he faced, then gave up a sacrifice fly and a popout to end the inning. Two days later his name was called again, this time to start the 8th inning just after the Royals had broken a 1-1 tie with two runs. He allowed just a harmless single in a scoreless inning. Two days after that, Soria came in with two outs in the 7th, two men on and the Royals protecting a 2-0 lead; after walking the first batter, he stranded the bases loaded on a foul out, then pitched a perfect 8th inning.

And two days after that, Soria was so dominant in the eighth inning that Buddy Bell sent him out to pitch the ninth with a 6-3 lead; Soria retired the side in order again, striking out the last two hitters. Soria was the de facto closer from that point until Octavio Dotel’s return from the DL, and then returned to the role after Dotel was traded.

I suspect it’s highly unusual for a pitcher – any pitcher, let alone one who had never pitched above A-ball in the United States – to be utilized as his team’s primary setup man in his second major league outing. But to get the closer’s job in just your fourth appearance? That’s almost literally unprecedented. The only pitcher I’m aware of who earned the role more expeditiously was Salome Barojas, who was anointed the White Sox closer in spring training in 1982, despite the fact that he had never pitched in the majors. Five games into his major league career, Barojas had five saves; he finished with 21, which exceeded the total number of saves he would earn (14) in the rest of his career.

Soria’s a testament to good scouting – Louie Medina saw him in Mexico and lobbied the Royals to draft him even though he had virtually no American experience – but he’s also a testament to the power of statistical analysis done right. Soria’s translated numbers with Mexico City in 2006 (his Davenport Translation) rendered an EqERA of 3.49, which is simply amazing given the steep difference in difficulty between the Mexican League and the majors.

Soria’s success is the product of an excellent cut fastball, a changeup that he hides very well and rides in on right-handed hitters, and with two strikes, the occasional sloooow curveball that he apparently borrowed from Zack Greinke. (Here’s an outstanding analysis of Soria’s repertoire.) And he throws all his pitches with precision. As one AL Central front office source told me after getting a look at Soria in April, “he’s the Mexican Zack Greinke.”

In terms of pure stuff, there’s really not much difference between the two, and the Royals considered moving both of them to the rotation late last year before splitting the difference, moving Greinke while leaving Soria alone. If anything, Soria’s a victim of his own success. He was so good in relief last year, and the Royals have such a long history of brutal closers stretching back to when Jeff Montgomery lost his stuff, that they don’t want to mess with a sure thing.

The question is: should they? It’s the classic dilemma in modern baseball: do you take a pitcher that has proven they can succeed in a high-leverage relief role, and move him to the rotation, where he might give you three times as many innings?

Soria threw 69 innings last year. Factor in two additional weeks that he spent on the DL for precautionary reasons more than anything else, and you figure he’s good for 75-80 innings as a reliever. Projecting any starter for more than 200 innings is risky anymore, so let’s say that he’s worth 200 innings in the rotation.

Now factor in leverage. A closer’s innings will be necessarily more valuable than a starter because (presumably) he is being leveraged in situations where a single run allowed has far more impact on the game than it would in, say, the first inning of a tie game. Fortunately, at Baseball Prospectus we have a statistic to measure that, conveniently called Leverage. Soria’s Leverage was 1.53 last year, so roughly speaking you can argue that his 69 innings were as important as 69*1.53=106 innings from a starting pitcher would be.

But Soria’s Leverage last year reflects the time he spent in middle relief as well as his time as the closer. The Leverage of the other four closers in the division last year ranged from 1.59 (Bobby Jenks) to 2.09 (Joe Borowski). A typical closer has a Leverage rating between 1.7 and 1.8. So Soria’s 75-80 innings in relief would be the equivalent of about 135 innings as a starter.

Then you have to account for the fact that, almost without exception, all pitchers will be more effective in relief than in the rotation. This is not a controversial statement, but the size of that difference might be surprising. Research that Nate Silver did as part of his annual improvements to PECOTA showed that, if you hold all other factors equal and move a reliever into the rotation, his ERA will rise a full 25%.

Now, Soria had a 2.48 ERA last season; tack on 25% and you’re at 3.10, and a starter with a 3.10 ERA is a damn sight more valuable than a reliever with a 2.48 ERA, Leverage be damned. But what if Soria’s true talent is more in the 3-3.5 ERA range? Would you rather have a closer with an ERA of 3.20, or a starter with an ERA of 4.00? In that case, you’d still want the starting pitcher. According to Nate, in fact, “a 2.00 ERA closer is roughly as valuable as a 3.69 ERA, 200-inning starting pitcher.”

Not all pitchers improve equally when moving from the rotation to the bullpen. As Nate found in a subsequent article, there are certain factors that make it more likely that a starting pitcher will blossom in relief. Those factors are 1) a high strikeout rate; 2) a high walk rate, i.e. poor command; 3) lots of isolated power, i.e. a flyball pitcher who gives up homers.

In other words, a pitcher with great stuff and little idea where the ball is going should see more improvement, moving from the rotation to the bullpen, than average. Nate brings up the examples of Bobby Jenks, Jonathan Papelbon, and J.J. Putz as guys who took to the bullpen like a fish to water. Conversely, that means that such a pitcher who was already in the bullpen would struggle more than average if he was moved to the rotation. For that reason, Nate argued against the idea of moving Papelbon back to the rotation after his rookie year (an opinion the Red Sox eventually agreed with, and an opinion that was borne out last year.)

But look at Soria. Soria had a terrific strikeout rate, 75 in 69 innings, true. But he also had phenomenal control – just 16 unintentional walks in 69 innings – and was almost impossible to hit for power; he surrendered just three homers, a triple, and eight doubles all year. His isolated power against was just .077, which is less than Tony Pena Jr’s isolated power (.089) last year. Soria is the antithesis of the Rick Vaughn closer stereotype – his strikeouts were the result of movement and placement, not from just blowing the hitters away with high heat. He would seem to be, in other words, the type of pitcher that would adjust better to the rotation than most relievers.

I understand why the Royals are keeping him in the bullpen, because it’s just so easy to look at him and envision a young Mariano Rivera on the mound. There are some visual similarities, and of course Soria has a great cutter, and like Rivera he used that cutter to just saw off the bats of left-handed hitters last year (they hit just .167/.217/.229 against Soria, which is filthy). And like Rivera he gave up very few extra-base hits.

But we shouldn’t be making momentous decisions like this based purely on his superficial similarities to one admittedly unique pitcher. Rivera has thrived in the bullpen as essentially a one-pitch pitcher, but Soria has four good pitches, he’s young, he’s worked as a starter for most of his pro career, and there’s a lot of statistical evidence to suggest he will adapt to the rotation just fine.

The Royals need their security blanket for now, so they’re leaving Soria in the closer’s role, but they have not shut the door on him returning to the rotation in the future. Unfortunately, time will shut that door for them soon enough. Assuming the Royals aren’t actually in a pennant race in the second half and can afford to experiment for the future, they need to put Soria in the rotation for the last month or two of the season and see what they’ve got. Worst-case scenario, in 2009 he goes back to being the team’s first bona-fide closer since Jeff Montgomery. Best-case scenario…the Royals have something even more valuable than a bona-fide closer.

In whatever role they use him in, he’s a joy to watch. Good thing we’ll be watching him in Kansas City for the next five years.


Anonymous said...

I think Soria is a potential future Cy Young contender. The sooner they make the move to the rotation, the better - although they should ease him into the rotation by monitoring his increase in innings each year closely.

Anonymous said...

Soria has a chance to be a legit No. 2 starter. I do not see the him becoming an ace, but a rotation of:
is not too bad. The bottom line is that the Royals will need to develop a closer. They are way too expensive to obtain via free agency.

Brett said...

I think in a perfect world, Carlos Rosa would be our closer in a year or two. The best bullpens are filled with your own failed or excess starters.

Clint said...

there's NO reason to not give Soria 10 starts this season.

Absolutely none.

Antonio. said...
This comment has been removed by the author.
Antonio. said...

Yes, closers are very expensive to attain in free agency, but starters aren't exactly cheap either.

I am a huge fan of moving Soria to the rotation.

Anonymous said...

I'm not sure the Royals are as hesistant to move him to the rotation as some seem to think. Posnanski points out that history shows that GM's don't make a young pitcher a closer and then move him to the rotation, basically ever. So there's a temptation to base our probability on history alone and assume it won't happen. Looking at it more practically though, the Royals don't really have another candidate for closer right now, and Soria probably needs to start out in the bullpen regardless so as to keep his innings down (I forget who it was, probably someone at BP, who showed that pitchers who increase their previous season's total by more than 25 innings are X times more likely to be injured). Moore understands better than anyone that he shouldn't be pitching 200 innings this season. You can keep a pitcher from 200 innings by keeping him on low pitch counts (pulling him from each game after 90 or so pitches) or by having him pitch half the season in the bullpen, half the season as a starter, a la Santana, Liriano, et al (Should we call it the 'Twins model'?). I do believe that there's a specific recipe for Soria's move to the rotation which looks like this: One or more pitchers in our opening day rotation pitches like Odallis Lima (probable) + the Royals are not in a pennant race (probable) + one of our other bullpen guys steps up and shows closer ability (not so probable). If those three things happen, Soria's a starter.

Anonymous said...

Oh yeah, and Brett, we did indeed write posts together on Joe Bazinet's blog the Pine Tar Rag. Not sure whatever happened to Joe, but that was a solid blog I thought. Good to hear from you again.

Minda said...

I had never been able to figure out if I liked the idea of moving Soria to the rotation. I think I was mired in the security of knowing he was there to shut things down in the (8th and) 9th. But this post makes me really, really, REALLY want to see what he could do as a starter.

Also, the quote about him being "the Mexican Greinke" made me laugh really hard. Rany, I very seriously love this blog. Thank you for doing what you do.

KB said...

If Greinke and Soria have similar abilities, etc. it would seem much wiser to make Zack the closer and put Soria in the rotation. Based on personality/history, I think Zack getting bored with only being in the game every five days will turn out to be a major factor.

Brett said...

I thought that was you. I enjoyed that blog a lot. I was dissapointed when Joe just disappeared. I like the idea of giving soria 10+ starts at the end of the season. Too bad we don't have a lot of in house canidates to close, but hey, if Joe Nelson can do it, then someone in our system should be able to.

Anonymous said...

NOBODY is factoring in the fact that Soria has already had elbow problems and required surgery for that. I'd be curious to see what take is on moving him to the rotation when considering this factor. It would seem like pitching an inning every few days (or 4-5 innings in 7 days) would be a lot easier on an elbow than starting would be.

Anonymous said...

I might be crazy, but I think the biggest reason that the Royals are leaving Soria as a reliever is actually his inning count. As we all know 200 innings or so per year is about the max that a pitcher can safely throw. After that there is a pretty steep drop in performance(See Bannister last year, those last 2 starts put him at around 230 innings pitched(both in the majors and winter ball) and it was obvious he was struggling).
Now if I remember correctly Soria was still playing winter ball in Mexico as a starter and was logging about 100 or so innings. So it would seem to me that until he stops playing winter ball, that the Royals will use him just in relief since about 100 innings would be his max.

Anonymous said...

Soria did play Winter ball in Mexico. But he pitched as a closer. Limited action in the season, about 10 innings. But substancial action in the playoffs, about 12 - 18 innings.
His team actually won the championship, after 27 years of constant dissapointment.

Antonio. said...

My idea is to go with a month by month basis...I think he'd throw about 130 or so innings, which would be more than the 25 innings increased posted above, but I was thinking BP had a higher increase listed. I know it's just one guy, but Verlander seemed to do fine with his increase.

Closer for April, set-up for May, short relief for June, long relief for July, swingman/spot starter for August, starter for September...with a shut down if necessary.

Antonio. said...

So no Latinos start in the major leagues? There are plenty of players that play in the Winter Leagues. And plenty of them are pitchers and plenty of them are starters. And his elbow surgery happened a long time ago. The situation and recovery has changed. Again, no one who wants Soria to start expects 220 innings out of him by this year. But by the time the Royals are ready to contend, it would be nice to have him ready to go 200 or so innings.

Anonymous said...

"Moore understands better than anyone that he shouldn't be pitching 200 innings this season."

Do we really know that, or are we just ascribing knowledge to the guy based on an assumption that we think DM is good general manager?

If anyone should understand that a pitcher like Soria shouldn't exceed 200 innings, it would be someone like Dusty Baker. Of course, Baker never learned from his failures.

To my knowledge, Moore hasn't had that sort of failure yet to learn from, although he certainly wasn't quick to step in at the end of last season when Bell started abusing both Meche and Bannister...


Anonymous said...

I love your stuff, and your comments on Mexican baseball were really insightful.

However, how can you miss that our Royal's own Jorge De La Rosa was born in Monterry, Mexico? Maybe HE's the one who could benefit from the pen, and move Soria to the rotation?