In his first season in a Nationals uniform, Rafael Soriano came to represent many things. He symbolized the Nationals’ win-now philosophy, a splashy $28 million addition late last offseason to a bullpen that appeared already set. He was also often the punching bag for the failures of the bullpen; his blown saves and seemingly aloof personality contributing to the team’s disappointment 2013 season.
Soriano’s first of two guaranteed seasons in Washington had mixed reviews. Like any pitcher, he had his peaks and valleys. His demeanor on the mound — drawing in the dirt and talking into his hat before his first pitch, and the signature untucking of his shirt after a save — belied his personality. Some of his production was good — a 3.11 ERA, 68 appearances, 43 saves (tied for fifth in baseball) and 2.3 BB/9 — but some wasn’t. He blew a career high six saves, the hits he allowed per nine innings jumped from 7.3 in 2012 to 8.8 in 2013 and his strikeout rate fell dramatically from 9.2 to 6.9. He will be 34 next season.
Thanks to the help of Pitchf/x guru Harry Pavlidis, we took at deeper look at Soriano’s season. How effective of a closer was he? Why did it seem like he had so few 1-2-3 innings? An examination of detailed statistics leaves us with a few conclusions: Soriano is an aging pitcher whose stuff, understandably, is losing velocity and effectiveness. He struggled with his slider much of the season and, during tough stretches, was fighting his own command, leaving pitches up in the strike zone. And although his walk numbers weren’t bad, it felt like he had an uncanny ability to walk a batter at an inopportune time and then cough up a crucial, save-blowing hit after it.
Below is a chart with a ton of information on closers who notched 30 saves in 2013. It is perhaps too much information to digest in one passing but each statistic is worth considering in context of the others so we’ll break it down one by one.
Soriano was on the lower end of the scale of effectiveness in this group. In terms of batters faced per inning pitched (BFpIP), Soriano was near the bottom of this group with an average of 4.16. Greg Holland, the Royals’ 28-year old closer and hard-throwing right-hander, led the group with 3.81 batters faced per inning pitched. He was closely followed by Minnesota’s Glen Perkins (3.83) and Atlanta’s Craig Kimbrel (3.85), perhaps baseball’s best closer. There could be any variety of explanations but the most obvious one is that those three pitchers have a combination of excellent arms, stuff and control. And because of it, they could accomplish the most basic principle of pitching: get outs and limit baserunners.
So just how effective, or ineffective, was Soriano? How many of his 68 appearances in 2013 were clean? For these purposes, clean innings are defined as appearances in which a pitcher faces three batters and gets three outs. An inning is still regarded as clean if, for example, the pitcher gives up a hit or walks a batter but still gets out of the inning cleanly, such as with the help of a double play, outfield assist, caught stealing or pick-off.
In 2013, Soriano had only 21 clean appearances and 20 clean saves. Only 31 percent of his appearances were clean, which ranks in the lower half of this group. Oddly enough, Soriano had a clean save rate of 47 percent, which ranks in the top half of the group. (Perkins paced all closers in both categories with an astounding clean appearance rate of 46 percent and clean save rate of 61 percent. That’s reliability.)
In terms of clean saves, Soriano actually outperformed his expected performance. The line in the graph below is the expected performance based on the batters faced per inning of the group. Each dot is each player’s actual performance, with Soriano as the dot circled in red.
Now, look at where Soriano falls in terms of clean appearances in the chart below. He falls in line with his expected performance, which is near the bottom, based on the batters he faced per inning. In simpler terms: Soriano was better at getting quick and clean innings in save situations than he was at getting clean innings in non-save situations.
One theory for this difference, suggested by Pavlidis, could be simply be random variance from season to season which can’t likely be sustained. But another theory could be Soriano’s mentality and pitching in non-save situations. On July 25, bench coach Randy Knorr, acting as manager for the ejected Davey Johnson, didn’t like how Soriano was throwing the ball in the ninth inning of a non-save situation that he inherited. So Knorr yanked Soriano after allowing two runs and putting two more men on and replaced him with rookie Ian Krol.
Some of the numbers in 2013 support that notion, albeit based on a small sample size. Soriano had a 2.55 ERA over 17 2/3 innings of non-save situations but a 3.31 ERA over 49 innings of save situations, which doesn’t support that theory. (A caveat: the large difference in innings is a major factor.) Soriano, however, may just have been more lucky. In those save situations, he held opponents to a .645 OPS. In non-save situations, he was hit harder, a .729 OPS. Over his career, there isn’t much different between the two scenarios: a 2.63 ERA and .582 opponents’ OPS in save situations as opposed to a 2.64 ERA and .609 opponents’ OPS in non-save situations.
But why was Soriano so ineffective and inefficient? The last four columns of the first chart give clues.
Soriano was 33 during the 2013 season and, with age, fastball velocity declines. Velocity isn’t everything for a pitcher but more speed means a larger margin for error. According to Fangraphs.com, Soriano’s velocity has steadily declined from 94.6 mph with the Braves in 2007 to 92.6 and 92.2 in 2011 and 2012 with the Yankees. In 2013, Soriano’s fastball averaged 91.9 mph. Soriano’s cutter is factored into that average because, according to him, he has a variety of fastballs, including a cutter, when he alters his grips.
Soriano’s fastball was among the slowest of the group but that’s not solely an cause of ineffectiveness. Edward Mujica, who has reportedly agreed to a deal with the Red Sox, relies almost exclusively on a splitter, which averaged 86.4 mph last season, and not a traditional fastball but still had a dominant first half of the season before he faded. The great Mariano Rivera’s cutter averaged around 91 mph last season, and has hovered around there for nearly five seasons, and he has thrived. Sergio Romo, of the Giants, doesn’t normally throw his fastball harder than 88 mph but he still gets plenty of swings and misses.
That is where Soriano struggled last season: missing bats. Holland led the entire group with a 39.3 percent swing and miss rate (shown on the chart as 0.393 whiff/swing). He was followed by Reds’ flame-thrower Aroldis Chapman (38.5 percent) and Angels’ Ernesto Frieri (33.6 percent). Soriano, on the other hand, was dead last with a 19.6 percent swing and miss rate. This is an indicator of the effectiveness of the pitcher’s stuff and Soriano’s stuff just wasn’t good as normal in 2013.
Throughout the season, he admitted to losing feel of his slider and that it lacked its traditional hard bite that missed bats. He noticed a flaw his mechanics; he wasn’t getting his arm high enough in his delivery and the lost confidence led to a career-low usage of the slider According to Brooksbaseball.net, opponents slugged .364 against Soriano’s slider in his career but last season they slugged .531 against it, the highest mark against any of his pitches.
Part of the problem — but perhaps also a reason for success, too — may have been Soriano’s ability to throw strikes. He was fourth best in the group with a 45.1 percent in-zone rate. A lot of his pitches were in the strike zone, and that’s actually lower than rates earlier in his career. The problem is that hitters were making more contact with his offerings, likely because of frequent high strikes and less sharp stuff. According to Fangraphs.com, Soriano’s opponents made contact with his pitches 81.5 percent of the time they swung in 2013 compared to his career norm of 75.6 percent. Some pitchers, like Frieri, can miss a lot of bats while throwing a lot of strikes: a 48.5 in-zone percent and a 33.6 swing and miss percentage, which is probably a dangerous combination long-term.
Despite being hit more often, Soriano seems to have gotten away without major damage. Opponents’ slugging percent when they made contact with any of his pitches was .459, around the middle of the pack. When opponents made contact with other pitchers’ stuff, such as Frieri or Padres’ Huston Street, they caused major damage to the tune of a .618 and .556 slugging percentage, respectively. Joe Nathan, who recently agreed to a two-year deal to be Detroit’s closer, had a measly .331 slugging percent against because opponents simply couldn’t make good contact with his stuff when they did connect.
Soriano’s issues in 2013 may have had more to do with stuff, mechanics and command than pitching approach. He has thrived in the past by throwing a lot of strikes and did well, to some extent, last year by doing the same. When his stuff was flat, he got hit hard. When he left pitches up in the zone, he got hit.
Soriano has logged relatively little wear and tear on his arm because of his unusual career path and previous injuries; he has averaged only 47 innings a season over 12 years. But even then, at 33, a decline is stuff is natural. Other older relievers have thrived late in their careers. Despite losing velocity over the years, Nathan is still one of the most effective closers in baseball at 38. Rivera closed until he was 40. Pittsburgh’s Jason Grilli had a career renaissance as a closer at 36. History shows that Soriano can still be an effective closer and he was during stretches of 2013 with his low 90s fastballs and array of pitches. It’ll just take some improvements.