keyboard_arrow_uptop

Every day until Opening Day, Baseball Prospectus authors will preview two teams—one from the AL, one from the NL—identifying strategies those teams employ to gain an advantage. Today: two AL teams! Projected division winners Angels and their diverse bullpen of specific pieces, plus the Red Sox and their collection of same-position hitters.

Week 1 previews: Giants | Royals | Dodgers | Rays | Padres | Astros | Rockies | Athletics | Mets | Yankees

Week 2 previews: Nationals | Tigers | Pirates | Mariners | Brewers | Indians | Marlins | Orioles | Diamondbacks | Twins

Week 3 previews: Phillies | Blue Jays | Braves | Rangers | Reds | White Sox

ANAHEIM ANGELS
Team Audit | Player Cards | Depth Chart

PECOTA Team Projections
Record: 91-71
Runs Scored: 742
Runs Allowed: 646
AVG/OBP/SLG (TAv): .258/.324/.409 (.278)
Total WARP: 36.8 (10.1 pitching, 26.7 non-pitching, including 0.0 from pitchers)

The premise of this series is that every team at least has a plan. Some of the plans might be hard to quantify, or hard to prove “work,” but every fan deserves to know that his or her favorite team’s front office is intentional about its decisions and is smarter than the aimless pack of idiots we might, in our darker moments, imagine them to be. When it comes to the least predictable part of team building, the Angels have a plan, and of this I can be certaineven if, ultimately, I’m at a loss to figure out how well it works.

Imagine that, back in autumn 2011, Jerry Dipoto had been forced to publicly campaign for the job of Angels General Manager. Based on what he has said and what he has done, my guess is his plank would have sounded something like this:

  • Get some talent in the upper levels of the minor leagues quickly, after a few years of barren high-upside drafts and international inactivity had left the Angels extremely low on subterranean depth; do this by drafting almost exclusively college players in the first 10 or 15 rounds.
  • Change the Angels’ offensive approach from slap-hacky to, if not exactly TTO, at least more balanced, with lineups capable of working counts and exploiting bad-control pitchers on bad-control days. Do this by acquiring players who, in his own words, “control the strike zone”; divesting themselves of Jeff Mathis; and updating player development strategies to reinforce the value of long at-bats and plate discipline.
  • Fix the bullpen, which had become swollen with bad relievers on multi-year deals.

There would have been other things, some too small to make it into a campaign ad, some too abstract for a politician to boil down into a soundbite, but those three would have covered everybody, from Lou in Gardena who melts down on the air every time a high-priced reliever melts down in the eighth inning, to the stathead bloggy types eager to hear that plate-discipline dog whistle. In all three areas there has been activity, sometimes successful, sometimes thwarted by lack of options or the inevitable restrictions imposed on a smart team competing against 29 other smart teams for limited resources. The bullpen, though, has seen both the lowest lows and the highest highs.

In 2012, Dipoto’s first year, the Angels’ bullpen ranked 13th out of 14 AL teams in FIP, 12th in ERA, 11th in FRA. In 2013, it ranked 13th out of 15 AL teams in FIP, 14th in ERA, and 10th in FRA. And, through June 30th of 2014, it ranked 11th in the AL in ERAfriendly home ballpark and all. Dipoto has spoken about his unwillingness to chase high-priced closers, opting instead for discounted reclamations (like Ryan Madson) or two-year commitments (as to Sean Burnett) or pre-arb trade targets (like Ernesto Frieri and Dane De La Rosa). Three of the four broke; that last, Frieri, simply fell apart.

On June 27th, though, Dipoto traded Frieri for Jason Grilli, and in the next couple months would add Huston Street, Vinnie Pestano, and Joe Thatcher in trades. They, and the incumbents, became one of the game’s best units, finishing the 2014 season with the seventh-best bullpen ERA in the AL, the third-best FIP, and the second-best FRA. The relievers were so good that, when the Angels’ ran out of healthy starters in late August, they turned to the bullpen to fill the fifth slot in the rotation. Cory Rasmus took the mound in the first inning against the A’s on August 30th and, in his first ever major-league start (and first start at any level since 2011), went three innings. Seven relievers combined to complete the game, and the Angels won 2-0. Rasmus would go on to “start” five more games in September.

Wrote Pedro Moura that month, “(Scioscia) has mostly deflected praise… When asked earlier this week about how so many opinions of him have changed over the last year, he had a quip at the ready. ‘Managers tend to get a lot smarter when you have a bullpen,’ he said.”

There are a lot of ways to build a good bullpen, many of them smokescreens for “got lucky when we signed him” and “boy sure got lucky when we signed him.” Further, every GM who has ever GM’d has had at least one good half from his relievers, but most don’t merit a season preview touting his GMing skills. But, while acknowledging that the Angels’ second half is a flimsy foundation on which to build a statue, we argue that Dipoto and the Angels merit the full write-up more than most. That’s because Dipoto has a bullpen philosophy that is different than most; he has clearly articulated that philosophy over the past three seasons; the unit he succeeded with arguably abided by that philosophy; and, hey, the dude was a former reliever, the only former reliever GMing today, and his reliever experience is part of what informs his philosophy. You get extra credit for convenient backstory.

Here’s the group that the Angels took into the postseason:

The first thing you might notice: No lefty. He tried to get a funky lefty and the funky lefty just flat out flopped. The second thing you might notice: Five of the eight (and the deposed funky lefty, too) were acquired via trade. By my count, Dipoto has acquired 14 relievers in trades, which seems to be his preferred way to avoid the long-term contracts he has called “a little bit of a game of Russian Roulette.” He gets relievers as throw-ins, he gets relievers back for his salary dumps, he gets relievers for prospects, he gets relievers back when he trades away relievers. He gets a lot of relievers. There’s still a cost to acquiring some of these guys, to be sure, but only Smith (the free agent) carried a long-term commitment out of this group. (Further, if less importantly, acquiring guys in trade rather than free agency has helped him stock his bullpen with players who still have options. Vinnie Pestano was 29 but still had an option; same with Fernando Salas. Each was sent down when the Angels needed roster flexibility.)

The third thing you might notice: This isn’t a hard-throwing group. In 2013, Angels relievers threw an average fastball 93.3 mph, which was in the top third among all teams; and they threw that fastball 72 percent of the time. But in 2014, the group’s average fastball was 91.9 mph, bottom third in the league; and they threw that pitch only 60 percent of the time. This might not seem greatthrowing harder usually makes pitchers betterbut there’s essentially no correlation between bullpen velocities and bullpen ERAs, or velocities and a components-based stat like SIERA. “We used to be a bullpen built on power arms and sometimes some high wire acts,” Dipoto told Peter Gammons. “A lot of 96’s with limited command.” Getting away from that was part of the unwinding.

What he was getting toward is the hardest thing to spot in this group: The “different looks” theory. He first mentioned it in April 2012, the first month of his first season:

In the meantime, he is gently steering the Angels… toward a collection of arms with what he calls “different looks” — different arm angles, different out pitches, different velocities. …

Later in the same article, he said: "You want to try to create as much diversity as you can.”

You can imagine how this would be useful if implemented right. You could more nimbly match pitchers up against hitters and situations they're suited toward—the high-velo guy against the slow bat, the funkiest delivery against a guy who, I don't know, struggles against funky deliveries, the sidearmer in groundball situations, the permanent slide-stepper with a runner on base, and so on. This would be complicated, but with enough intelligence and information it could work. It'd be like a modified version of using a lefty against a left-handed batter, putting the pitcher in a position to succeed, and ultimately helping his stuff (and that of the rest of the staff) play up. Implemented effectively, it would help break the tyranny of the assigned innings—the seventh-inning guy, the eighth-inning guy, maybe in a radical envisioning even the closer guy—and save the team the expense of purchasing a pre-cooked eighth-inning guy who wants three years and $18 million to face three batters at the same time every night. That's ultimately the goal: To quit having to shop on the free-agent market for an eighth-inning guy. Much better to have three inexpensive options who, deployed correctly, complementing each other's strengths and weaknesses perfectly, can mimic the effect in the eighth (or any other) inning. "The important thing is not to recreate the individual," as Billy Beane says in Moneyball. "The important thing is to recreate the aggregate."

One way to test whether he did this last year is release points: The Angels’ right-handed pitchers do have, among all teams’ relievers, the third-most variation in vertical release points. Another is general descriptions: Joe Smith is obvious, as only one reliever in baseball (Brad Ziegler) has a lower release point that he does. But what about the rest? How would we describe them?

Nobody ever called Rasmus deceptive. Which is the ultimate deception. (He’s a reliever who throws four pitches with near-equal frequency, for what that’s worth.) All of these guys do something different than the others, and in a subjective sense many seem to do something different than a typical right-handed reliever.

Now, I’m going to walk this way back. Doug Thorburn, asked by me to assess all the Angels’ primary relievers, contends that “overall, the Halo bullpen does not strike me as especially varied in terms of ‘look,’ whether talking about delivery or repertoire.” The 2013 unit, he says, despite being much less effective, actually has more diversity: “Individually they're not especially interesting, but as a group they do offer some ‘different looks.’” If there’s a theme to the bullpens, it’s “the sheer number of RP's they have had (since 2013) who invoke the ‘hunch’ into or out of max leg lift. That shouldn't affect things in terms of the hitter's POV, but I could see how that might be interpreted as ‘different looks.’” I don’t argue with Doug on things like this.

The truth is, measuring the value of sequencing relievers would be nearly as hard as measuring the value of sequenced pitches. You’d have to know which characteristics to build your taxonomy around, you’d have to figure out the ways to measure those characteristics, you’d have to know how those characteristics influence the matchup, how those characteristics’ value is affected by other pitchers having (or not having) the same characteristics, how batters in general can or can’t adapt to these characteristics, and how specific batters individually are prone or not prone to these characteristics. Then you have to figure out whether there’s even an implementation of these varied characteristics that outweighs the always reliable “get good pitchers, then use the best ones as often as possible and in the most important moments possible” way of building and deploying a bullpen.

So I’m not going to go so far as to say this plan worked. I’m going to go so far as to say that there’s a plan. If there really were public campaigns for GM positions, I’d expect an awful lot of candidates to promise to fix the bullpen. I don’t think more than one in each race would have a policy paper and a stump speech on how exactly he’d do it. Dipoto does, and for that I’d vote for him.

Appendix

Because a Doug Thorburn assessment is too good to waste, here’s how he described the eight regulars in the Angels’ postseason bullpen:

  • Street has the electric slide to start his windup, but otherwise is pretty pedestrian with his "look" in terms of arm slot and stuff – the only weird thing is that he doesn't throw very hard for a RP (avg FB 90.2 mph).
  • Salas throws a knucklecurve, so that's fun, with velo that averages 91.8 mph on the FB. He has a late posture-change to his delivery, manipulating a higher slot, and yet his average release point has basically the exact same vertical height as the well-balanced Street.
  • Jepsen is now a member of the Rays, so I don't know if he fits. He was the hardest thrower of the group (avg FB 96.3 mph), though his "look" was not especially interesting.
  • Smith is "different" for obvious reasons of submarine goodness, but he actually stands out among the sub cohort as well. Most submarine guys hunch over during the stride (as they drop down) and then pop up to have solid posture into release point; Smith, doesn't really pop up, instead he just keeps on drifting with spine-tilt toward the 3B/arm side at release. His release-point window is ridiculous.
  • Grilli is now a Brave, so same boat as Jepsen. Grilli's basically a 2-pitch guy (FB/SL), with a very quick yet boring delivery. He pitches from the stretch at all times, and he might "look" like he's different because of his quickness (~0.85 seconds from first movement to release point).
  • Pestano has the hunch-over of a submarine guy but pops up too early to earn the moniker. Nonetheless, he finishes with strong posture and a very low release point (~4.6-5.0 feet vert height, depending on season, Z-score of -2.0 or less), so he does provide a different look that falls short of the Smith extreme but is far from "normal." He's another 2-pitch guy (80%/20% on FB/CB last year), and his velo has been declining.
  • Rasmus actually has some of that submarine hunch in the early-going, but it's short-lived and he starts tilting to the glove/1B side on top of foot strike. It results in a ~10:30-11:00 arm slot and a vert height of 5.6-5.8 feet (for a guy who's listed at 6'0"). He finishes by falling off to the 1B side, and the side-to-side of his balance might be construed as a "different look."
  • Morin has some late tilt and his delivery kinda explodes after foot strike (from slow momentum to big torque), but it's far from unique.

***

The following articles were essential: