MLB teams
Sam Miller, ESPN.com 10y

The guide to sustainable farming

MLB, Los Angeles Angels, Texas Rangers, St. Louis Cardinals, Kansas City Royals, Cleveland Indians, Pittsburgh Pirates, New York Yankees, Tampa Bay Rays

ONE YEAR AGO, on backfield No. 6 in Tempe, Ariz., where players breathe the helium that floats spring training optimism, 165 Angels minor leaguers took a knee and 55 Angels minor league coaches stood at attention. Scott Servais, the club's assistant GM overseeing player development, was in the middle. It would be his only chance all year to address everybody at once.

"How many of you here played in the playoffs last year?" he asked. Denny Hocking, the Angels' new rookie ball manager, hired from Baltimore's Double-A affiliate, was the only one who could raise his hand.

"Baseball America thinks we have the worst talent in the game," he continued. "How does that make you feel? You play in the worst organization in the game."

Third baseman Kaleb Cowart, LA's top prospect at the time, recalls looking at the faces of his teammates as Servais spoke: "They're like, Wow. Was he really just that open about it?"

But it was true. The critics had spoken, have spoken, continue to speak: Even now, a year later, Baseball America considers the Angels' future to be windswept and barren, again ranking them 30th in organizational talent.

Servais is the guy who was hired in 2011 to fix them. It's a job he's done before. As the Rangers' director of player development, he took an organization that had baseball's 28th-best farm system and in two years turned it into the best. But if that was a nifty trick -- like juggling eight balls and a chain saw -- turning around the Angels' system is like repeating it, blindfolded and on roller skates. The best routes for adding talent aren't available to him: The Angels forfeited three top draft picks to sign Albert Pujols, Josh Hamilton and C.J. Wilson, and the loose rules that had allowed the Rangers to sign half of Latin America and hoard draft picks have since been tightened.

Since the Angels have few avenues to acquire impact talent, Servais is left to develop it. So he recently burrowed into his home on 35 acres in Colorado to compile a guide to player development, a manual to be read by everybody from ownership on down. It would lay out the Angels' organizational philosophies and offer step-by-step development strategies for each skill a player must master. It would help modernize the Angels' player development department; it would make the team more like the one every organization in baseball wants to be -- the Cardinals.

St. Louis reached last season's World Series with a roster that featured 17 players the club had drafted and developed. None of the three other teams in the LCS had more than six drafted and developed players. The Cardinals are, as Servais puts it, "the hot team in player development." And three years ago, they gathered all their organizational intelligence into a manual: the Cardinal Way, a proprietary 117-page document. As Cardinals Double-A manager Mike Shildt told a newspaper reporter at the time, "This is the blueprint."

When I meet Servais this winter in a Denver hotel lobby, he says, "They've got it going on."

What do the Cardinals do differently?

"Well ..." He smiles. "Let me see what I've got here." He reaches into his attaché, pulls out a document -- 117 pages. It is the Cardinal Way.

"They don't know I have that," he says. I open it. A warm, golden glow envelops me.

This is really it?

"That's the secret," he says.


TWO THINGS MAKE predicting a prospect's future confounding. One is that everything beyond his next breath is exponentially unpredictable. A major leaguer might surprise you; an A-baller might conceivably do anything. One year, baseball's third-best overall prospect turns into the superstar you know as Mike Trout. Another year, he turns into the cautionary tale you know as Brandon Wood.

To understand the state of the Angels -- bloated budget, barren farm, four-year post-season drought -- you might start with Wood. "It haunts me," says Abe Flores, the Angels' former farm director who oversaw the development of Wood (and Trout). "Haunts us. We were counting on him. He was a guy we absolutely needed, and then it snowballs. That's one I've struggled with for a long time."

In 2005 Wood had one of the greatest minor league seasons ever, compiling 101 extra-base hits while playing shortstop. He went to the Arizona Fall League and hit a league-record 14 homers. That kind of production, of course, is worth much more than the runs scored. Writing for The Hardball Times in 2008, Victor Wang, now an assistant scouting director with the Indians, looked at all the players Baseball America had included in its annual top-prospects list during the 1990s. He found that elite hitting prospects were the minor leagues' safest bets: A hitter who ranked in the top 10 -- as Wood did -- produced, on average, about 11 wins more in his first six years in the majors than a typical replacement from the bench would. The website Pirates Prospects updated the research in 2012, using more recent Baseball America rankings, and the expected value of a top hitting prospect had gone up to nearly 18 Wins Above Replacement. At the going rate for free agents, 18 wins are worth about $125 million.

But those 18 wins are an average, which leads to the second frustrating thing about projecting a prospect's development. We plan for the future by averaging our expectations-30 percent chance of rain, 20 percent chance of a spade on the river, 50 percent chance Wood is worth 11 WAR -- but we play it out only once. You never catch 20 percent of a spade. So Wood was never going to be worth an average WAR. He was going to be worth one thing, and that one thing could conceivably be anything.

As it turned out, Wood started 2010 as the Angels' third baseman. He was arguably the worst player in franchise history. No Angels hitter with 200 career plate appearances has ever produced a lower OPS than Wood's .455.

The exponential unpredictability of the future doesn't stop there. Partly because of Wood's sub-replacement performance, the Angels had a losing record and missed the playoffs in 2010. Ditching Wood, and desperate to make a "big splash" that winter, the Angels traded Mike Napoli for Vernon Wells, who was booed during the first homestand and never recovered. The Angels again missed the playoffs in 2011, again spent the offseason seemingly on tilt, and committed more than $300 million to Pujols and Wilson. Each veteran underperformed in 2012, when the Angels missed the playoffs a third time. That offseason the team signed Hamilton for $125 million -- and missed the playoffs again when the outfielder flopped.

To encapsulate the snowball effect Abe Flores witnessed: The Angels will pay Wells, whom the team has long since discarded, $18.6 million this year.


IN 2011 THE Angels fired or demoted nearly everyone in the front office, including Flores. When Jerry Dipoto took over as GM, he said, "The analytics department is [now] baseball ops." Sabermetrics, in other words, would be part of everything the front office did -- including player development.

"We can beat people in player development," Servais says. "Scouting -- they all look at the same players. It's really hard to get a leg up. You have a much bigger chance to have an impact in player development if you're willing to try new things. Why would we want to do what everybody has always done? We'll get the same results. I want different results."

Yet more than any other part of the game, player development resists analytics. It's not that coaches dismiss numbers; most will use them if they're useful, available and well-explained. (In conversation, the Angels like to drop wOBA, a hitting stat that is notable not just for its complexity but because it has such a funny name.) Rather, it's because data-driven analysis requires actual data, and there's no practical way to quantify player development's performance. "We would always talk about ranking the best farm directors," says Jim Callis, former executive editor at Baseball America and current prospect expert for MLB.com. "But how do you quantify it? It's the chicken and the egg -- how much of it is scouting, and how much of it is player development?" Does a player flop because he was mishandled, or had scouts simply been wrong about him in the first place?

Even if we could confidently credit Flores with everything Trout became, or Wood didn't, a larger problem becomes clear: What might have worked with Trout might have worked only with Trout. Most of the lessons of the sabermetric revolution are based on what's called large-N analysis: looking at all the players who ever played and finding, in millions of data points, answers about player tendencies and optimal strategy, and meta-answers about the reliability of statistics. But developing a prospect is an N=1 problem: Each player's combination of skills, genes, experience, health, neurology, psychology, size and style makes him unlike any other player. Studying one player might tell you nothing about another -- might mislead you, even. And a breakthrough with one player might be irrelevant to another's potential.

"I'm often asked what I think 'the next big thing' in sabermetrics will be," former Indians consultant Russell Carleton wrote for Baseball Prospectus this year. "The next big thing is understanding how each player works and, instead of strategies that can be applied across entire organizations, understanding the nuances of each player on a case-by-case basis."

Take a prospect like Alex Yarbrough. His story resembles those of many American-born big leaguers: youth ball at 6; switch-hitting by 9; on the showcase circuit as a teenager; known to online prospect hounds by 15; always faster than the game around him; star shortstop in high school; star second baseman in college; drafted by the Angels in 2012, fourth round; healthy bonus ($300K); funny kid, competitive, tweets a lot, has a crush on Miley Cyrus. He's like everybody else, except that Yarbrough has always done one thing differently: He takes the ball out of his glove wrong.

After most people catch a ball, they reach into the glove with the throwing hand palm down. Yarbrough reaches in with his palm up. To go from there to a throwing motion, his arm has to come out wide, a long arm stroke far from the torso. So it takes Yarbrough longer to throw, especially on what Servais calls "the biggest play in the game, the momentum changer" -- the double play.

It's a small detail, but second basemen who don't have the arm to fill in at short or third, the athleticism to move to the outfield or the power to force promotions have to be perfect. "More often than not, the margins are that small," Yarbrough admits. This small detail could be the difference between 10 years in the Show and private school for the grandchildren, and a career that tops out at Triple-A.

So what's a stathead to do about that?


SERVAIS HAS SPENT a lot of time contemplating the N=1 problem and thinks maybe the answer isn't having an analyst codify millions more data points on a player's idiosyncrasies. Maybe it's something else, something more basic -- something found in the Cardinal Way.

"That's the secret," Servais says again as I flip through the manual in the Denver hotel lobby.

Command the fastball, the document reads.

Uhh ...

Pitch ahead in the count.

Sort of obvious ...

Develop a breaking ball.

Oh come on. The golden glow fades.

"There are no secrets," Servais says. "Read through that. It's as simple as you can get."

Exactly one sentence in the book has been highlighted: "Limiting OBS will be one of the main considerations in evaluation of minor league pitching."

It's entirely anticlimactic, but what I come to realize is that the manual is a big part of how the Cardinals' PD department works -- by making every instruction in it rote, unexceptional, harmless background noise. How a coach teaches pitchers to back up a base isn't, ultimately, all that important. What's important is that no coach has to spend more than two minutes of his life thinking about it. That frees him to focus on the N=1 problems, to make 165 little breakthroughs, one for each player, every year -- then on to the next 165.

"If you get everybody on board," Servais says, "you've got a chance."

That's where the new Angels player development model comes in. Although Servais hasn't completed his own manual just yet, the organization has, broadly speaking, rallied around three central tenets.

1. Spend, Big and Small. Moneyball, at its heart, was really just about identifying places where a dollar spent can bring back $1.10 in value. And the minors provide countless places where small investments might nudge players toward their best selves, or at least remove the obstacles in their way.

The Angels, upon Servais' arrival, were spending around $9 million a year on player development, meager compared with most teams. Their Dominican facility, with no air conditioning and terrible food, was dilapidated; MLB threatened to shut it down. Their use of technology was 10 to 15 years behind the rest of the league, according to their own estimates.

Since last year, the Angels have tripled their spending on Dominican facilities and bumped overall spending to $12 million annually. Smaller expenses are no less important to Servais: a nutritionist, hot breakfasts at spring training, a supplements budget, cellphones for coaches, travel budgets for coaches' families, bedspreads with the Angels logo for the Dominican kids. Bedspreads probably won't win the Angels their next title, but they might just make players a little more comfortable -- and they cost less than Josh Hamilton earns taking batting practice.

2. Try Everything. The Angels' PD staff -- half of it hired or in new positions since Servais took over -- has become used to seeing its unconventional ideas tested. In the Dominican Republic, the Angels started measuring time of possession. (The quality of pitches increased significantly the quicker the pitcher worked.) They're filming instructional leaguers from ladders so players can see their footwork on defense. They're incorporating elements of advance scouting in rookie ball: Idle pitchers will sit in the stands, chart the way opposing pitchers attack hitters, then run back to post the results in the dugout. They measure new stats -- like how often batters fight back from a pitcher's count -- and share them with the players.

The Angels expect to finally install TrackMan video technology, at a cost of $500,000, at one of their minor league parks this season, giving them access to troves of information on pitcher spin, hit trajectories and other granular data -- and shrinking the large data gap between them and other forward-thinking organizations. A team can have one of three relationships with advanced metrics. Tier one: Ignore them. Tier two: Embrace and employ what's already out there. Tier three: Create its own intelligence.

"We want to get to three," Servais says.

3. Get 55 Coaches to Feel, and Act, as if They're on the Same Team. This might be the most important tenet. If the High-A hitting coach spends a year on a player's swing but passes the baton to a Double-A hitting coach who undoes it, the player loses faith in the system and goes back to the hitting guru he's known since sixth grade.

That's where Servais' new manual will come in. But it's also where the cellphones and family travel budgets play a role -- it's why Servais once threatened to quit if his coaching staff at Texas wasn't paid more. And it's where listening to everybody, and giving each idea an honest trial, comes in. If the 55 coaches are one unit, they're sharing information -- they know Yarbrough has size 12 1/2 feet, limiting how graceful he can be. They know how he'll respond to criticism, how diligently he'll stick to endlessly repetitive drills. He, in turn, will trust them.

And he has. When Servais raised the throwing issue in the summer of 2012, Yarbrough agreed to make it his primary focus. When the coaches broke down his fielding mechanics, he went along with the inquiry. When they made him throw countless balls from his knees to undo sloppy footwork, he did it; when those throws became a regimen of six or seven drills, some invented specifically for Yarbrough, he followed them hour after hour in spring training. Then he kept working with High-A shortstop Eric Stamets before games all season. He didn't just do the monotonous drills. He felt good doing the monotonous drills. And he got better.

"He's not Robbie Alomar, cool and smooth," Servais says, "but Alex Yarbrough is going to play in the big leagues."


ABE FLORES MIGHT remain haunted by Brandon Wood, but few besides Flores think it would have mattered which organization the shortstop was in; he was destined to fail. His swing was the problem -- he had a big loopy hitch that made it difficult for him to catch up to hard stuff up or in -- and swing mechanics are either totally fixable or impossible to change. "It wasn't ever like this guy's a slam-dunk guy," Flores says. "It was always in the room, always that concern. That high-punchout/low-walk rate -- you know that's not how it works up in the big leagues. Those ratios aren't going to work. They bite you."

But if some players like Wood aren't fixable, will baseball at least be able to identify their failings better when it's smarter about everything in 10 years? "My gut feeling is we'd be more skeptical about him faster," says MLB.com's Jim Callis. "I just think there's a limit to what you can fix. My gut says it's innate talent more than it is being able to develop something."

That's what Servais is testing now, and there are signs that the systemic approach is working: The Angels' system improved from 22nd in on-base percentage in 2012 to second in 2013 (according to their internal stat reports), and their 2013 draft class hit more homers last summer than any other club's, even though the Angels didn't draft a hitter until the ninth round. Perhaps the best sign of all: His coaches told Servais that the Rays -- "an organization that actually does the studies," he says-were snooping around, trying to find out what LA was doing differently.

There's an old joke that farm directors are like parents -- they all love their kids more than anybody else loves them. The system still lacks high draft picks, still suffers from underinvestment in Latin America, still requires three, five, maybe seven years to assess. What Servais knows is this: At the end of the 2013 season, he went from affiliate to affiliate, watching his teams play in their league playoffs. Four of six Angels affiliates made the postseason, and a fifth just missed. High-A championships aren't worth much in this sport. Still, in every city Servais visited, players flagged him down, dozens in all. "Hey," they'd call, "remember that day in spring training?" The day he asked how many of them had played in the playoffs the year before? The day he told them how bad the world thought they were?

"It'd be kind of funny if he does that again," says Kaleb Cowart, now the team's No. 3 prospect. "It'd be a lot of hands come up."

Follow The Mag on Twitter (@ESPNmag) and like us on Facebook.

^ Back to Top ^