Globe iconLogin iconRecap iconSearch iconTickets icon

MLB News

How advanced metrics influence the Draft

Sabermetrics helping clubs understand how talent will translate to pro level
June 8, 2016

Kyle Lewis ranked in the top 15 nationally in batting average (.395), home runs (20), RBIs (72), on-base percentage (.535), walks (66) and runs scored (70). He was just named Baseball America's Player of the Year. He will be a top pick in Thursday's first round of the 2016 MLB

Kyle Lewis ranked in the top 15 nationally in batting average (.395), home runs (20), RBIs (72), on-base percentage (.535), walks (66) and runs scored (70). He was just named Baseball America's Player of the Year. He will be a top pick in Thursday's first round of the 2016 MLB Draft, if not the top pick.
But Lewis, who MLB Pipeline ranks as the No. 3 Draft prospect, compiled those crazy stats at Mercer, in the Southern Conference and not at, say, Vanderbilt, in the SEC. And so teams evaluating Lewis' worth face a question that has dogged scouts for decades:
How does this player's amateur performance translate to the bigger stage?
Whether the question is applied to Lewis or any among the multitudes of players eligible for the Draft, which begins on Thursday night at 6 p.m. ET with live coverage on MLB Network and, the answer is and always will be inexact. But like every other facet of this industry, advanced analytics have infiltrated and greatly influenced the way teams prepare for the event.
Complete 2016 Draft coverage
"It's getting toward where the Major League levels are," an NL executive said. "It's the same in the sense that you try to get as much information from all angles and combine the statistical side with the scouting reports to get a full picture of the player."
Obtaining the statistical side is a potentially costly process. It's not like MLB is supplying the same Statcast™ data for prep and college games that it does in the big leagues. And it's not like logging onto the NCAA web site and sorting stats by batting average or RBI is all that helpful, either.
So teams have to find usable information however they can, often by enlisting the help of third parties.
At the college level specifically, the technological service TrackMan, which is a component of Statcast™ and uses Doppler radar to provide such useful data as exit velocities and spin rates, is available for hire. The data compilation service College Splits, which was once a public site and now mostly services clubs, can be used to suss out the worthwhile stats.
All of this information can be combined with the good old-fashioned scouts' eye to form a club's ultimate evaluation.
"Every team looks at player performances as part of their process," said Christopher D. Long, a sabermetrics analyst and consultant for the Tigers. "There's no doubt about that. It's more a question of how that's evolved and how sophisticated it's gotten. Certainly the better teams have an analyst who, this time of year, is dedicated to doing Draft analytics."

Any claim that statistical information is an affront to scouts is missing the point. Whether they realize it or not, scouts have long been in the data compilation business. Be it a player's speed down the line or a catcher's pop time, any data point can be helpful in the decision-making process.
What's changed in baseball is the way this data is compiled, the augmentation that arrives in the form of more modern metrics and the way all of this information is combined with other factors to make final determinations. The Cardinals are a great example of a club that has gotten great value out of the Draft -- never more than in 2009, when Matt Carpenter (13th round), Trevor Rosenthal (21st) and Matt Adams (23rd) were all taken in late rounds from the college ranks -- because of the blend of stats and scouts. To hammer home that point, the name of the Cards' Draft prep database -- STOUT -- marries those two words.
"Even at the big league level, you're not going to make a big trade just based on the PITCHf/x data," an NL executive said. "So at any level, the numbers you're looking at are just part of the equation. You look at the stats, the scouting reports, the medical history and the makeup."
The core difficulty of Draft analytics comes with taking the numbers and determining how much they are affected by environment.
One notable (and somewhat extreme example) example is Nationals first baseman Ryan Zimmerman, who played his college home games at Virgina's Davenport Field, the college equivalent of Petco Park, in terms of its power suppression. In the entirety of his college career, Zimmerman hit just seven home runs. In his first 63 professional games, he hit 11 and now has more than 200 in the Majors.
"It's hard to know how each park affects each kind of hitter," said Tom Tango, a renowned analyst who has consulted for the Mariners, Cubs and Blue Jays and now works for "That's a huge challenge just on its own. The thing to add on top of that is the opponent for whatever player you're looking at. There's a wide gap of talent there as well. So you've got two huge things conspiring against you, just trying to figure that out."

People are trying to figure that out.
At College Splits, for instance, efforts are made to account for not just park and schedule factors, but even variations within specific rosters. If a team wanted to know how A.J. Puk, the Florida left-hander who many are projecting to go No. 1 overall, has specifically fared against batters who played in the Cape Cod League -- i.e., players on more of a professional track -- that information is available.
This is how those eye-catching surface-level stats, such as the ones Lewis has accumulated in his transcendent collegiate career, can be given proper context.
"On the one extreme, there's looking at a guy's numbers at Mercer and a guy's numbers at Vanderbilt and just not recognizing the difference that's there," said a person familiar with the amateur analysis process. "The other extreme is just discounting the Mercer numbers entirely and saying there's no valuable information there.
"The truth is somewhere in between. And you don't have to guess anymore. You can account for the level of competition, park factors and all that stuff. That work can and is being done. Your confidence and belief in it is up to your judgment, but those questions can be answered."
Lewis certainly helped assuage some of the fears about his college competition when he hit .300 with seven homers (good for fourth in overall) in the Cape Cod League last summer. Not only does the Cape feature many of the game's top college prospects, but it also uses wood bats, which helps scouts get a better sense of how a hitter's skills will translate in pro ball.
You might have noticed this conversation has centered around the college level. The truth is that at the prep level, the statistics are even more unreliable. And while nobody wants to the be the scouting director who passes on the next Mike Trout, showcases and college commitments are already considered quality guideposts in finding Draft-worthy talent.
But Long, for one, thinks it might be worth the while (and the cost) to do more.
"There are some really strong high school programs," Long said. "If you're a smart MLB team, why not install a TrackMan unit there? If you think there's a small chance it would have long-term value, the math works out."
It could be the cost of ensuring a quality player doesn't fall through the cracks.
"Analytically, the Draft is the hardest part of baseball," Long said. "But it's far and away the most important. If you draft Mike Trout, that's such a huge, huge edge for years. Or if you draft poorly in the first and second rounds for a few years, there's almost no way you can come back from that."
That's why the good teams cut no corners -- and consider all numbers -- in their prep for the Draft.

Anthony Castrovince is a reporter for Read his columns and follow him on Twitter at @Castrovince.