Technology Threatens Scouts, Could Also Set Them Free

After the Houston Astros elected to part ways with eight scouts last month, I wrote a piece on whether scouts and Statcast could coexist following the move by Jeffrey Luhnow. (Luhnow said the vacancies would ultimately be filled.)

Carson Cistulli was also interested in this idea of redundancy, asking both Dave Cameron and Eric Longenhagen about it on different episodes of FanGraphs Audio. Dave was optimistic about the future role of scouts, while Longenhagen reported on some of the anxiety in the scouting community — anxiety with which I’m also familiar to some degree.

Ken Rosenthal then wrote an excellent deep dive for The Athletic in which he spoke to a number of executives about the subject.

From Ken’s piece:

The Brewers, under GM David Stearns – Luhnow’s former assistant with the Astros – employ about 12 pro scouts and do not cover major league games until September, sources said, a rarity. The team’s pro scouts work in the minors until the final month and also help out on international and amateur coverage.

The Brewers, like the Astros, believe strongly in video scouting, sources said. They also believe that information on a player’s makeup – information many teams consider vital – is more valuable at the lower levels of the minors, where performance data is limited. To others, however, the video and data offer only so much.


The Diamondbacks recently let go of two full-time and two part-time pro scouts, but will hire a new pro scouting director and increase the size of their staff, according to general manager Mike Hazen.

“I look at pro scouting as one of our most important departments, if not our most important,” Hazen said. “When you get into making trades, the pro scouting department is your most important department. They are the guys who are out there seeing these players, gathering every bit of critical information that you can’t get anywhere else. It’s the contacts, and the relationships and the critical makeup information that otherwise you don’t have access to.”

How scouts and technology are integrated — whether by way of a peaceful, complementary process or one subject to conflict and infighting — has the potential to emerge as a significant story over the course of the next decade. I belong to the camp that believes there’s a place for both, that a $10 billion industry should have space for robust scouting departments. But I’m also not a business owner. I’m a journalist/blogger susceptible to romantic ideas.

I asked Texas Rangers general manager Jon Daniels in July if he anticipated conflict or collaboration in the near future. He saw more of the latter. Daniels noted that technology should free scouts, allowing them to gather more intelligence on makeup, the human element, etc. But I wanted to follow up with a more specific sense of how technology might complement the game’s talent evaluators.

More recently, I asked Pirates general manager Neal Huntington — who worked as an advance scout before becoming one of baseball’s more forward-thinking general managers — about how scouts and Statcast might get along. Huntington suggested that Statcast would likely free scouts to better fulfill the main duty of the profession for as long it’s existed: to watch games.

This seems like the most basic of job requirements, but it can actually be a problem.

As Dave noted on the podcast, scouts have essentially been glorified data-entry personnel for years, recording radar-gun readings, pop times, run times, etc. The lack of technology has too often forced scouts to divert their eyes from the field.

It’s a problem with which I, as a former beat reporter, can relate.

There’s a similar, but reverse, problem in press boxes across the country. The need both to interact on social media and also to post game stories immediately after the final pitch has often taken reporters’ eyes off the field — perhaps too often. My experience in the press box has suggested that, too often, reporters are compelled to look more at their laptops and phones rather than toward the action. I was guilty of this myself while working for the Pittsburgh Tribune-Review. My friend and fellow baseball scribe Dave Manel of Bucs Dugout would occasionally joke (or, half-joke) that we were at a baseball game, but not really watching it.

Reporters have become prisoners to technology. Technology might also threaten scouts. That said, it also offers freedom to better do their jobs.

“Instead of having to write, our guys are able to watch,” Huntington said regarding the influence of player-tracking systems on the organization’s scouts. “If you are charting the game, your head is going down.”

That seems like the most basic of problems, but it has existed for decades.

Huntington recalled how, as an advance scout in Cleveland, he would debate about whether to chart the game in person during the live event or wait and go back to his hotel room and use the nascent video to replay the game and chart afterward.

“I served as an advance scout and charted every pitch,” Huntington said. “There were times when I had to fight ‘Did I want to wait and chart the game after and just watch the game?’”

It’s what scouts miss, it’s what reporters miss when their heads are down, when distracted, that has come to be so important: the details. And what can be inferred from the details? To what question will the details lead?

Huntington cited PITCHf/x as proof scouts and player-tacking systems can coexist in harmony.

“When we got the pitch-tracking data… it didn’t eliminate scouts. It freed scouts form having to chart pitches and grind through the data to find trends,” Huntington said. “Now much more is automated. But a scout that sees the pitcher’s tells, a scout that recognizes patterns — patterns that a computer won’t even recognize — there’s still incredible value in that scout that is in the stands… They’re able to watch the pitcher, his mannerisms, his action, his body language. The small things, instead of writing down, ‘It was a fastball down and away and hit to second base.’

“From an advance standpoint, it’s been an incredible advantage… Instead of worrying about the [stop]watch and getting that contact-to-contact, we are able to watch the game. We are able to watch the second baseman not move on a play where he should have moved. That catches our attention. Where are his defensive instincts? Where is his energy? Where is his love and drive for the game? Or does he just like to hit? There will be some other things, perhaps some better things, for our guys to focus on: the intangibles. They’re not [engaged in] glorified data entry.”

For the Pirates’ scouts, the radar gun is still a tool in the bag, but now it’s something of a tool to be used to perhaps verify readings, not as a crutch. Statcast offers the promise of putting down the stopwatch and eliminating play-by-play cataloging.

“They don’t have to put spray charts down by hand,” Huntington said. “It will free our scouts to maybe not even get run times because they will be able to watch the gait, they will be able to watch the defenders and their reactions. They will get a better feel for defensive instincts. It’s one thing to say, ‘The route was efficient, the jump was X, the max speed was Y.’ It’s another thing to understand defensive instincts. How engaged he away from the ball? Is he a spectator or a participant?

“They are able to just watch.”

That might seem like a simple thing, just to watch, but for so long scout’s haven’t been able to give their full attention to the game. Statcast can allow them to travel to the ballpark with a full focus on the game. It is perhaps the most powerful evidence suggesting that the technology and the scout can peacefully coexist.

Technology is perhaps not a threat to scouts, it is perhaps an advancement to set them free.

A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.

Newest Most Voted
Inline Feedbacks
View all comments
6 years ago

I like the optimism but hasn’t the industry always said that machines would set the People free but in the end Jobs were discontinued (but of course also others created)?

Baseball makes a ton of Money but they try to be cost efficient too. If one data Analyst can do what 10 Scouts did MLB Teams are not going to continue to pay 10 Scouts on top of the data Analyst.

I don’t think scouting will die completely as some aspects are still not easy to quantify (especially hitting and pitching mechanics – although there are technological advancements in that field too) but I totally see Teams drastically reducing the number of Scouts just like Roboters reduced the number of maual workers on cars.

I’m not judging that and the Change is also creating new highly qualified Jobs (for scientists who are willing to work twice the hours for half the salary- at least if you buy in to some stories that Jeff told in his Podcasts:)) but for the traditional Scouts Technology is absolutely a treat. Eric expressed similar toughts when talking about it so I guess there is quite a bit of fear currently in the Scout industry.

6 years ago
Reply to  Dominikk85

You’re not wrong but you are ignoring the big difference between manufacturing and baseball scouting, which is that a big part of scouting minor and major league baseball players is determining what direction their performance will potentially go, or is likely to go, in the future.

As Longenhagen observes a lot, development of players isn’t linear, its something where a player may have a breakthrough or epiphany on a certain issue and suddenly become greater than the sum of the parts. A big part of player development scouting is figuring out who might potentially make these breakthroughs, or what they might be. Since there’s not free flow of information about players out of their organizations, you need direct observation to infer what is going on. Game data can tell you what has happened, but it can’t tell you why or how it might not continue, and even if it could, it would involve a lot of non-game data (batting practice data, etc) that a team would never release to the other clubs.

That’s why player development is filled with stories like Cody Bellinger, a kid who got drafted for his athleticism and hit tool and had no power at that point, but who unlocked something with a change in swing and became a monstrous power hitter. Some scouts somewhere looked at Bellinger in HS (where he hit 1 (ONE) home run) and thought there was potential he could become a major league baseball player at a corner position. That person was right, but the emergence was sudden, and it didn’t involve tools he was displaying at the time.

You also see that with MLB players. In another Dodgers example, they traded for Tony Cingrani, failed starter, who has another 3 seasons of team control, because they saw something in his pitch selection and location that made them believe they could make him a lot more effective in the long term, even if he didn’t contribute this year. Data may have suggested to them that Cingrani’s slider was a lot more effective than his other pitches and not thrown often enough, but the belief they can help him is also predicated on believing he would be receptive to that coaching and instruction and information and would attempt to change his approach in response.

That may still lead to less scouts, but it isn’t going to eliminate them, because there’s basically no chance whatsoever that data capture will ever be able to record all the data that scouts are able to collect, or that it would be able to make determinations about the value of that data without knowledge of the player’s personality.

6 years ago
Reply to  mikejunt

I’m not sure you’re rejoinder is really addressing Dominikk’s points (or why his original comment, at the time of this writing, had net -1 thumbs). It isn’t just manufacturing that works this way. I’m strained to think of an example where technology doesn’t lead to a decrease in total jobs, regardless of industry. Healthcare (especially nursing), agriculture, textiles, financial industries, tour guides, security guards, sales reps, graphic design…. Maybe education? Where so much of the job is likely to be person to person forever. What, if anything, is the difference between baseball and any other industry, vis a vis how the technology relates to job numbers? Dominikk is right that it’s wishful thinking to suggest that anything but job loss will result. Not saying that’s good or bad, but I don’t know how you could claim baseball is different.

To be clear, I think you’re both saying the same thing and I don’t understand the disagreement. Is it that you take issue with the parallel between baseball and car manufacturing? If he’d compared scouts to, say, fact checkers, would that have made it easier? Surely a little abstraction is at work here, but the historical pattern is simply undeniable: technology reduces the amount of human labor necessary to produce a commodity.

And there is an *extremely* high chance that data capture will be able to record everything a scout can collect, and more. Maybe not be 2018, or even by 2025, but in our lifetimes, you bet. But more importantly, you make the excellent point (as other’s have here, in the podcast, and outside Fangraphs) that this opens up a ton of room for scouts to do something that, at least in the short term, will always require a human job: making determinations about psychology and make up.

Related, do scouts have a union? Do they have any bargaining power so that they can position themselves to ensure that whatever the new scouting looks like, their members will be able to be the ones that get the jobs? Rather than, say, the legion of 27 year old whiz kids who love the game, have PsyDs and will work for nothing? Recent layoffs suggest no. Maybe a dumb question, but part of the conversation?

6 years ago
Reply to  John

My point was that while in manufacturing you have an action being replaced by machine and data, you don’t see that in fields that involve inference and heavy personal interaction, other than situations where automating tangential tasks frees up time for the core work and means less people are needed (just like scouting). You don’t see automation replacing psychologists, managers in any field, human resources professionals, or teachers. Scouts are very much in that category.

That may lead to less scouts. It won’t come close to eliminating them. And the attrition of scouts who aren’t strong enough in those areas of emphasis will probably exceed the lost positions in the end.

A lot of these scout firings are probably organizations getting rid of people who aren’t open to or able to function in this new environment and work with the data.

6 years ago
Reply to  mikejunt

Agreed! At least mostly. It seems like scouts by this account are somewhere in between? Part of their job is easily replaceable by machines, and part of it very much not. Here’s to hoping that many can find ways to adapt, and that they’ll have institutional support in doing so. Yes, scouting seems unlikely to ever be eliminated but it will change, and considering historical examples, there’s no reason to think that that change will necessarily be liberating.

Like the OP, I also find the optimism hard to really believe. Adapting (which is how I understand being “set free” or as you put it “function[ing] in a new environment”) is a lot tougher than people make it out to be. I take your point that some people just don’t want to and hard as it may be for them, letting them go one way or another is an unpleasant reality. But in my experience, even people who *do* want to adapt are often forced out or fired so that an organization can start over with new (often lower paid, or part time) workers. We’ll see.

I’m just writing so this point isn’t lost here. It strikes me as an important one that’s obscured by the optimism. Transitions of this kind are often necessary, always painful, and sometimes inconsiderate and disrespectful. Don’t pretend everything will probably be fine.

Jetsy Extrano
6 years ago
Reply to  John

For every 10 scouting jobs today, maybe there will be 5 new scouting jobs, and 2 of those will be held by current scouts. The other 8 are out.

Imagined numbers, but change hurts, usually. Even when it gives while it takes away, it doesn’t have to be fair about it.

6 years ago
Reply to  mikejunt

Actually things like machine learning starts to cut into those higher qualification jobs and is even supporting in researching psychological patterns. Sales is also changing with more analytics.

Those jobs won’t ever be totally replaced but machine learning algorithms are getting more adaptable and soon won’t just help with analytical tasks but soon also creative tasks.

Humans are still going to be needed but a lot less of them are needed as efficiency grows.

In japan they are even experimenting with robo nurses in retirement homes.

Other jobs are generated of course but generally as efficiency grows the need for human capital becomes lower (sometimes offset by good economic growth).

But people shouldn’t assume that robots just eat low quality jobs, it will expand more and more into into higher qualified and creative jobs.

Also I think that psychological evaluations of players will be done by psychologists and with the help of computer tests rather than scout opinions. Mechanics will also more and more be evaluated in biomechanical labs rather than in the eye test.

Again I don’t think the need for scouts will totally cease but almost all classic scout fields will be shrinking and other professions cut into their jobs.

6 years ago
Reply to  Dominikk85

Dominikk85, what you’re describing is more science fiction than fact. You keep giving examples of automation replacing humans in repetitive tasks, but you have yet to provide one where they have replaced humans in subjective analysis. I can’t think of a field where that has happened, as it’s not something that machines can do well, and I doubt that they will any time soon.

6 years ago
Reply to  jdbolick

jdbolick, what field of work is exclusively subjective analysis? Every single job, from healthcare to art to baseball scouting is composed of significant amounts of repetitive tasks. Right now, technology is allowing us to be more efficient and do the repetitive tasks for us. But in fields like scouting, we are now seeing that scouts spent a lot of time doing repetitive tasks. Without wasting time mapping spray charts, we now only need half or a third of the scouts to do the exact same task as 10-20 years ago.

In healthcare, we now no longer need doctors to pore over x-rays to find the hairline fracture in a bone, or analyse MRIs to diagnose cancer. Machines do this faster and more accurately than we do. This ‘frees’ doctors to do more subjective analysis, but it also means that we need fewer doctors to accomplish the same tasks as before. What happens when machines can more accurately diagnose other diseases based on a patient’s symptoms than a tired, overworked doctor?

So while machines are not eliminating the need for all highly skilled professions yet (though they have severely disrupted ‘highly skilled’ fields like finance and insurance, in addition to agriculture and manufacturing), they ARE ensuring that we now only need fractions of the amount of human labor we needed in the past. And when every single profession is made more efficient, eliminating all the repetitive elements by giving them to machines, what will humanity do with all this ‘efficiency’ and unemployment?

6 years ago
Reply to  willl

What happens when machines can more accurately diagnose other diseases based on a patient’s symptoms than a tired, overworked doctor?

Then WebMD becomes an actual tool rather than a source of annoyance for everyone in the medical profession. Just like Dominikk85, you’re describing things that machines cannot currently do and aren’t realistically expected to do anytime soon. The addition of radar guns to most stadiums didn’t meaningfully affect the number of scouts employed by major league teams. PITCHf/x didn’t meaningfully affect the number of advance scouts either, as far as I know. The point is that the things that the machines can do well are very different than the things teams need scouts for.

6 years ago
Reply to  mikejunt

I don’t think scouting will ever be totally gone but the volume probably will be reduced, especially at the pro level.