Ender Inciarte’s defensive prowess hasn’t gone unnoticed by baseball fans. (via Arturo Pardavila III)
In a box in my bedroom, stashed away somewhere under my bed, is a stack of old scorecards. The graphite scratchings pressed into the stiff card stock show every inning, every batter, every stolen base and caught stealing and groundout to short from a particular game.
On these pages, I learned the building blocks of the game. The hits, the walks, the outs, the balls and strikes: everything came together on that 9×9 grid as I carefully watched every play.
It’s been years since I’ve filled out a scorecard. When I watch a game, I know there is someone sitting in the press box recording every play. I know there’s an intern in an office in Pennsylvania charting the location of every pitch and batted ball. Even if I didn’t know that, I know where I can look up the information. I don’t need to do it myself.
Then again, we don’t do it because we have to. The season-ticket-holder keeping score in her inch-thick laminated notebook in the left field bleachers of Chase Field, or the kid excitedly penciling in “RF – Judge” when the Yankees came to Progressive Field this summer—surely they knew the official scorer existed. And me, well, I’ve still got those scorecards stored away under my bed.
We do it because it engages us, connects us to the game on the field. It teaches us the basics of the game and keeps us sharp when the innings sometimes drag on. It gives our passive watching an active purpose.
It’s the same reason we vote on our All-Stars, and why MLB continues to let us even when we do things like almost vote eight Royals into the game just so Mike Trout can feel like he’s in an episode of The Twilight Zone. When I sat on my bedroom floor as a kid surrounded by paper chads going through all the permutations that let me vote for Tony Gwynn and Barry Bonds on this ballot while still getting in my support for Ray Lankford and Brian Jordan on that one, I was officially a part of MLB. It didn’t even matter that I’d probably never get the ballots I took home back to the stadium in time to actually cast them. I was hooked, and MLB knew it.
At its core, the Fan Scouting Report — FanGraphs’ and Tom Tango’s annual invitation to fans to assess and predict player performance — does the same things. It connects us to the game, teaches us, gives us a say. It has the potential to create the same kind of lifelong attachments that the scorecard and All-Star ballot do. Yet the Fan Scouting Report remains tucked away in the analytic corners of the web, unknown to the average fan.
What if that changed? What if the Fan Scouting Report were as accessible as a scorecard or All-Star ballot? What if we could bring the Fan Scouting Report to the ballpark?
A New Dimension
The Fan Scouting Report fills a niche that neither keeping score nor All-Star balloting can. Keeping score focuses on the who and what, but never really delves into the how and why. All-Star voting gives the fans a voice, but very little in the way of direction or responsibility.
You can learn a lot from keeping score. You learn each fielder’s responsibilities and assignments on various defensive plays when you have to track every assist on every putout. Tracking the baserunners helps you notice things like a runner on second being more likely to score from a single with two outs than with nobody out. You start to pick up on the situations where advancing a runner takes some of the sting out of an out, or when just getting on base is imperative.
It doesn’t really dig deeper into the skills that produce those outcomes, though. That’s not to say you can’t notice those skills from keeping score; if you pay enough attention to what is going on on the field, you’re bound to pick up something. But the act of recording creates a framework for how you think about the game and affects how you interpret what you see.
With scorekeeping, that framework is very results-oriented. You record what happened, but you’re on your own to figure out why. Even the things from a couple paragraphs up, which eventually become second nature to fans, are going to be much easier to notice for the first time if someone tells you to look for them.
All-Star voting has a very open-ended framework. Some fans vote for who they think is the best player. Some vote for who they think is having the best season. Some vote for who they think has the most accomplished career. Some vote for Jarrod Saltalamacchia because it creates more jobs stitching letters onto the uniforms. It’s all about whom each fan most wants to see, and by giving them a vote, you create in them a vested interest in the players on the field.
This framework is all about narrative. It’s about weaving the story of baseball through the names of its greatest practitioners over the years. It’s about seeing Mickey Mantle and Willie Mays and Hank Aaron and Moose Skowron all together on the field at once and knowing that these are the names that will come to define your generation. It’s about giving fans the freedom to shape that story however they want, and there are no wrong answers (at least I hope there are no wrong answers, because I’m pretty sure I am partly responsible for a 40-year-old Ozzie Smith getting voted into the starting lineup of the 1995 All-Star Game).
There’s certainly a place for that at the ballpark. You can’t watch someone keeping score and think there isn’t. And if you were nine years old in 2001, you probably don’t look back at that home run Cal Ripken hit off of Chan Ho Park and wish you’d gotten to see Corey Koskie there instead.
But there is also a place for a framework that focuses on something more in-depth and pushes fans to see more objectively the game’s inner workings. By seeing the scouting categories, fans know right away what they are looking for. Most of them are things long-term fans have an intuitive understanding of, but seeing them explicitly laid out brings a clarity to their importance that can be difficult to enunciate without that framework.
As a result, bringing the Fan Scouting Report to a broader audience would help build a newly educated fan base with deeper connections to the game, as well as give experienced fans a new outlet to invest in.
Engaging the Fans
Reaching that broader audience and creating those connections will be difficult as long as the Fan Scouting Report is just an annual survey. As is, the Report is set up to give useful scouting information to people who want to use the results, but not so much for the impact on fans casting the votes.
To create a lasting impact beyond the face value of the data itself, you need a more persistent presence. Keeping score relies on real-time interaction with the game on the field. All-Star voting ties in with months-long promotional campaigns that encourage fans to get involved and support their team’s players.
This interactivity drives engagement: the more actively a fan can connect with MLB, the more productive that activity becomes in building a long-term relationship. With the Fan Scouting Report, you fill it out once a year and then likely forget about it.
If the Fan Scouting Report were instead available at games for fans to fill out as they watch, it would add a level of engagement missing from the current implementation. MLB could publish an app or a web page that presents the scouting categories for fans in attendance to observe and vote as they see them in action, just like how a professional scout would take notes.
MLB could even take it a step further and integrate the results into its programming. Imagine going to a game as a kid and watching Ender Inciarte get a great jump on a fly ball, or noticing how quickly Brandon Crawford gets rid of the ball turning two, and then going home that night and seeing Harold Reynolds walking through the play on MLB Network highlighting exactly what you saw. You’d feel like the next George Genovese or Edith Houghton.
Now imagine that Al Leiter pulls up a graphic showing Crawford’s release or Inciarte’s first step according to the Fan Scouting Report right after you voted on it earlier that day. You’d feel like Roland Hemond himself.
Those are the kind of things that build lifelong fans. There might even be some sort of residual benefit to the future crop of scouting talent when the kids who grew up with this kind of framework start moving into the profession.
So I think there is clear potential for MLB and fans alike to benefit from this sort of initiative. They aren’t alone, though—analysts who rely on the Fan Scouting Report for its data could also find the expanded scope of the project useful.
Scouting with Small Data
On July 4, 2003, Bo Hart was hitting over .400 with a wOBA of nearly .450. Fifteen games into his big league career, his stat line looked eerily similar to the last Cardinal rookie sensation from two years earlier.
As it turned out, Bo Hart was not Albert Pujols.
This was, of course, no surprise. While it’s impossible to look back now without the tint of hindsight bias, I think it’s fair to say no one expected or even hoped for the kind of things from Bo Hart that they had from Pujols. Even if you didn’t know anything about their respective minor league numbers, lurking behind that doppelganger performance were some pretty obvious distinctions—the way the ball exploded off of Pujols’ bat, the near-Bondsian plate discipline, the perfectly repeatable coordination in the swing—that hinted at a brighter future for Pujols.
That’s not to say scouting data is perfect or that you can ignore performance data. After all, there were players who played with Barry Bonds who said that Wily Mo Peña hit the ball harder than anyone they’d ever seen (which may have been true, but it certainly didn’t portend great things). Even Pujols himself (infamously drafted in the 13th round and having entered his rookie season at the back end of the top 50 prospects) is hardly an endorsement for the prognosticating efficiency of scouting data.
The point is, though, that scouting data is especially useful when you don’t have much performance data. There are going to be times when Bo Hart hits like Albert Pujols for a few weeks. When that is all you have, it really helps to know which one looks like he could maybe have the tools to be the next Joe DiMaggio, and which one’s success inspires stories about how he used to work at Nordstrom and batted ninth his senior year in high school.
This means scouting data is particularly suited to a couple of purposes. One, it helps get a quicker grasp of the skills of players without much of a track record, like recent call-ups or bench players who don’t get a lot of playing time. And two, it helps identify when something has changed with a player.
This is how fans naturally think. If your team’s ace has a rough April, what do you do? You check his if velocity is down. Maybe see if the breaking ball looks as sharp as it did a year ago. If the stuff looks the same, it’s easier to accept that it could just be a bad month that doesn’t necessarily mean anything going forward. If not, well, maybe something’s wrong.
This is especially true for defense, where even the best metrics rely on a large sample size to get a reliable estimate of a player’s abilities. The Fan Scouting Report is great, for example, at helping to figure out whether a large swing in UZR is a fluke or whether something fundamental has changed. This usefulness is blunted when the Fan Scouting Report is only available after the season ends, though.
By running the Fan Scouting Report during the season, we can use up-to-date scouting data to supplement our understanding of small samples of performance data. You can do things like check whether it looks like a fielder has lost a step after an offseason surgery, or whether those annual “best shape of his life” reports from spring training translate to a tangible difference in the player’s tools.
The other area where the Fan Scouting Report should shine—allowing us to get a better grasp on the skills of call-ups or bench players—is also shortcoming in the current implementation. These players have the least reliable defensive metrics, but are also the least likely to get enough votes in the Fan Scouting Report for their ratings to be usable.
(You don’t necessarily need a huge sample of fans to get good scouting data. Tom Tango, the creator of the Fan Scouting Report, has noted that you can get usable results for a player with just five or so votes, and that 10 is generally pretty good, but even those thresholds can be a problem for players who don’t get a lot of playing time.)
Having an in-stadium push backed by MLB would greatly expand the number of fans who participate and help fill in those gaps, but part of the problem is also that it can be hard for fans to form a solid impression of a player they only saw a handful of times. If you’re a Padres fan, you probably have plenty of memories seeing Manuel Margot in center this past season, but maybe not so many of Franchy Cordero. When it comes time to fill out the Fan Scouting Report in October, it will be much harder to come up with meaningful ratings for the latter.
Allowing fans to scout players in real time means they don’t have to rely on long-term impressions. It doesn’t really matter how well you can remember Cordero’s play by the time October rolls around if you already recorded your observations when you saw him back in June.
An in-stadium implementation would also open up additional possibilities that the current implementation isn’t capable of. For example, the Fan Scouting Report currently relies on fans rating players on their own teams—the ones they see on a regular basis. Relying instead on reports from individual games would allow us to compare how each fielder is viewed by his home fans compared to on the road.
You could also allow fans to take scouting assignments ahead of time. It can be difficult to gauge a player’s initial reactions to batted balls when you are watching a game casually: by the time your attention shifts to the fielder (or, if you’re watching on TV, by the time the camera shifts to the fielder), he has usually already reacted.
If you are at a game and can select an option to, for example, watch specifically for the initial reactions of outfielders for the next inning, you can get a more reliable view of their first step and instincts. We could even compare the general ratings for fielders to the ratings given by fans who took out an assignment ahead of time and see what kind of difference looking more closely makes, or break down the votes into those cast in attendance at the stadium and those cast from watching the TV feed to see how much difference it makes to see the plays live.
Even something as simple as month-to-month splits to track the evolution of a player’s perceived skills over the course of a season could be useful. On an even more granular level, we can use the ratings from individual plays to highlight the best examples of each skill on display, or to compare to Statcast ratings for the same plays to see how well the algorithms track with human perception.
As is, the Fan Scouting Report reaches a broad enough audience that it works well, at least for the vast majority of regular starters, but it does limit what we can do with the data. Opening up the data further would expand the possibilities and introduce new options that would make the Fan Scouting Report even more useful.
Scouting in a Statcast World
Fifteen years ago, when the Fan Scouting Report was just starting out on Tangotiger.net, UZR was not readily available. FanGraphs didn’t yet exist. Defense was a huge hole in our statistical understanding of a player’s value, and the Fan Scouting Report was a major step toward filling that hole.
Now, in an age where Statcast is constantly giving us new data on the same underlying skills the Fan Scouting Report was meant to uncover, its place feels somewhat different. It’s not that Statcast has made the Fan Scouting Report obsolete—there is definite value in having another independent perspective, both as an additional data source and as a point of comparison and sanity check for the Statcast data.
Even if Statcast were to someday make the Fan Scouting Report completely redundant to analysts, though, the Fan Scouting Report would still have its place. It’s like learning basic arithmetic when you have a calculator that can do the same thing. Sure, Statcast can give you a pretty good idea of who the fastest outfielders are or who has the strongest arm, but knowing what to look for and how to spot it yourself still gives you a better idea of what is going on.
Much like the ubiquity of scorecards has helped introduce generations of fans to the terms and concepts of statistical record keeping, the Fan Scouting Report can bring that same level of literacy to scouting concepts and give fans a clearer vocabulary to define concepts they understand intuitively. Having those concrete terms can help crystalize our understanding from something nebulous into something we can easily explain.
Besides, we don’t do it because we have to. It may have started that way, back when the Red and Green books were mythical treasures and fans got their stats from the back of Bowman and Topps cards, but it evolved into something else. Something worth more than the sum of the squares on the grid or the marks on the ballot. We do it, above all, because it connects us.
References and Resources
- Associated Press,“MLB Won’t Print Red and Green Books”
- Jack Curry, The New York Times,“Cardinals Rookie’s Slam Upstages a Mets Rally”
- Gordon Edes, Boston.com, “This Raw Material Hints at Lode of Talent”
- FanGraphs Library, “FSR”
- Bryan Grosnick, Beyond the Box Score, “A Brief Interview with Tom Tango!”
- Tom Tango, Tango on Baseball,“The 2016 Scouting Report”