This started with a crush on Starling Marte.
Marte is an outfielder with the Pittsburgh Pirates and my favorite baseball player to watch. He's fast enough to have stolen 47 bases last year, third-most in the majors. His arm is one of the best in the game, capable of firing 101 mph bullets from the outfield. He is strong enough to have hit a baseball 460 feet, something only 48 other hitters have accomplished outside of Coors Field over the past two years. (Bryce Harper hasn't. Mike Trout hasn't.) He might be the most creative base slider in the game. He has demonstrated a talent for getting hit by pitches. Over the past three years, he has been the best defensive left fielder in the game, and this year, he will finally get to be a very good defensive center fielder.
I also know (or think I know) exactly how good Marte is: He was the 28th-best position player last year (by WAR) and the 53rd-best hitter (by OPS+). You might think those advanced stats are junk, but whatever stats you prefer, you have some idea how good he is: The 13th-best hitter (by batting average) or the 91st-best (by runs scored). We've all got stats. We all use our stats.
What if we had none? Not just no WAR but no nothin'. What if some ministry of information outlawed the collection of baseball statistics and we were all left to judge players exclusively by what we saw, what we perceived and what we remembered? Who would be perceived as the best player in baseball? Who would be the first player chosen in a franchise draft? Or, the more important question: With how much eye-rolling would actual major league general managers respond to a weird thought experiment on the subject?
I decided to find out. Over the past week, I've asked seven anonymous major league insiders, ranging from veteran scouts to new-school GMs, who the first pick in a franchise draft would be if there were no stats.
The rules were simple enough: Everything about baseball is the same, but counting anything other than team wins and losses is illegal. Scouting is allowed. Radar guns are allowed, and even Statcast data are generated, but everything except what we can hold in our memories gets wiped clean before each pitch. From this limited data -- our eyes, our memories, circumstantial evidence and team records -- teams must be built.
Many names came up. More interestingly, four philosophies about the eye test emerged.
1. The Utter Anarchy philosophy
"Been thinking about this for two hours," the Utter Anarchy exec said, "and I'm stumped. Every time I think I have an answer (Trout, Harper, Corey Seager), I have no idea if it's because I know the numbers or because that's who I would pick out based just purely on seeing them."
The underpinnings of this philosophy probably trace to Bull Durham's gork scene.
If the difference between an All-Star and a fourth outfielder is a gork a week, and there is no way without stats to notice those gorks, then you're left with 750 major leaguers spread out in a massive ocean of gorks.
Imagine you're watching a game on Opening Day. If you're paying close attention, you might notice that one guy goes hitless, strikes out a couple times and makes the final out of the game. Maybe he's lodged in your memory because of that final out, so you notice he goes hitless in the second game too. Finally, he gets two hits in the third game and one in the fourth and one in the fifth. Is he good? How about if he goes hitless in the next game but then homers in the seventh, then goes hitless, then two singles, then hitless, then 1-for-4, then 1-for-4, then 1-for-4, then hitless, then three hits. Is he good?
Most likely, you have no idea. Even right now, staring at that paragraph, you have no idea because you aren't allowed to add all those games up and figure out whether they add up to something good. In fact, I can tell you what they add up to -- a .259/.359/.370 slash line -- and you still don't know if they're any good unless you know everybody else's slash lines. And that's just one guy, whom you happen to be paying suspiciously close attention to. There are 750 active players, spread out across 15 games every day. What are the odds you'd remember all 60 plate appearances from one player if you were trying to keep track of hundreds?
Also, that's the game log of Mike Trout in the first two weeks of his 2016 MVP campaign! His performance in those two weeks was wildly misleading, and the only way you would know that is if you somehow remembered what he did in hundreds of other games spread across dozens of other cities over his previous other years.
"As professionalism infiltrated the game, teams began to bid for star-caliber players. Stars were known not by their stats but by their style. Every boy would emulate the flair of a George Wright at shortstop, the whip motion of a Jim Creighton pitching. But (Henry) Chadwick recognized the need for more individual accountability, the need to form objective credentials for those perceived as stars. So in 1865, in the Clipper, Chadwick began to record a form of batting average." The Hidden Game of Baseball, by John Thorn and Pete Palmer
In the Utter Anarchy school of this exercise, everyday hitting consistency -- which makes up a huge part of a player such as Trout's value -- is useless as data. Even filtering down the 750 players into a manageable pool of candidates would be challenging. I might propose certain heuristics, such as, for instance, not bothering with short players. (But Mickey Mantle was "amazingly short," Pat Jordan writes in "A False Spring.")
Or you'd limit yourself to players who hit a ton of home runs, assuming you can keep even the rough frequency of those in your head. (Mark Trumbo and Chris Carter led their leagues in homers last year and finished 86th and 108th in Wins Above Replacement.) Or maybe you'd trust that a pitcher who throws harder than everybody else and doesn't seem to walk many batters can't possibly be bad. (Franchise cornerstone Nathan Eovaldi?)
If it's Utter Anarchy, I can start making cases for all sorts of wrong picks. Could you prove that Javier Baez isn't better than Trout without using numbers? Baez plays shortstop -- very well -- which is something maybe 20 people in the world can do. He makes incredibly smart plays, athletic plays, miracle plays, really. He's strong, he runs well, he smiles, he's handsome, he's younger than Trout, and he plays for a much more successful team. Are you that confident that you can tell Baez or Jose Ramirez or Edwin Encarnacion or Brian Dozier or Kevin Pillar or Yasmani Grandal or any number of very good players with extremely different skill sets apart -- or that you could prove any of them isn't better than Trout?
Or Marte? Throws harder than Trout. Runs faster than Trout. Can hit the ball about as far as Trout when he lays into one. Gets about as many hits as Trout. Strikes out a little less than Trout. And seriously, have you seen how good he is at sliding? The actual difference between Trout and Marte is massive, as big as the gulf between Barry Bonds and Bobby Bonds. It's also only a homer a week and a bunch of walks. Would you know that?
Utter Anarchy exec never did commit to a pick.
2. The Pretty Good Filters philosophy
This philosophy -- and the next one, for that matter -- takes as a given that we wouldn't be smart enough to get the question right but that we're too smart to get it all the way wrong. That there are only 20 or 30 players who would be considered, and all of them are great. If the Utter Anarchy philosophy is that you'll end up with Marte if you're lucky, this one is that you'll get Marte in the worst case.
For instance, in the example of Trout's first two weeks of last year: While he was hitting .259/.359/.370, some of his teammates were hitting better. Kole Calhoun was. Yunel Escobar was. Geovany Soto was. If you were such a close watcher that you noticed they were all hitting better than Trout, there'd still be enough circumstantial clues to tell you Trout is the better player. While Soto was playing three times a week, Trout was in the lineup every day. While Calhoun was batting sixth and playing right field, Trout was batting third and playing center. And while Escobar might have been hitting better, teams were intentionally walking Trout -- not Escobar.
In other words: Even if we can't see and judge and record Trout's every at-bat, his manager can, and if his manager thinks Trout is the best hitter and capable of handling a premium position, we can at least trust that data. We might intuit that Trout was good because he was a teenager when he debuted in the majors. We would see how rarely he was pinch hit for or asked to lay down a bunt.
Of course, that still might not be enough to differentiate Trout from Andrew McCutchen, Adam Jones or Lorenzo Cain, but it narrows our pool considerably. You start to pick a favorite player who checks off a number of filters: Big, strong, hits in the middle of a good team's lineup, mobile enough to move around the field, young -- and you take Kris Bryant. Or tall, middle-of-the-infield player with leadership charisma, obvious pop and youthful handsomeness on a successful team -- and you take Carlos Correa.
The most popular pick from the Pretty Good Filters types was, perhaps surprisingly, Francisco Lindor. Our respondents kept bringing up boxes he checks: He plays a premium position for a winning team, and a couple days of watching him makes it clear that he's not merely acceptable but excellent at the position; he's fast; he makes a lot of contact, which (I speculate) is easier to notice than many other offensive skills; he hits the ball hard; he's affable; he smiles; he's young.
He's also small, and I'd be nervous picking somebody who wasn't a physical specimen -- nervous that I'm getting Erick Aybar, who in his prime would have checked most of those boxes but was a much worse player than Lindor. But Aybar never batted third for a good team, as Lindor does. There are enough filters, this philosophy posits, to eventually eliminate the Aybars. (For that matter, Aybar was very good as well.)
3. The Wow! philosophy
If you showed me Giancarlo Stanton and Lindor standing next to each other, it might take an entire career before you convinced me that Lindor is better. To the extent that you can ever build a case that outweighs "look at the demigod," it will rely on your memory. And what's going to stick in your memory: a slightly higher percentage of quality at-bats or a baseball traveling 504 feet?
There are other players with "wow tools," as one exec put it, but not all "wow tools" elicit the same volume of Wow! The hitting equivalent to Stanton -- Tony Gwynn, say, or Joey Votto -- could slap hundreds of singles the other way or take hundreds of tough pitches just outside of the strike zone, but none of those hundreds would stay with you like a baseball traveling 504 feet would.
Range, as a defender -- such as Kevin Kiermaier's or Andrelton Simmons' -- would scarcely be noticed, nor would a pitcher with great command or a great changeup catch your eye. Throwing arm and speed would be noticed, but less so than Stanton's power. The only thing that might compare would be a pitcher's stuff, particularly his fastball velocity.
Which brings up Noah Syndergaard, the other pick from somebody in the Wow! school of execs. He wouldn't pick between the hitter Stanton and the pitcher Syndergaard, but he said they both have "the most 'wow' factor, which I think is what scouts will remember more than steady success." Both, a different exec pointed out, have good baseball skills in addition to their tools. Syndergaard throws strikes along with throwing 100; Stanton is a solid defender and keeps his strikeout totals under control. You wouldn't mistake them for Eovaldi or Trumbo based on a single loud tool.
At the very least, one figures there is a high floor in picking somebody who can hit a baseball 500 feet or throw one 100 mph repeatedly. It might not get you the best player in baseball, but what are the odds anybody picking without any numbers will get the best player in baseball? Well ...
4. The This Is Actually Easy philosophy
"How many games do I get to see?" one exec asked.
As many as you want.
Okay, then. Five years ago, when Trout was still an elite runner as a rookie, I suspect I would have accepted this certainty. Five years ago, the physical obviousness of Trout would have made this a very different article. But these days, he's merely a good runner, and as noted above, recognizing his massive edge on somebody such as Marte requires noticing a lot of small acts in the batter's box, spread out over months.
Nah, smarter people than I said. "I still think it's Mike Trout," a This Is Actually Easy exec said. "Five-tool [center fielder] who can make spectacular plays defensively and do everything in the world offensively. Watching him play is special, and he would make any team better. Seems pretty simple to me."
"I think you'd quickly narrow it down" to just a few candidates, another said, and from there, "the consistency of [Trout's] impact on a day-to-day basis stands out. He's basically always good."
The main thing here isn't just that the answer is Trout but that these execs are confident it wouldn't be hard to see. One said he had no doubt scouts would recognize Trout's greatness.
"I like Marte, I really do. This guy is pretty remarkable as well," he said. "But Trout -- I saw this guy in the Arizona Fall League [in 2011] when he had a horrible statistical performance, but it was all still there, even the way he wore his cap was like Mickey Mantle. His strength, his body, his eye-hand coordination, his bat-to-ball skill combo, you can see it in BP. The way he has an incessant desire to save runs -- climb the walls, run into walls, dive for balls, always backing up plays, never out of position, great teammate, his incredible, positive arrogance. You name all those factors plus, and he's a superstar."
I believe him, I think. I'm giving up on my alternative universe in which Marte is the most famous athlete in the world. But it's worth pointing out that the industry once had the chance to prove it. Trout was a draft prospect. His high school stats were basically worthless as data, as all high school stats are.
The teams who made the trip to Millville, New Jersey, saw Trout and were impressed -- but they didn't put him first among draft prospects. Or second or third or even 23rd. Even two years later, when he was embarking on the greatest rookie season ever, there wasn't a consensus that he was even the best prospect in baseball.
That's not the same thing as what we're talking about. Making draft picks and ranking prospects are all about projecting growth and development, not necessarily picking the best player for a team today.
But it'd be easier to believe the This Is Actually Easy group if they had just shown us how easy it was -- instead of picking Donavan Tate, Bobby Borchering and Jared Mitchell ahead of Mike Trout.