Q: Are We Not Men? A: We Are Robo!
Reader, I gasped.
Sorry, I’ve just always wanted to write that. I’ll tell you later on why I gasped. Let’s start here. Last Tuesday, while making his Grapefruit League debut with the Blue Jays, Max Scherzer challenged a pitch. Then he challenged the challenge system. Scherzer’s start against the Cardinals marked his first experience with the automated ball-strike system, which is being rolled out in some spring training facilities this year, continuing its inexorable, years-long creep toward implementation in regular season games. Scherzer, for one, does not welcome our new robot overlords.
On his 11th pitch, Scherzer fired a 1-0 fastball to Lars Nootbaar, just clipping the outside corner, at least according to home plate umpire Roberto Ortiz, an organic life form who uses an inefficient pair of weird, goo-filled orbs to assess pitches. Nootbaar – who, we should note, played some rehab games in Triple-A last season, and so was at least somewhat familiar with the challenge system – immediately patted his head. That’s the official way to request a challenge (though I would strongly encourage the league to require the player to rub their stomach with their other hand too). Scherzer, never a fan of waiting around when there’s pitching to do, canted his head from side to side like a racehorse in the starting gate. The machines spoke: The pitch was 2.3 inches off the plate, or as the humanoid Buck Martinez put it, “way outside.”
Just like Scherzer, this was my first experience with the challenge system, and I found the graphic adorable. That’s the point, I guess: implement an all-seeing eye that judges everything and everyone with detached, ruthless precision, then soften it with a lovable cartoon face. Scherzer recovered to strike out Nootbaar, then made his own challenge in the second inning. The right-hander, who tracks his pitches using the same goo-based technology as Ortiz, didn’t agree that he’d missed low with a 1-0 curveball to JJ Wetherholt, and he pounded the top of his cap like a bongo drum.
Reader, that’s when I gasped. Then I laughed. I gasp-laughed. According to the delightful ABS graphic, the pitch was hilariously low. It was nowhere near the strike zone. This pitch was in the Cactus League. It was so far away that Social Distortion wrote a song about it called “So Far Away.”
In fairness, we should acknowledge a few things. First, one of the consensuses that emerged during last year’s test of the challenge system was that teams should disempower the pitcher from making them. Catchers are right there, and they have a much better sense of the actual location of the pitch. Second, Scherzer indicated after the game that the challenge was more an experiment than an expression of his certainty that the pitch had clipped the zone. “That was a rare occurrence for me, with a curveball down, to actually see if that’s actually a strike or not,” he said. You’re allowed to take that notion with a grain of salt. Part of me believes Scherzer, but, uh, he was bopping himself on the head with a lot of conviction.
We should also note that the steep shape of a curveball makes it hard for the pitcher in particular to judge the exact spot where it crosses the plate. The really interesting thing is that curveballs are actually relatively easy for umpires to judge. That’s not necessarily intuitive. Curveballs approach the plate at such a steep angle that they hit the catcher’s glove (or the dirt) far lower than they cross the plate, which might fool the umpire into thinking a pitch was lower than it was. And curveballs that come in at the very top of the zone leave the pitcher’s hand so high and possess such a loopy shape that they also might be hard to recognize as strikes. Where a fastball or cutter pushes straight through the zone cleanly, a breaking ball slices through it at an oblique angle, and it just seems logical that the more of the zone a pitch catches, the more likely it is that it will be recognized for doing so. But apparently that’s wrong. I broke down the 2024 stats for curveballs and fastballs (four-seamers, sinkers, and cutters) in three areas of the zone: the heart, the top and bottom of the shadow zone down the middle, and the top and bottom thirds of the zone down the middle. I’ve highlighted those areas in pink.
In all three areas, Curveballs had higher strike rates than fastballs. On pitches over the heart of the plate, it was a matter of a few tenths of a percent, but in the middle graph, curveballs were ahead 83% to 80%, and on the right it was 83% to 81%. Maybe it’s just that curveballs are easier to judge because they’re slower, but umpires are better at recognizing when they’re strikes, so in that sense Scherzer picked a bad pitch to challenge. However, much like Scherzer’s curveball, we’re drifting away from our main objective here. We’re focusing on how far the pitch was from the zone, and just to reiterate: It was far.
However, you might notice something about that graphic: There’s no distance measurement. When Nootbaar challenged in the first inning and earned his Nootbaal, the graphic zoomed way in to show us the exact size of the miss down to a tenth of an inch.
When Scherzer challenged, no measurement popped up, and I suspect that I know why. I think this is a deliberate decision made to avoid embarrassing a player who challenges a pitch that’s not particularly close. Nearly all challenges that end up as balls will show the miss distance. But if the pitch doesn’t even touch the shadow zone – that is, if it’s not even within one baseball-width of the strike zone – the graphic leaves off the exact distance so as to avoid blowing up the pitcher’s spot. Max Scherzer, trailblazer that he is, has showed us that although robots don’t feel, they can still be programmed to blush.
Don’t worry. We’re still going to blow up Scherzer’s spot. Because of all the cool graphics, it’s still really easy to get an exact measurement for pitches that land in the Zone of Embarrassment. We know the measurements of just about everything else on the screen. We know the strike zone is exactly 17 inches wide and the ball is approximately 2.9 inches wide, and through the magic of Statcast, we know that because Wetherholt is 5-foot-10, his strike zone is roughly 18.55 inches tall. I threw a screengrab into Photoshop, measured each of those constants, then used the ratio of pixels to inches to calculate the distance. The ball was 3.98 inches from the strike zone. It missed the shadow zone by more than an inch. It crossed the plate just over a foot off the ground.
That looks pretty damning, but allow me to blow your mind for a moment. If we’re being fair to Scherzer, we need to acknowledge that the pitch was actually much closer to the rulebook strike zone than Statcast makes it look. Let’s think about it under the rules of the current, non-computerized strike zone. Keep in mind that this was a curveball breaking downward. Now let’s look at the way that the Hawkeye cameras measure a pitch, courtesy of an MLB.com explainer by Anthony Castrovince.
Keep your eye on the diagram on the right. Statcast’s strike zone is two-dimensional, and it’s measured from the very center of home plate. That’s a perfectly reasonable way to design an ABS system – an earlier version was 3-D, so it seems safe to assume that this 2-D version is, for some tangible reason, an improvement upon it – but it’s not the way the strike zone has worked for the entirety of baseball history, including right now. The rulebook definition starts like this: “The STRIKE ZONE is that area over home plate…” and that’s really all we need to know. The strike zone is three-dimensional. It’s seven-sided, a pentagonal prism, and the ball just needs to clip any part of it in order to be a strike.
At The Athletic, Jayson Stark had the good fortune to be present in the clubhouse after the game, when Scherzer found out that the robo-zone didn’t match the rulebook zone: “Wait, I thought it was the whole plate,” he said. “So now we have to redefine what the strike zone is? You said it was a 3-D zone. Now we’ve got a 2-D zone? Hasn’t it always been a 3-D zone?” The answer to that question is yes. It has always been a 3-D zone and it still is, but now there’s also a 2-D zone. There are two strike zones. We’ll dig into the philosophical implications of this dichotomy later, but for now, that’s how umpires are judging pitches, so why don’t we try measuring things that way?
Let’s start with how curveballs work. Their path gets steeper and steeper as they approach the plate. There are plenty of reasons for this. Curveballs actually leave the pitcher’s hand traveling slightly upward; the classic way to recognize a curveball is seeing it jump up out of the pitcher’s hand. The magnus force created by the ball’s topspin pulls it downward, and that force compounds upon itself over the length of the pitch. Here’s where it starts. The baseball is traveling horizontally, and the topspin interacts with the air to start pulling it down.
Now it’s traveling at a steeper angle, but guess what? The topspin is still pulling it downward, so its angle is going to keep getting steeper and steeper as it goes.
Moreover, gravity amplifies this effect for a pitch that’s always breaking downward. Air resistance slows the pitch down as it nears the plate, but gravity is pulling the pitch downward at a constant rate. So say it takes a tenth of a second for the ball to travel the first 10 feet toward home plate, and in that time, gravity pulls it down five centimeters. By the time it reaches the plate, it’s going slower, so over the last tenth of a second, it only travels eight feet, but gravity is still pulling it down five centimeters. All the numbers in this example were completely made up, but you get the point; the ratio of downward movement to horizontal movement is increasing. A curveball’s approach angle keeps getting steeper. You can see it in Statcast’s 3-D pitch visualizations.
These are two actual Scherzer curveballs from last year. We’re going to focus on the bottom one, which came in a bit below the plate. The red line shows a straight line between the position of the ball when it crosses home plate and the position when it’s 50 feet away.
Now, let’s zoom in and look at the path of the pitch over the last few feet of its journey. As you can see, our new purple line is significantly steeper.
None of this should be particularly surprising if you’re familiar with Alex Chamberlain’s primer on vertical approach angle, but the point is that curveballs, with their sharp downward movement compounded by gravity, are the steepest pitches of all. According to Alex’s pitch leaderboard, Scherzer’s curveball averaged a vertical approach angle of -9.9 degrees last season. For now, let’s assume this pitch had the same VAA. With help from our friend Pythagoras, we can calculate that a pitch traveling at an angle of -9.9 degrees would be 1.48 inches higher when it crossed the front of the plate than when it crossed the middle of the plate. Here’s how that works.
OK, so measuring at the front of the plate, the pitch comes in 1.48 inches higher. It’s now missing the zone by just 2.50 inches. It’s well within the shadow zone. That certainly makes it sound a little closer, don’t you think? Here’s what that looks like in our original diagram.
You know what? It’s still pretty far away from the strike zone. Stark’s article mentioned that after the game, reporters told Scherzer that his pitch would have been a strike according to a 3-D zone. They were way off base. In order to do so, the pitch would have had to arrive at the plate with an absurd VAA of 25 degrees. That ain’t happening. This pitch is still unequivocally a ball. There’s no system – goo-based, camera-based, vibes-based, none – in which this pitch hits in the strike zone. It was so far away that Carole King wrote a different song about it, also called “So Far Away.”
That said, I do suspect that this particular curveball actually had a steeper VAA than -9.9 degrees, making it a bit closer than the graphic above indicates. Just using the old-fashioned goo-orb test, it looked sharper than the typical Scherzer curveball. Second, I was talking things over with Michael Rosen, our resident pitching genius, and he got curious and pulled data for a Scherzer curveball, just one random curve from 2023. That pitch had a VAA of -10.1 degrees over the last 10 feet.
That VAA would move Scherzer’s pitch a few hundredths of an inch closer to the zone, and this one solitary, particularly sharp curve could’ve been even closer. It’s still not a strike according to any definition of the strike zone, but it highlights the disconnect between the two current competing versions.
So far we’ve only been talking about the front of the plate, but this would also be true of both the back and the sides. A pitch with a steep horizontal approach angle can clip the corner of the plate before it reaches the midpoint. The back gets tricky because of the plate’s pentagonal shape, but it’s still possible; the closer to the center a high pitch is located, the better a chance it will have of dropping down and catching a piece of the rulebook zone. ABS would tell you that every one of the pitches illustrated below is a ball. But according to a normal three-dimensional strike zone – which is what umpires are calling – that’s not actually true. It’s smaller than the rulebook zone.
As things stand, when the league does implement an ABS challenge system for regular season play – and at this point, that seems like a virtual certainty, though which regular season is still undecided – then the game will officially have two different strike zones. It’s possible that the league could change the rulebook definition for umpires so that it matches the Statcast zone, but that strikes me as unlikely for many reasons, chief among them it would essentially turn the iconic shape of home plate into a vestigial appendage. In the two-zone world – the world that Triple-A players have been living in for a while now – a pitcher would be able to throw a strike, get robbed by the umpire, challenge that incorrect call, and lose the challenge because according to the robot umpire, the pitch really was a ball. Even crazier, the pitcher will throw a strike, the umpire will get the call right, and then the batter will challenge it and that correct call will get overturned! The umpire and the computer will make two different calls, and both will be correct because they’ll have two different zones.
As the numbers from our curveball example show, we’re not talking about a couple of unlikely edge cases. The differences in movement from the front and back of the plate to the middle aren’t minuscule. Some pitchers’ curves average above 11 degrees of VAA, and the sweepiest sweepers average more than six degrees of horizontal approach angle. We’re often going to be talking about well over an inch of difference. This is going to happen all the time. I’m not the first person to notice this. On Wednesday, Baseball Savant’s Tom Tango crunched the numbers and announced that in 2024, one percent of all takes would have fallen into this category, just for issues with the front and back of the plate.
As things look right now, baseball will soon officially have a human strike zone and a robot strike zone. The robot strike zone will be so thin as to be non-existent, while the human strike zone, as it always has, will be shaped like an infinite number of infinitely thin home plates. Honestly, I don’t know how any pitcher who’s had it fully explained to them will avoid succumbing to paralysis halfway through their windup and toppling off the mound simply because they’ve exhausted their ability to process the disjuncture of the situation.
I mentioned earlier that setting up the robo-zone in two dimensions rather than three was a perfectly reasonable choice. The more I think about it, however, the more I think it might be the only reasonable choice. Calling balls and strikes is incredibly difficult. I’ve had to do it before, and I’d approximate that I felt 100% certain on about 30% of the pitches I called. But even then, I doubt I was really thinking about the strike zone the way the rulebook demands. The rulebook zone doesn’t have four corners; it has 10 corners. And it doesn’t have an edge; it has 15 edges. The difference between a two-dimensional plane and a three-dimensional space is the difference between a topographical map and a mountain.
On the one hand, this makes me wish the robo-zone were three-dimensional, just because I’m imagining how much more fun the challenge graphics would be. We’d see in precise cartoon glory not just whether the ball nicked the corner of a box, but one particular corner of a 10-cornered pentagonal prism. It would rule. On the other hand, it’s absolutely preposterous that we ask human beings to process information with anything approaching this level of precision. Wherever you’re sitting right now, try to imagine a pentagonal prism floating in the air next to you. Now try to picture yourself deciding whether a Tarik Skubal fastball nicked one of its seven sides. Now do it again, but first squish your prism down a bit because Nick Madrigal is up next. So maybe it does make sense to have two zones; we’ve just got them reversed.
Scherzer was candid and engaging with reporters, and after processing all of this information, he closed with the takeaway that most of us saw in the headlines: “Can we just play baseball?” he asked. “We’re humans. Can we just be judged by humans? Do we really need to disrupt the game? I think humans are defined by humans.” When he puts it that way, it’s a pretty reasonable request. Right now, umpires and batters track pitches using the exact same equipment, and that makes plenty of sense. If the game is played by humans, it’s certainly not laughable to feel that human eyes and brains should be deciding what’s a strike and what’s a ball. I don’t mean to say that there’s wisdom in every mistake simply because it’s made by a human, but once a computer is making the decisions, the objective of the game becomes slightly less fun, for the same reason that playing chess against the computer isn’t particularly enjoyable. It becomes less of a game and more of a problem solving exercise.
These days, there’s no end to the ways that computer programs are judging us – CAPTCHA requests, Spotify recommendations, suspicious login emails, targeted advertising, personalized search results, automated insurance denials, the artificially indiscriminate firings going on throughout the federal government – and with vanishingly few exceptions, the people being judged would like nothing better than to smash all of these robot judges with a hammer.
Don’t get me wrong, I would love to smash the computers that turned Google into such a joke with a hammer, but the difference here is that many of those systems were designed as shortcuts, either to save time, to replace human workers, or to shift accountability away from the person instituting a crappy policy and onto the circuit board that implements it. On the other hand, the challenge system is a particularly elegant solution to the problem at hand. It will introduce an extra layer of accountability into umpiring without replacing the umpires or undermining their centrality to the game. It won’t obliterate the value of pitch framing, but it will hopefully reduce the amount of shouting umpires have to bear. Now that we have the ability to know the exact location of every pitch, it’s probably not completely defensible to just ignore that knowledge. Instant replay was instituted for the same reason. “I like it when somebody screws up and somebody gets screwed over” is not exactly a winning campaign pitch.
Let me hit you with one last disconnect. The really funny thing is that depending on how you look at it, Scherzer is both the best and worst messenger for this argument. He’s a sure-fire Hall of Famer and a longtime union rep. He’s not afraid of a fight, and his standing in the game ensures that when he speaks, people will listen. His comments warrant plenty of counterarguments, but “Max Scherzer doesn’t know what he’s talking about” is not among them.
On the other hand, Scherzer has never had that much use for umpires in the first place. Since Sports Info Solution started tracking pitches in 2002, 328 pitchers have thrown at least 800 innings. Scherzer’s 14% swinging strike rate ranks ninth among them and his 27% whiff rate ranks 19th. His 17% called strike rate, however, ranks all the way down at 212th. Scherzer has always succeeded by racking up whiffs, pumping his fastball by hitters and tempting them into chasing sliders and curves. Relying on the umpire for called strikes has never remotely been his game. In fact, since 2008, Statcast says he’s had 1,262 would-be strikes stolen from him, third-most in all of baseball. Few players have relied less on human umpires or accumulated more reasons to be fed up with them than Scherzer. Maybe we should tell him that after his next start. I’m sure he’ll have something interesting to say about it.
Davy Andrews is a Brooklyn-based musician and a writer at FanGraphs. He can be found on Bluesky @davyandrewsdavy.bsky.social.
If it’s not 3-D it’s not accurate, and should be shelved until it can be made accurate. Of course, what should be and what will be are often different things where the Lords of Baseball are concerned.
There is no reason to believe it can’t be accurate in 3 dimensions right now. The league has just chosen not to do that for some reason.
You have it exactly reversed…
There is no reason to believe that it can be accurate in 3D right now. And there is a reason to believe that it cannot be…
The reason to believe that it can be accurate in 3D right now would be a demonstration that it can be accurate in 3D.
The fact that nobody has seen this demonstration is evidence (fairly weak evidence, but evidence nonetheless) that it is NOT accurate in 3D right now.
Erego we have 0 evidence that it can be accurate in 3D right now. We have some evidence that it cannot.
Would theoretically be pretty simple to change this, but as of now, your conclusion is the opposite of what logic would demand.
They were using a 3D zone for testing in the independent league. The problem is that it kept calling strikes on sinkers and changeups that barely clipped the bottom front edge and were essentially unhittable.
To me, that’s a reason to redefine the strike zone, not get rid of the 3D zone altogether. Or make both the ABS and umpire -called zones 2D.
Except that by the rule book, that’s a strike, unhittable or not. It’s more hittable if the hitter isn’t standing at the back of the box.
Yeah, I’m not suggesting that redefining the strike zone is THE solution to such things. Batters can stand further forward in the box. Pitchers who really can clip the edge of the zone should be rewarded, etc…
But if the answer to “It’s too accurate” is to scrap it in favor of something worse, that’s some serious illogic.
I guess I would suggest that redefining the strike zone should absolutely be in play.
The strike zone was not handed down by God as objective truth; it’s entire purpose is to force the pitcher to throw hittable pitches so that the game is worth watching.
I think there is no question that pitchers now are throwing with spin and velocity that was unimagined when the strike zone was originally defined. I would also bet that some of the supposed inaccuracy and human error of umpires is (in small part!) a result of human interpretation of what is actually hittable. It seems very likely to me that there are slivers of the defined strike zone that are darn near unhittable for anyone (without adjusting for it so much that it makes other pieces unhittable).
Redefining the strike zone in this case would be bringing its original purpose back into alignment with what is possible.
To be fair, the strike zone has been redefined a few times over the years. Maybe we need to go back to the lower part being the top of the knees as it was before the current zone changed it to just below the knee cap in 1996.
You sound like a boomer shouting at clouds. Anything is more accurate than the umps. Using “it’s not perfect” as a reason to ditch something that’s greatly improved over the current methods is the epitome of boomerism.
I don’t see it as greatly improved if it’s just a challenge system. Changing a few pitches might be an improvement, but it comes at the cost of the game flow, and I’d hardly call it a “great improvement.” Plus what’s with the dickish boomer comments? Fuck you.
I believe the problem was actually that they tried the 3D zone, but it felt very unnatural to everyone involved, and was calling strikes that no umpire would call a strike. While there is a rulebook strike zone, the way the strike zone is called in practice differs significantly from it, and replicating the established practice of how the strike zone is called is more important than fitting the arbitrary rulebook definition
I agree with your larger point, but “arbitrary rulebook definition” is quite a phrase.
The problem is actually worse than this. It’s not just inaccurate because it’s not 3D, it’s also inaccurate because the top and bottom of the ABS strike zone are not correct based on the rulebook definition. They’re being set based on still frames of the batter’s stance at some previous time.
The technology is simply not being implemented by people who care about accuracy to the rulebook strike zone — or the de facto umpire strike zone — anywhere near as much as they care about expediency.
As other commenters and articles have raised, there’s also a major problem about whether an accurate, precise, rulebook strike zone is actually what people want to be called, or whether umps actually do a bunch of finessing to their de facto zone that, in fact, makes the game better and shouldn’t be discarded. But, given that there has never been a technology that even came close to calling the rulebook strike zone, that debate is considerably more theoretical than is commonly acknowledged.
I am ok with a non-3D zone, because it’s STILL BETTER than what we have now with humans. A human umpire isn’t any better at determining whether the ball is in the strike zone in 3D than 2D, and either of those, he would still be worse than an automated system in 2D.
Where’s your source on that? For all we know, umpires very well could be getting more calls right in a 3D space that the 2D system gets wrong than the calls that the umpires get wrong.
Source:my observations plus any site I have seen where they grade umpires. I don’t have time to look for this information to prove it to you (something that should be pretty obvious without having to prove it with a source), but you can research this yourself if you disagree and get back to me.
It is calls like THESE we are trying to fix:
https://www.youtube.com/watch?v=8zbWxE1GAWA
How long does it take for a challenge to go through? I haven’t seen it in action yet, but it seems deeply weird to undo the big win from the pitch clock (game moving faster).
The whole process takes about 15-20 seconds or so, it’s actually pretty seamless.
It’s capped at 4 challenges per game- 2 for each team. Even at 30 seconds per review, that’s 2 minutes.
There could be some interesting strategy to it- Do you challenge a call early? Hold them until the end? Only with men on base? I’m sure there are cases where 1 strike/ball call has a huge sway in the outcome, but, I’m sure it would take a few years for teams to figure the best scenarios out.
To me it seems like sort of a tenure thing. Saw Muncy do it a couple times already and the pitches were way off. Like I would venture some rookie isn’t going to do it in the 1st inning, and most people won’t do it on a 2-0 count, but almost guarantee that the better players will have free reign to do it whenever they want.
Yes. This. I mentioned it in another comment before I saw yours. There will have to be team meetings to strategize whether someone will challenge a call or not, but it’s hard to quantify it in such a way that will be obvious to the player at a single point in time.
I think the hitter or pitcher should indicate if they feel a call was bad enough to be worth a challenge, but it’s up to the manager to actually initiate the challenge.
I thought it was 2 unsuccessful challenges per game (in other words, if you make a successful challenge you retain it).
It will speed up games because irate managers won’t be running out of the dugout screaming after a pitch in the dirt is called a strike. LOL. I watched two AAA games in Worcester last season and the Hawkeye came up almost instantly, like tennis, and the pace of play barely changed. There were 3 challenges in each game and all were successful.
but irate managers running out of the dugout screaming after a pitch in the dirt is called a strike is fun and a worthwhile reason to keep the game longer
The ones I have seen so far have taken less than 10 seconds each I would say.
I have no idea why they insist on implementing all this stuff with a secondary set of challenge rules that amounts to some kind of resource-management minigame instead of just trying to help the umps get the calls right.
I’m not sure if I care whether the game selects human umps or robo umps, but man, I abhor challenges. I think they’re a significant negative when it comes to excitement and viewing experience.
I feel like just telling the ump in real time every pitch if it’s a ball or strike? Why not just do that? Umps still get paid, no bad calls. Or you could even create a shadow zone if you want to leave some human element. Balls that possibly nick plate and inch off the ump decides.
I just don’t see why baseball would not want accurate and consistent ball and strike calls. I am not the least bit interested in continuing the “historic part” of the game when all that means is we get to watch the next generation of Eric Gregg’s and Angel Hernandez’s mangle the strike zone. How many here watched Game 5 of the 1997 NLCS? This game is so infamous that it is called The Eric Gregg Game. Everybody who is wondering whether ABS should be instituted should find the game on You Tube and just see why I am so strongly in favor of taking balls and strikes away from human eyes.
In the abstract, sure, of course – everyone wants more consistent and accurate calls.
The questions for me:
1. What threshold of accuracy and consistency is enough?
And
2. What are we giving up in entertainment experience to get there?
I vividly remember this game. The last strike to McGriff was more than a foot off the plate. This game almost broke me as a person. If for no other reason, the injustice of that day will always put me on the side of robo-umps.
Did he ever give an explanation as to why he called that game that way?
In any case, this is Manfred’s MO. Champion rules changes, ball changes, strike zone call changes, all sort of other kinds of changes, and then everyone else has to deal with the unintended consequences. Who knows if this is the time when it blows up enough to finally embarrass the league, but I’m guessing the bar for embarrassment is pretty high.
I don’t think there’s much risk here. If it’s not working as intended, they can turn it off. Or, they’ll just tell us to get used to it, like with slides into second and third base.
Personally, I’m opposed to using any video review for purist reasons, but I don’t foresee any danger to the sport or Manfred’s “legacy.”
Eh in comparison to the other changes this one feels pretty small. Although it has actually been sort of cool to see inside players minds. On back to back pitches Muncy challenged ( it was 5 inches high) and the catcher challenged (it was 5 inches off the plate). Was funny to me to see a catcher think a ball that far off the plate was a strike.
They should measure it at the front, middle, and back of the zone. Otherwise the ABS should not be used. A curveball that clips the top of the zone in front is going to be found to be an inch or more low. Similar issues could occur horizontally or vertically for other pitches.
Just front and back should be sufficient
“Ridiculous gelatinous orbs” was RIGHT THERE!
The two main problems with replay in any sport are that it slows down the game and that it works from the assumptions that perfection is an attainable and desirable goal, which is untrue.
We have all seen great acts of athleticism nullified because a ball, toe, or elbow touched the wrong blade of grass, grainline of hardwood, molecule of clay, or crystal of ice. We watch sports to be entertained. That’s not entertaining and waiting to find out if you have been entertained is even less entertaining.
Gamblers watch sports for other reasons, and have more pull than you might think.
Indeed, it’s always seemed the drive to get every call correct every time is the need/desire of those for whom the games are a means to an end rather than the end in themselves.
That is an absurd statement given that a single pitch called the wrong way changes an at bat (outcome probabilities) so dramatically. The difference between 3-1 and 2-2 is huge. In any given game you can spot several instances where player X should have K’d or should have BB’d prior to a detrimental outcome according to your rooting interest.
But all kinds of random factors can change outcome probabilities. That’s why they play 162 games.
1000%, so succinctly well-put
It’s about consistency, not perfection.
I’m with you. When a runner is called out on review for imperceptibly (in real time) coming off second base, the entire game is ruined for me. It replaces the natural tension of the sport with the feeling of waiting in line at the DMV. I can’t tolerate it. That’s a me thing, and it’s resulted in me watching like 10 games a year instead of 800. I mostly just watch clips and recaps now.
I didn’t turn on baseball to watch tedious semantics. I want to see tension and excitement. And if you kill that in the seventh inning with semantics, then I feel betrayed and that my time was wasted. Baseball is too long a game for betrayal.
The thing with this is it always goes both ways. I was at the Armando Gallaraga “perfect” game. Everyone in the stadium knew the call was blown. You could see it with your own eyes. Umps make harder calls in their sleep. That was a great act of athleticism and the history books don’t accurately list as such because of that missed call. We could really go down this road forever, and nobody know where the best trade off with replay use and game delays are, but it does appear that the general consensus is most people are willing to give up some time to get things “right/consistent.”
The history books don’t list it, but our memories sure as hell do! It’s seared into our minds in a way that Bud Smith or Edwin Jackson or Dallas Braden could never hope to capture. The name Jim Joyce lives in (tiny) infamy.
That, in itself, is a “win” for the sport imo, however much it sucks for Tigers fans or Gallaraga. They wrote a friggin book together about it.
Exactly. And these ‘history books’ that list perfect games don’t actually exist. Speaking of cutting both ways: what is the most discussed perfect game of the last 25 years? Gallaraga’s.
Sure for now, but all this stuff fades. We will all be dead so who cares, but eventually the “list of perfect games” is all anyone knows from a TV broadcast in 2053, which will simply list them all and none of the quirks. These weird moments only last as long as they are in the public consciousness, the actual list of perfect games the MLB holds as official is just lists 24. That’s it. Plus we are like the 99th percentile of knowledge on this website. Tons of people have forgotten all about it, just like tons of people have forgotten about how Strahan got his record breaking sack and in random TV NFL broadcasts it just puts his name as the season leader and that’s that.
Yea, you could have stopped after the first half of the second sentence. I could give a shit about Rogers Hornsby or Ty Cobb let alone all the players who had a singular moment of glory. Even the handful of fucks I give about Babe Ruth are because of Sandlot and an autograph my dad once traded a BB gun for that may or may not be Ruth’s.
Within reason. Going back to the NCAA basketball example, and I think it applies to baseball also – if you can’t tell a call was definitively wrong within 30 seconds, call stands, move on. I don’t need the refs taking 4 minutes to add 0.2 seconds to the game clock. It is an exercise in tedium and makes the game-watching experience demonstrably and unquestionably worse.
I think a lot of that is dumb process stuff that can and should be improved. When the NFL replay was first rolled out the whole “going to the special replay booth” was a joke. All of these leagues have the money and means to make these calls fast from some dude at HQ somewhere.
Speaking of, has anyone tried to watch the end of an NCAA basketball game lately? I say “try” because I don’t believe anyone can actually sit through that tedium. Checking replay monitors, regular foul not a flagrant, checking monitors, add 0.3 seconds on the clock, checking monitors again…good Christ it’s interminable.
& the other thing with college basketball is so many teams flop like mad (Looking at you Duke & Wisconsin) that they stop the game for flagrant foul checks 3-4x a game.
Seriously, I watched Wisconsin on Sunday & an MSU player put a light elbow in a Wisconsin guys chest for a box out check & the dude fell down like he was shot. Halt the game, check the monitor for 5 minutes to find out it’s nothing.
It stopped a clear MSU fast break on a rebound & they didn’t even make the Wisconsin player leave the game until the next TO (The NFL DOES do that now, if they stop the game for you, you must sit out at least 1 play). Total idiocy, MSU gets shafted out of a fast break, Wisconsin has a risk free way to do so. You’re rewarding guys for acting like they got hit in the head with no penalty for faking.
I’m trying to leave a comment with paragraphs but when I hit return twice it takes me to Shohei Ohtani’s player page.
Same. I had to keep clicking “cancel” when the warning dialog appeared.
Same..when I try to have a space between paragraphs, I go to Ohtani’s page. The man is a menace!
Yep, that is happening! We’ll get it fixed.
Mine goes to Bo Bichette’s! (I guess maybe it goes to your last player page visited?)
I was thinking the one you visit the most? The last pages I had visited before were Jeff Kent and Ryne Sandberg, so it wasn’t that for me.
Practicing their April Fools jokes.
Excellent.
Ok, I’ll cop to it: that picture is bothering me.
The Devo reference in the title, on the other hand, is excellent.
Great article!
Scherzer misses the point, though, when he says we should continue to use flawed humans to judge pitches.
Robo umps aren’t about improving from 95% accurate to 100% accurate (if that were even possible).
It’s about consistency. Rookies and veterans should have the same strike zone. A 3-0 pitch should have the same zone as an 0-2 pitch. The first pitch of the game should have the same zone as the last pitch of a perfect game.
Human umps are either unable or unwilling to have that level of consistency. Some of it is just the inevitable bias of wanting a blowout in misting rain to end. But some is clearly biased umpires, like CB Bucknor and Angel Hernandez. Some is umpires betting on games, like with Pat Hoberg.
They say that, but it is never true. It’s about covering their asses and avoiding the whole “they lost because of a bad call!” news cycle. They don’t like it when umpires make the news.
I admit I’m a cynic about these things.
Hear, hear. Well said. I hate the differing strike zones/rule books for certain favored nations.
I never liked the idea of total Robo strike zone. I’d rather have it be a buzzer in the umpires ear which sounds louder or softer depending on the dwell time of the ball in the 3D strike zone. This should be rather easily do-able given the technology. The umpire does HAVE to call a pitch that barely gets a nip of the zone and can use his judgement with the help of the same tech we have at home.
doesn’t have to… sigh.
I’d rather it be a shock collar with increasing intensity – let’s talk about a vested interest in getting those calls correct!
Couldn’t they use multiple 2D zones to simulate a 3D zone? Like, wouldn’t 5 “slices” eliminate nearly all of the 1% of conflicting takes? If the ball catches any of them, it’s a strike.
I worked for an autonomous trucking company several years back. Commercially available sensor technology (2D and 3D) is astonishingly good and even custom systems are not exorbitantly expensive. Any processing that would need to be done would surely be less involved than real-time 3D-tracking traffic around a semi. I imagine the difficulty is much less in the technology itself than in finding the right combination and setting it up in an unobtrusive way for ballparks.
Does anyone know how they’re physically set up now?
I read the primers while I was still writing, but I don’t recall enough of the details. My understanding is my proposal would require no change to the installed technology beyond some lines of code. That could be wrong though.
“once a computer is making the decisions, the objective of the game becomes slightly less fun, for the same reason that playing chess against the computer isn’t particularly enjoyable”
I just don’t see how this follows at all. The players are still playing against each other. They were never playing against the umpire.
Have you seen a game featuring Angel Hernandez?
The loser was always everyone watching.
Angel Hernandez was an object lesson in the hazards of unions.
How you gonna thumbs down this? You’re so pro-unions that you’re pro-Angel Hernandez? That’s embarrassingly uncritical. Nothing is only good or bad.
Pitch framing involves a sort of competition with the umpire.
Let’s think about this philosophically: pitch framing = deception = cheating. If one accepts that, then why not allow batters to fake the act of being hit by a pitch or fielders to fake the act of catching a ball? If one accepts catcher deception as a legitimate part of baseball, then should we not get rid of the entire video review system? The question is, does the game accept deception, or does it not?
But the point of pitch framing is not deception. The point of pitch framing is giving the most flattering presentation. Catchers don’t always know if a borderline pitch will just inside or just outside the zone and it isn’t their job to know or decide that. People forget that framing is about not losing true strikes as well as stealing balls.
It is kind of like a defense attorney who might not always know for sure if their client is guilty or innocent of every charge. It is their job to give the best possible defense and the jury and judge can make a decision.
… as is making a flattering presentation of a catch that is not a catch. My point is, if one is going to support presenting false outcomes as not false, apply it to all aspects of the game.
More precisely, pitch framing sometimes involves deception, which even when it does, is both within the rules and expected.
Batters do occasionally fake the act of being hit by a pitch. Fielders do occasionally fake catching a ball that they trap.
And just to slide gleefully down your slippery slope, I would very much prefer getting rid of the entire video review system. It’s a game, entertainment, not brain surgery.
And the game certainly does accept deception in its very essence. Pitcher vs batter, fielder vs baserunner, pitcher vs baserunner, all involve deception. Player vs umpire? Why not? It’s fine. Again, it’s a game.
(It appears we are philosophical opposites)
I have no problems with people arguing against the video review system as long as their arguments are consistent. Yours is given that you prefer to keep all decision-making with the onfield rule enforcers. Personally, I prefer to use all available technology to ensure a more equitable outcome.
Cool, glad we can both enjoy the game!
One does not accept
Anything is more accurate than the umps. This is just someone nitpicking a system they can barely understand, yet alone implement. It’s foolish to let “perfect” be the enemy of “much improved”. Of course the dinosaurs like Scherzer are going to complain about it because they complain about anything new, and of course some reporter trying to get chummy with a player is going to say moronic things like “it would have been a strike with a 3D system derrrrrp”.
Why isn’t the 2D zone set at the front of the plate? That could miss some high breaking balls dropping into the zone, but I think people are mostly good with pitches like that being called balls.
This would also somewhat track with checked swings being called by breaking past the front of the plate, so it makes some sense to me.
I think the challenge system is the best way forward rather than full ABS, but I also think that having two different strike zones should send MLB back to the drawing board.
I was a D1 pitcher and umpired to high school level. I think there are 2 possible solutions
1. Keep the “plane of glass” approach and move it forward a couple inches, to halfway through the rectangular part of home plate
2. Do a 3d zone that is about 6″ deep, starting near the front of the plate, and it has to catch the zone for the entire depth to count
The rule book strike zone has never been what is called in practice. When they first started testing this, the window was at the front of the plate. No one liked it when a ball just nicked the bottom of the zone only at the front of the plate. So they moved it back. This seems to have received a better response. But to me, the answer is to have two windows that the ball needs to touch. Perhaps the current one and one five inches in front of it. Would this work to most people’s satisfaction? I don’t know. But things like this need to be tested. Just like where exactly are the top and bottoms of the zone. Once a standard is set, everyone will get used to it and it will be fine.
To me the egregious ball or strike call that directly effects the outcome of a game is at the top of my very short list of things I don’t like about baseball. “I “don’t” like it when somebody screws up and somebody gets screwed over” defines my position exactly. Why, if accuracy is valued in everything in life, would an easy to adapt method not be utilized to improve balls and strikes? I don’t care if it is a 2-D or 3-D method as long as something is implemented. Why should plays in the field be subject to review analysis but not the thing that is the basis for all of baseball? One of Scherzer’s comments struck a chord. During spring training it seems that it would make a great deal of sense if there were a lot of challenges available so both pitchers and hitters would be able to test the parameters and get a feel for the system.
“Why, if accuracy is valued in everything in life, would an easy to adapt method not be utilized to improve balls and strikes?” I disagree with that premise. Accuracy is and should be valued where it is crucial, like in science, medicine, engineering, law, etc. In games and entertainment, it can be important, but not often the most essential thing.
There is no reason why the TV folks should not show the full 3D shape of the strikezone and the trajectory of the ball. A graduate in computer science should be able to write the code within one day. (Or maybe a project for a first-semester CS course.) Sure, you might want to reduce the zone to a block for simplicity, but even then, it would be a step up from the 2d image shown currently.
“…shaped like an infinite number of infinitely thin home plates. Honestly, I don’t know how any pitcher who’s had it fully explained to them will avoid succumbing to paralysis halfway through their windup and toppling off the mound”
Ha! Time add those calculus classes to spring training.
Here’s hoping that pic of Max doesn’t stay on the home page very long – it is creepy!
Just FYI Davy, I was patiently waiting for a Dire Straits “So Far Away” reference that never came.
Same!
Song outro as long as the song, the good version – “So Far Away”
Song outro as long as the song, the less-good version – “Hey Jude”
This is why the strike zone should be automated: when you’re watching a game, and a human umpire messed up, how many times was that pretty obvious to the viewer? And how many times have you looked at an automated strike zone result and it was obvious to you that it was wrong? I would say a LOT for the first scenario, and zero for the second one. If the automated zone gets it wrong, it is essentially wrong by rounding error. It a human gets it wrong, it can be SUBSTANTIALLY wrong. To me, this is the key.
As for how long the challenge system takes, and whether a manager gets pissed at a pitcher, catcher, or hitter for wasting a challenge at a non-strategic point in the game, that’s another topic of discussion, at least until the strike zone is fully automated with no challenges.
I may be a persnickety quibbler — okay, I’m definitely a persnickety quibbler — but I constantly notice problems with the on-screen strike zones on the various TV broadcasts while I’m watching the game, if that’s what you mean. The ABS system and the “Fox box” aren’t actually backed by the same technology, but since no one is actually seeing the ABS zone, I think you must be talking about the one on TV.
Yes, I am talking about the ones on TV. I can’t recall A SINGLE instance of when I saw something that shows up as a strike in the box on the telecast and thought “no, that’s clearly WRONG”. But for human umpire calls, it happens A LOT.
I am not saying the electronic strike zone is 100% infallible, but it’s FAR better than what we have now with human umpires that make GLARING errors. THOSE are what we should be eliminating. Whether a pitch is in our out of the zone by 0.5″ doesn’t really bother me that much. There’s always going to be margin of error. But that margin shouldn’t be 4″ or more, where we’ve seen human umpires screw up by that magnitude A LOT.
It’s calls like THESE we are trying to fix
https://www.youtube.com/watch?v=8zbWxE1GAWA
It seems like there will be a time where
1. The umpire makes a correct call, given the rule book 3d zone.
2. A batter challenges because it’s a very consequential call.
3. The ball is not a strike according to the 2d zone.
4. The robo ump incorrectly overturns the call.
Even if the net effect is an increase an accuracy, the one time out of one hundred that this does happen will be a firestorm. If it happens early enough in the season, there’s a chance the firestorm sucks out support for the challenge system and ABS. It would be like if, the first time someone used replay for a play at the plate, a correct call was overturned.
MLB needs to make sure this does not happen. The best way would be to have the robo umpire call a 3d zone, so there aren’t two conflicting zones. Another way (if this isn’t feasible) would be to identify which pitches ABS tends to get wrong and apply a ‘call stands due to a lack of clear and convincing evidence’ standard for pitches that might be in the 3d strike zone, like with replay.
Agree with several others here in the comments. We can’t have two different strike zones. The strike zone has always been 3D. MLB should NOT implement a 2D strike zone. Either make the computer make the calls using the same 3D strike zone or drop the computer trial altogether.
Article aside. I would like to read it but the picture is offensive.
My daughter has heterochromia. It’s not a big deal and hers is more subtle than Max’s but making fun of someone because of a medical condition is not appropriate.
You have clearly misunderstood what the picture is doing. The article is about robo umps, and Scherzer is a key person quoted.
If Scherzer weren’t in the article at all, I might agree with you, but this is clearly a nod to Terminator, not his heterochromia.
Corrections to physics & mathematics presented in the article:
1) It is a Magnus effect, not a Magnus force. The force comes from the collisions of the ball with the air.
2) The source of the drop of the ball to the ground is the Earth, not gravity. (This one is excused because so many, even physics Ph.D.s use this incorrect terminology. Although, this one is not as bad as the fictitious phrase microgravity.)
3) Pythagoras was not the source of our definition of trigonometry in Western culture. Hipparchus is viewed as the one who created tables of arc and chord values as a function of the angle.