The Rockies Believe They Have an Unbreakable Code
PITTSBURGH — For as long a there have been coded messages designed to secretly pass information before prying eyes, there has been someone trying to break the encryption, from the Babington Plot to the Zimmerman Letter. For as long as there have been signals in baseball, there has been an opponent trying to identify a pattern and steal the signs. And with every game televised, with cameras everywhere, teams have never before been more paranoid about protecting their messages.
Complicating matters is the commissioner’s concern about pace of play, which has manifested itself this season in the form of a limit on mound visits. Now a pitching coach’s capacity to deliver a message directly is even more constrained. Pitching clocks might be on the horizon. The need for signals is even greater.
In the face of all this, at least one club appears has responded with their own innovation.
Last Sunday, the Washington Nationals broadcast noticed an unusual card sheathed in clear plastic on a wristband that was adorning the left arm of Rockies catcher Chris Iannetta. The MASN cameras zoomed in for a close-up in an attempt to satisfy the curiosity of color man F.P. Santangelo and to discern the contents of the card.
This author went into investigative mode, paused the television, pulled up the game on my laptop via MLB.TV, and took a screenshot of the image.

Attempts to unlock the code via crowdsourcing on social media were unsuccessful.
Is the cipher unbreakable? The Rockies think so.
The wrist card wasn’t just something the Rockies implemented Sunday. Iannetta has worn the card since the Rockies’ season began in Arizona.

And it’s not something employed just by Iannetta: Tony Wolters uses a cheat sheet similar to what we see on the wrists of quarterbacks.

The Rockies were in Pittsburgh this week, so I traveled down the turnpike from Cleveland to investigate.
While other teams might have already done something similar, this wristband — or, at least the scope of it — seemed to be unusual in nature. And with the amount of information available in today’s game, when it’s possible to know every batter’s performance against every pitch type in every count, I have wondered if teams would try and get more information to catchers and players on the field.
For example: Mariners’ outfielders are carrying cards in their back pockets to help with positioning this season. In college baseball, the coaches — lacking the trust in their young battery — often call the vast majority of pitches, which can slow things down.
So what’s on the card?
Iannetta said the information is mostly related to controlling the running game, and he explained some of the mechanics of the process.
“It’s just a random three-digit number that corresponds to a sign and then we have 10 different cards with random numbers,” Iannetta said. “As soon as they [the MASN broadcast] zoomed in… we heard about it and switched cards immediately. We switched to a different card with a whole new set of numbers. There’s no way to memorize it. There’s a random-number generator spitting out a corresponding number [for the cards], and the coaches have the same cards.”
In explaining the process, Iannetta said he’ll look toward the dugout see a coach use his fingers to send in the three-digit code and then look on his card for the corresponding call. It could be a throw over to first or nothing, no action. Iannetta said three-digit codes are never repeated in-game for the same call.
“If I get ‘1-4-3,’ and it’s a throw over to first base, we’ll never use ‘1-4-3’ again to throw over,” Iannetta said. “There will never be repetition… It’s pretty impossible to steal signs if you use the system we are using.”
The cards and coding system are the ideas of Rockies bench coach and former major-league catcher Mike Redmond.
Redmond had wanted to implement a better signaling system for years. He was inspired by watching college games, noting how often teams were able to switch their signs and calls when using a similar system. He also wanted to make life easier for catchers, a simpler method so they could focus more on calling a game. He began to implement the system at the beginning of last season when the Rockies had a younger, less experienced starting catcher in Wolters, who had begun his professional career as an infielder.
“I’ve been thinking for years about putting in a system that stops guys from stealing signs and keeps up with the flow of the game,” Redmond said. “With the video and cameras on you all the time, if you don’t get creative and really mix up your signs it could cost you a game. What I wanted to do was eliminate that and make sure we had an easy system that worked well and flowed, and one that nobody could pick up.
“Essentially we took what we do and put it on a card.”
What does “EAT” signify on the card? Redmond laughed and said he could not give away all of his secrets.
Redmond and the analytics team create new cards before every game, tailored to specific opponents, and encrypted with three-digit code from a random-number generator.
“It’s really impossible to steal signs,” Redmond said.
Redmond said when the Rockies added veteran catchers Ryan Hanigan and Jonathan Lucroy last season, they also appreciated the system.
“They realize how easy it is,” Redmond said. “They don’t have to worry about the signs and calling plays.”
The cards are double-sided and flip open. While the Rockies say they are largely used for calls in the run game, there is scouting information on opposing batters. That’s important, he says, when considering the limitation on the number of mound visits.
“As we talk more and more about pace of play, our cards have the ability to have reminders of scouting reports, and how to pitch guys, as a quick guide, so the pitching coach doesn’t have to stop the game and say, ‘Hey, this is how we’re going to pitch this guy,’” Redmond said. “It’s all right there.”
Could the club take it a step further and make it a pitch-call list and begin to call pitches more often from the dugout? After all, to store and recall all the scouting information on an opposing lineup simply through memory is challenging for just about every catcher.
“I still believe the catcher’s job is to call the game,” Redmond said. “To work with the pitcher and have a game plan and call the game and learn that aspect of it.”
For instance, Iannetta notes that the catcher’s vantage point allows for a unique view that is different from the one along the dugout railing. Catchers can better get a sense of how batters take pitches, the conviction of swings, and the quality of the pitcher’s stuff.
“The ultimate thing is, ‘What can they execute?’” Iannetta said. “Ninety-nine times out of 100, I’d rather them throw a pitch they are confident in, that they know they can execute in that moment, than the quote-unquote ‘right pitch.’ You go with what they can execute. If you throw a breaking ball down and away on the black, it’s not going to get hit very often. It’s a pitch-to-pitch thing. If they lose command for two pitches, you have to recognize that and help them get it back. If their best pitch is a slider, and they don’t have a feel or it early in the game, you have to find another way to get outs while throwing [the slider] enough times to get the feel back… That’s just part of the game.”
For Iannetta, there’s a blend of science and art when calling a game.
Moreover, Iannetta said there would also be a pace problem if getting pitch calls relayed from the dugout to Iannetta to the pitcher and then back again if there’s a shake-off.
“If I have to look at a card and give a three-digit signal to a pitcher and he doesn’t want it, looks at his card and wants curveball and gives me that number… it would slow down the game,” Iannetta said.
Rockies manager Bud Black agrees with the Iannetta viewpoint. Asked about potentially expanding the card to call all the pitches, Black began with the qualifier that he appreciates the value in the wealth of data valuable to him. But he motioned toward the blank, white cinder-block wall in the visiting manager’s office at PNC Park and said, for him, each game is a blank canvas. On any given night, the pitcher is going to have a different feel, a different comfort level, with certain pitches that will be unique to that contest.
“We know all those percentages. Travis Sawchik, on 2-2 counts, you hit .050 on the curveball,” Black said. “[Pirates shortstop] Jordy Mercer on 1-0 counts doesn’t hit the changeup. We know all that. But where you run into issues is every pitcher is unique. Every pitcher doesn’t have a great 1-0 change or 2-2 curveball. You have to rely on what the pitcher does, his strengths, and really what is going on that day.
“Each game is unique. As much as you want to draw big conclusions from big data — and, yes, there is something to that — but at a particular moment in that game, big data might not be there.”
Black said he does call some pitches from the dugout, typically in high-leverage situations.
“As far as calling pitches from the bench, do we call some? Yes.” Black said. “Do we call all 150 pitches? No.”
The Rockies believe they may have an unbreakable signaling system, but they seem to respect that even the perfect encrypted message has its limits.
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.
Rockies = innovative…I’m confused
It took till 2017 for MLB to discover QB wristbands? They are pretty sweet…
The wristbands are definitely not new. Teams have been using them for a long time and I remember using them in high school in 2008. The code is different and cool though. It seems like a good system.
If someone from the opposing team got a still shot of this wristband, it would probably be pretty easy to learn what the codes mean. Then you’d just need to know that they are still using a specific code sheet (one that you’ve got a picture of) and you’d know exactly what they are doing.
They never re-use a code sheet so you’d have to get the still shot and break it pretty quickly during the game.
Well, yeah. Considering MASN was able to do that, and they changed the codes mid-game, I think it is possible. And while they change the 3 digit codes daily, are they changing the right side of the codes daily (as in, does “5-1” or “EAT” mean something different each day)? If not, once you find out what those mean, you’d have their exact codes as soon as you got a picture of the catcher’s arm. I honestly have no idea how hard that would be, but again, MASN was able to get one, so it’s definitely possible.
Yes, this is exactly correct, they do still need to keep the card secret, but given the prohibition on use of any aids by the opposing team, that should be OK.
The only concern is that they don’t want to say what EAT is–suggesting there is a potential weakness here. You could go BACK in the televised game and see what the signs they already gave were associated with, but you’d need to have the right angles.
“Redmond and the analytics team create new cards before every game”
“three-digit code from a random-number generator”
Seems pretty secure to me.
They could have somebody block the catcher, Javy Baez style
I would put money on this being breakable – specifically I would bet that the random number generator is academic since the process for creating the cards (and specifically the information the cells in the spreadsheet communicate) is probably not random. My first guess would be that they’ve got a deterministic (though probably complicated) algorithm for creating the spreadsheets, and are merely slapping a code on top of it – if everything was truly scrambled every day the overhead of learning how to decipher it would be just as hard on catchers as learning to call a game in the first place.
Now, is it breakable in a time frame and manner that would be useful in -game, probably not, and the value of breaking it is probably minimal compared to the rules you’d have to break to make use of the information (ex: you’d need to have some video and comms to the dugout that I think is verboten).
I have no idea why people are down voting the possibility that it is breakable. It’s “unbreakability” relies on no one seeing the code sheet. This is obvious because when they heard that MASN had showed a picture of it, they took the extra step of changing sheets mid-game.
Unbreakable does not mean nobody knows it happens or knows what the codes mean…
It means nobody knows what is about to happen.
Because the title of the article is “The Rockies Believe They Have an Unbreakable Code.” And I think posters are reacting to that, because of course it is breakable!
But as Oddball says, this would be a mess to break. It would require getting shots of enough cards to decipher what the letter side of the card is with in-game events over several games, hoping they don’t change the letter side of the card, and then getting a shot in the game of what the current number system is. And then the utility of it is difficult since you’d have to somehow have to get that information out and memorized without the Rockies’ knowing and switching cards.
A better way of thinking about it is that the Rockies have a code where opponents can’t get usable information to their side in time to counter what the Rockies’ are doing.
I think the headline is technically passable but you guys are right too. One-time pads themselves are “unbreakable,” but of course if you have to use yours with a camera and a stadium full of people looking over your shoulder, that’s another story.
If a one-time pad is used only once, it is impossible to decrypt. A ciphertext that was encrypted from a one-time pad could have been literally any plaintext string of the same length. For example, suppose I want to encrypt the words “BEAR” and “CATS” by modular addition along the standard 26-letter English alphabet. If I use “KDVJ” to encrypt “BEAR”, I get:
B + K = 1 + 10 = 11 = L
E + D = 4 + 3 = 7 = H
A + V = 0 + 21 = 21 = V
R + J = 17 + 9 = 26 ~ 0 = A
and the ciphertext would be “LHVA”.
If I use “JHCI” to encrypt “CATS”, I get:
C + J = 2 + 9 = 11 = L
A + H = 0 + 7 = 7 = H
T + C =19 + 2 = 21 = V
S + I = 18 + 8 = 26 ~ 0 = A
and the ciphertext is also “LHVA”.
So if you can only see the ciphertext “LHVA”, there’s no way for you to know if you’re seeing the encrypted version of “BEAR” or “CATS”, or literally any other length-4 alphabetic string, without knowing the original key.
eta: they need to randomize+avoid reusing the positions on the card too. otherwise you can ignore the numbers and look at the positions as a ciphertext, which would get reused, which would allow you to determine which positions on the card mean which plays.
one time pass are completely breakable when one person has the pad on their wrist and there are high power telephoto cameras around. But, otherwise, yes.
The quality of the random numbers is exactly the question — the list of meanings can be fixed as long as the RNG is cryptographically strong.
I’m talking here about attacks that rely either on no “stolen” info (I know their random numbers based on today’s date), or limited (after hearing three from today I can guess the rest). Attacks by getting the whole card are a different game.
You sound like somebody who just learned about the concept of pseudorandom but didn’t really understand it.
Basically they are using one-time pads: https://en.wikipedia.org/wiki/One-time_pad
and the players and coaches have the keys on their wristbands.
You can’t crack them because the pattern is random and never the same. We might learn what “EAT” or “5-1” or whatever means but it doesn’t matter. The codes are different every game and looks like they have several for each just to mix it up.
Read Neal Stephenson’s Cryptonomicon for a long, insanely fun story about cryptography.
Hey Travis, you might want to choke-up on the bat in those 2-2 counts… jus’ sayin’.
“Travis Sawchik, on 2-2 counts, you hit .050 on the curveball.”
somebody needs to work in the batting cage
I was thinking he was giving Travis way too much credit.
Considering that he hits .000 against the four-seam, changeup, and slider I think he’s doing pretty well.
one card…SSS lol
Never played catcher. Love to hear from someone who has!
OK, how about this:
Color indicates the action type (positioning, pitch selection, etc.) so the number points to the action to take. The colors and actions probably don’t change much if at all. The actions included would vary based on the opponent. I’d include a lot of fluff to make tracking harder (guess: all those blue P items are ‘you pick the pitch’).
Guessing: white RPT is repeat what you just did. same for green EAT items. Keeping it simple!
Why only odd numbers? No 5-2 or 5-4 items.
Green items are 2B, 3B or EAT. What scenario, defensively, would a catcher be focusing only on 2B and/or 3B?
The Rockies should consider covering these up to discourage the FP Santangelos out there!
I notice the other night that Gary Sanchez is using one of these as well, so the Yankees seem to be going in this direction too.
Goody! Something on my favorite Baseball Stats site that I can finally make a (hopefully informative) math comment upon!
Signs, stealing signs, and even wiggling your fingers boils down to Information Theory a la Shannon. I think three digits is actually a good choice, really (note, kiddies: do not use this as the basis of your password at home). You basically have ten bits of entropy. Which is quite a lot.
Now, the thing is, you as a coach probably only need to convey three or four bits of signal for those ten bits. That’s a bucket-load of “noise,” or in other words the “coding” that prevents the other side from guessing your signal over a nine inning game. And it’s probably around the right level where the on-field guy can look at the wiggling fingers and look at the wristband and not get confused.
Kudos to Mike Redmond, btw. (I always said that catchers were the smart ones.) But he’s being disingenuous:
“They don’t have to worry about the signs and calling plays.”
Yes they do, Mike. If they are top-end catchers who know their pitchers — yes they do. Stats is all good and well, but these guys are out there on the field, in the moment. The human element is augmented by clever signals: not replaced by them.
They’re “one-time-pads” – a common form of encryption used by spies during the cold war.
edit – I see bluestraggler already made this connection
Every team should use such cards when playing the Red Sox.
You’d still lose 17-4. Hee hee.
(OK, I know, early sample. Indulge me.)
“We know all those percentages.”
Apparently doesn’t know Ian Desmond’s percentages if he’s let him have 81 ABs this season
Wow, this is a fantastic article! Really interesting to see how teams are balancing the old-school and new-school thought processes.
My daughter’s softball team used these at the U11 level. Downvote THAT, I dare ya!
A lot of high school teams are starting to do this with runners. Coach shouts a code and the runners and hitters check their wristbands for the right sign. Not a terrible idea.
P is pitch, RPT is repeat. I’m guessing 5-1 is pick-off to first, given its frequency. H and/or 5-5 could mean hold the ball. 2B/3B could mean check the runners. Maybe there’s also something in there for a quick-pitch or slide step?
As to whether the code is “breakable?” Well, in some ways, it is. To start off with, there are four colors visible: mostly white, some pink, some blue, some green. Pick the color you want to “break,” which probably isn’t boring old white.
And now we come to statistical analysis: specifically, Bayesian. Assuming that the “code” remains the same over all nine innings (thank you Mr Manfred for your truly exquisite attempts to limit visits to the mound), you can still compute (in real time) the “priors,” which is to say the observed probabilities based upon the signs. And you can, of course, further narrow your field by looking for the pitch you want the opposing guy to throw — say, a curve on a 3-2 count with RISP.
Clearly, you haven’t cracked the code. But you’re probably capable of improving the odds.
Jeepers. This is beginning to sound like a 1960’s kid with a Joe Bazooka decoding ring, isn’t it?
Those are my priors, and this is my story.
“Iannetta said three-digit codes are never repeated in-game for the same call.”
Not entirely relevant. Though not entirely irrelevant, either.
First of all, the color scheme clearly means something. And if it doesn’t, then it’s just going to confuse the average baseball player. (And me, for that matter.)
And secondly, and even if we accept Iannetta’s comment at face value: you don’t have to repeat the same three digits every time to offer up a pattern.
Which is the point of the extra 6 bits of entropy. Consider the case where a pitcher throws 50% fast-balls for the “out” pitch, for example. It might be a curve or a slider or a change-up, which means you’ve just lost two bits of entropy. But let’s stick with the fastball. High or low or inside or on the corner, whoops, there goes another two bits of entropy.
There’s a limited amount of room for these signals, and by limited I mean really limited.
Now, if you’re going to tell me that the catcher gets a one-time pad for each batter … I’d be inclined to accept that this is effectively “unbreakable.”
But it’s also a remarkably good way for the pitcher to cross up the catcher, if you ask me.
[Edit]: “In-game” probably involves 120+ pitches. 120 pitches, generally speaking, exhausts the possibilities within the range of three digits.
(F.P. — I’m on this. Just give me a day or two — Serb)
Let me help with 252 and 341!
Hi guys. This is my first ever comment and I hope it helps…
With the way those cards are currently designed, they are 100% CRACKABLE, with one very huge caveat, and that is, an opposing team would need a zoomed-in image of the current wristband card being used by the catcher or of the corresponding card the manager is using in the dugout. Without a zoomed-in photo like the one shown in the article above, the code is 100% UNCRACKABLE. Like many of you, and Iannetta, have already stated, the 3-digit numbers are completely randomized by a computer and the numbers are only used once. Nobody can crack this system without being able to see the current cards. If you were sitting at Coors field, watching Bud Black closely, while inputting into your computer every set of numbers he signals to the catcher, neither you, nor your computer, no matter how good that computer is, would be able to use those numbers to figure out what the Rockies next move would be.
Now, lets say a team has captured a readable image of the opposing teams card(s) there are two ways to crack it. The first way would be to simply watch the number being called from the opposing manager and then match it to the subsequent defensive action. Let me give you an example from the Rockies card in the photo above. When Bud Black signals to his catcher the number 314 he is telling Iannetta to call for the defensive action corresponding to 314 which you can see on the card is “P” thus Iannetta would give his pitcher the “pick-off” sign and the pitcher would then do a pickoff move to first base. After they have used each “defensive action” one time you will own the KEY to their card’s “initialisms” and “abbreviations”. However, most of the Rockies codes pictured above seem rather easy to decode. Here are my guesses…
Purple (runner at 1st):
P = Pickoff
H = Hold Pitch
HP = Hold Pickoff
HSO = Hold step-off
NL = No Look Pitch
LG = Long Hold Pitch
5-5 = Pitchout
Pink/Red (bunt defenses):
5-1 = 3rd baseman(5) and the pitcher(1) crash the bunt.
5-3 = 3rd baseman(5) and the 1st baseman(3) crash the bunt.
IM = Inside Move Pickoff to 2b. This is typically done to see if the batter shows that he is going to bunt. Knowing if the batter is going to bunt allows a manager to make a more accurate defensive call.
Green (1st and 3rd defense):
2B = If the runner at first base attempts to steal 2b then the catcher has been told by the manager to make the throw to second base to try and get that runner out. The runner at third base is not the priority and the manager want to try and get an out a second base.
3B = If the runner from 1st base attempts to steal 2b, the catcher will ignore it and try to get the runner at 3rd base out, possibly from getting to far off third base expecting the catcher to throw to second base.
EAT = If the 1st base runner steals 2b, let him. Don’t throw the ball, period. Eat it.
Yellow:
RPT = Repeat the previous action/play
(The Rockies unnecessarily have this on their card. I’d rather see RPT simply replaced with more of the other coded actions such as HP, 5-5, etc)
Lastly, whenever the manager signals in a number that is not on the card the catcher simply calls for the pitcher to make a pitch. And yes, the catcher decides on what pitch to call, or put more accurately, suggest to the pitcher what pitch to throw. The manager only tells the catcher when to pickoff and pitchout, but almost never does he tell the catcher what pitch to call during the inning.
Before I checkout, I want to thank Fangraphs for the great information/content it produces. Good stuff.
Bo
Pretty slick.
People seem to think the only key to crack the code is an image of the card. It’s possible that it’s only one part of the key.
What if the catcher has a predetermined number that he adds to the signal he’s given, so that the number flashed by the manager is not the number the catcher looks at on his card? This predetermined number could be constant through out the game or change every pitch based on the count or which way the wind is blowing.
As an example, the predetermined number could be +1 for the day. So when the manager signals 2-4-2, the catcher looks at 243 (242+1).
Not knowing what this number is (if it exists), would make it pretty much impossible to break.
no, that would be quite easy to crack. only sooner numbers are in the card, get an action and you can see if the number is even on the card. the more complicated it is the more likely you a are to cross up. best to just keep the card secret.
um, only took everyone a full season to realize that they were using this system? They’ve been using it since Black/Redmon took over prior to last season.