How to break Magic the Gathering.
Stand-up Maths
0:00 Are you ready for the biggest number I believe has ever been used in a game?
0:04 And the way it's generated, I think is like super interesting.
0:07 It happens to come out of the game of Magic the Gathering,
0:11 which is why I have all these cards around me.
0:13 Now, I don't play Magic.
0:15 Um, friends of mine do, who are massive dorks,
0:17 and they assure me, and they're in the room right now.
0:20 They assure me it's a lot of fun, and I get it.
0:23 There's lots of cards.
0:24 The cards interact, which is how we get this ridiculous number.
0:27 You got some uh like creatures.
0:29 So, here I've got a worm spelled incorrectly.
0:31 And then you got spells you can play that change things
0:35 in the game because it's like a two-player game against someone else.
0:38 Uh strangling soot.
0:39 Don't know what that is.
0:43 Uh and Armageddon, destroy all lands.
0:47 There you are.
0:49 I'll play that.
0:51 I regret calling them dorks.
1:00 This video brought to you by Jane
1:04 Street and their brand new neural network puzzle.
1:07 Details later, we're going to need somewhere new to film.
1:10 Right, new location.
1:11 Now, the thing about Magic the Gathering is it's
1:14 been around for now over a third of a century.
1:17 And if you're thinking that's not right, Matt, it came out in 1993.
1:21 I have some terrible news about 1993 for you.
1:24 But the point is it started well it was
1:27 the brainchild of someone with a PhD in mathematics and it's
1:31 only got more complicated since and over the three
1:34 decades since then every possible edge case has come up.
1:38 There are now almost 30,000 different cards.
1:41 And so the way they interact has become just an unwieldy level of complex.
1:47 And for that reason,
1:49 they've had to introduce some rules to stop things from getting too out of hand.
1:55 Because a lot of the cards involve calculations and numbers,
1:58 regulation 107.1 states that only whole numbers are allowed, only integers.
2:04 Let's try and keep this nice and simple.
2:09 If it's something you can't calculate using integers,
2:12 well, regulation 107.2 says, tough.
2:14 If you can't calculate it, you get a zero.
2:18 We're keeping it neat.
2:19 Now, once you get to loads of cars interacting, you think, well, hang on.
2:23 What if there's like a runaway feedback loop?
2:26 Well, regulation 104.4B says no to that.
2:29 If you've got some kind of, you know, feedback cycle,
2:34 which is going to carry on infinitely long, no deal.
2:39 The game is officially a draw.
2:41 And that's assuming it's a cycle where no player can stop it.
2:45 If a player can stop it, then that's fine.
2:47 and they can choose some some arbitrary level at which at which they quit.
2:52 And so within those rules, in theory,
2:54 you can only play moves which will give you a manageable end result.
2:59 And that worked.
3:00 Everything was nice and manageable for a very long
3:03 time until someone found a loophole somewhere above big numbers,
3:07 but before infinity.
3:09 Our ridiculous combination involves these three cards which we
3:14 have managed to get our hands on the original.
3:18 However, for us non-magic the gathering people, h they are dense.
3:22 There's a lot of like wonderful artwork.
3:24 There's a lot of information on these.
3:26 And so what we've done is we've made three simplified standup maths
3:31 versions and we're going to use those to explain how it all works.
3:36 Starting uh with this one, card A, double it.
3:40 Now, this card doubles the number of tokens that are put into play.
3:44 And the one nuance here is you'll have like
3:47 your original creature cards or like you know, spell cards.
3:50 And you can also have copies of them that are called tokens.
3:54 And the game deals with them subtly differently.
3:56 The original and all the token copies.
3:58 This is if you're putting in token copies of something, they get doubled.
4:02 Now, we'll get to card B in a moment because before you play B,
4:05 you have to have card C in action.
4:06 And card C says once you do play B,
4:09 you're going to get another token copy of it appearing.
4:12 And once you've got A and C ready, our ridiculous loop is set to kick off.
4:18 Now, it's time for card B, which in the original is called Astral Dragon.
4:23 We've just called make two because when you
4:26 play this, you make two token copies of A.
4:29 So, in theory, we need two copies of this to now enter the game.
4:34 But because they're tokens entering the game,
4:36 card A's power is to take entering tokens and double them.
4:39 So, we'd actually get four.
4:41 So, just by playing this one card,
4:43 we should get four token copies of A appearing as if by magic, the gathering.
4:50 Now, to get those copies, well, okay,
4:53 this card over here, I know I'm going a little off.
4:56 Um, this mystical teachings card says I can search search your library.
5:04 So, I'm going to use that to head to the library.
5:12 So, at my local library, I was able to make some extra photo copies
5:16 of card A, which if I now cut them apart, gives us four token.
5:22 And you can tell they're token cuz they're now grayscale copies of of A.
5:28 So there we are.
5:29 We can put them all into play.
5:30 I'll just put them down here.
5:32 We're now going to resolve the C issue,
5:34 which is by playing B, we also get an extra token copy.
5:38 And I've got a couple down here.
5:40 I've made these a little smaller.
5:41 You'll see why in a moment.
5:42 So we put one token copy in of B because of C.
5:46 But look at this.
5:47 There's five A's over here.
5:48 And they all see this happening.
5:50 And each one of them, well,
5:52 the first one here decides that it's going to double this one coming in.
5:56 Okay, so now we got two of those.
5:58 And then the next one sees both of these coming in, so it doubles those.
6:02 And now we've got four of them.
6:03 And then the next one sees these four and doubles them.
6:06 Then we get eight.
6:07 And well, there's five of them.
6:08 So we have to double it five times.
6:10 So we get two to the power of five token copies of B.
6:16 There are now 32 extra B's in play.
6:19 We've just put in 32 newbies.
6:21 And every single one of these tokens says, "When this enters play,
6:24 put two token copies of A in." So, I need to add another two token copies of A.
6:29 But guess what's over here?
6:30 Already five A's.
6:31 And they're going to do the same thing
6:33 to the incoming two A's I did to the incoming 1B.
6:37 Now, a single B got doubled five times.
6:39 We had 32 of them.
6:41 If we have two A's coming in, it's going
6:44 to get doubled five times, which would be 64.
6:49 So, here we have this is 64 A's, which I'm now going to add in to our A pile.
6:58 Oh, no.
6:59 Because if I put these in, we now have a total of 69 A's over here.
7:06 And we've only deployed one.
7:08 We've done that one.
7:10 We've done one of our 32 Bs.
7:13 So now the next the next B comes in.
7:18 It's going to make two new A's and they're
7:22 going to be seen by 69 pre-existing A's.
7:26 So it's the two is going to get doubled 69 times.
7:31 Wow.
7:32 We're about we're going to have to add two to the 70 A's.
7:39 And here we go.
7:42 So, I have put these um we're now running 64 to a sheet and we have
7:47 an additional 2 to the 70 copies of A plus the 69 that were already here.
7:53 How many do we have?
7:54 Could we work it out?
7:56 I've worked it out.
7:57 Is that many?
7:59 Look at that.
8:00 It's a number with 22 digits.
8:02 Oh, little confession.
8:04 I didn't photocopy them all cuz if we wanted
8:06 to have This is basically 10 to the 21.
8:09 And if we wanted that many card A's at this scale,
8:12 the amount of paper it would take would fill
8:15 more than one entire shipping container with just paper.
8:20 In fact, it would fill multiple shipping containers.
8:24 It would fill enough shipping containers
8:27 to fill 1 million shipping container tankers.
8:30 Uh, that's a lot of paper.
8:32 We'll talk more about outrageously big numbers in a moment.
8:35 But just to complete this journey, we're going to keep going through B's.
8:39 And each time we go through B's, we get a number of additional A's equal to 2
8:44 to the power of the number we have already+ one.
8:47 So our sequence is each next term in the sequence is the previous
8:51 term plus 2 to the power of the previous term plus one.
8:55 And we should do that well through all 32 of the new bees that showed up.
9:03 And so the sequence goes five 69 that a lot and then a lot a lot a lot a lot.
9:11 You get the idea.
9:13 Now to deal with the elephantized number in the room.
9:17 What does a 22digit number even mean?
9:20 And a lot of people when I said a big number in a game might
9:23 be thinking about the classic 52 factorial because
9:25 if you take a deck of 52 cards, you give them a decent shuffle.
9:30 There's 52 factorial possible arrangements of this deck.
9:33 But here's the thing about that number and the card number.
9:37 We could write them down.
9:38 I mean, I have written that one down.
9:41 You could write down 52 factorial.
9:43 But the next step up of this, could you write that down?
9:47 because it's going to h well in binary it'll have
9:50 that many digits because that's the rate at which this is growing
9:55 and in base 10 we just divide it by it's about
9:59 a third going from base 2 to base 10 so it
10:01 would have a third of that many digits in base 10
10:04 the next one up and there's no way we can write
10:07 that down the one after that absolutely not the number of atoms
10:12 in the knowing universe we believe is about this many Now,
10:18 um, I've made that number up.
10:20 That's 81 random digits.
10:21 It's about 10 to the 80.
10:24 So, it could be this.
10:25 The exact number it happens to be right now will look exactly like this.
10:30 And anything above that, there aren't enough atoms in the universe,
10:34 one per digit, to write them down.
10:37 And this very quickly exceeds a power bigger
10:40 than the number of atoms in the universe.
10:43 So, the order of magnitude is just nuts.
10:46 In fact, if you put this whole recursion thing,
10:48 I mean, we tried it in Excel, that crashes real fast.
10:51 We put it into Python, that's as far as it gets.
10:54 We then use something called hypercal.
10:56 And if you do continue the number of iterations,
10:59 you can represent the number as this, a power tower that's 30 high.
11:04 So, what we have here, a power tower.
11:06 You start at the top, you work your way down.
11:09 At the top, we have 10^ 3.6* 10 20.
11:11 And we could write all those digits out.
11:13 There's a lot.
11:14 um if we could even work them out.
11:16 And then the next one down.
11:18 So you calculate each layer as you go down.
11:21 There are 30 of these.
11:22 As you go down each one,
11:23 the next one has as many digits as the value of the one above.
11:29 And it's nuts how fast you you run out of universe to write this in.
11:36 Like this is just an insanely large but finite and very specific number.
11:42 So, what's amazing about playing this in Magic the Gathering is
11:47 that you are technically playing an exact number of Astral Dragons.
11:50 And I reckon if you got that many, I mean, you might win this game.
11:55 And you can argue it's not infinitely large.
11:59 It's finite.
12:00 And it's an exact an exact value.
12:03 Now, this is a lot smaller than Graham's number.
12:08 Graham's number is just it's like iterative power towers
12:12 or power towers although it was just a threshold.
12:15 It was just like an upper bound below which the value to the problem must exist.
12:20 And even though this is vanishingly smaller
12:22 for some definition of vanishing and smaller,
12:25 it's an exact value that exactly solves this problem.
12:28 So, in that regard, this could arguably be the biggest
12:32 number that's ever actually been used by humans in this game.
12:36 It just so happens that it's so big, we're unable to compute it.
12:41 Oh, wait a minute.
12:43 You remember rule 107.2.
12:45 If you can't calculate a number, it equals zero.
12:50 You can't calculate this number.
12:52 It exists.
12:53 It's it's exists as much as the number seven does.
12:57 And compared to infinity, it's about as big as seven.
13:01 But as humans, we can't compute that.
13:03 So if you did ever play the shenanigans in an actual game,
13:07 and there's only three cards, easy to achieve.
13:09 We found them all.
13:10 Like some of them are a little obscure,
13:13 but these are all absolutely attainable cards.
13:15 If you somehow contrived to play this in a game,
13:18 depends if if the adjudicators will let you get away with this.
13:23 If they insist you need to be able to calculate
13:27 the integer value other than just proving the integer exists,
13:31 you've now got zero dragons.
13:34 And that's the smallest possible number of dragons.
13:38 Now, if you've enjoyed following along with Oh, this ridiculous.
13:42 My goodness, you might be the perfect person for what the sponsor of this video,
13:47 Jane Street, has in store for us.
13:49 Because speaking of shuffling things,
13:52 they have a machine learning neural network puzzle
13:55 which is going to be explained by exploded matt.
13:59 If you're not familiar with Jane Street,
14:02 they are a research-driven trading firm where curious people work on deep
14:06 problems and they believe that deep
14:09 learning is the future of quantitative trading.
14:12 and they have an in-house machine learning team who work
14:16 on neural network models which help drive their trading strategies.
14:21 They also have to build the infrastructure required
14:24 to make those inferences and continue that research.
14:27 If you're at all interested in this kind of neural network research,
14:31 Jane Street's machine learning team have put together a neural network puzzle.
14:35 They trained a neural network with 96 layers and then they
14:39 shuffled them and you got to put them back in order.
14:42 And you know, forget 52 factorial.
14:44 To my naive understanding, this is like 96 factorial.
14:48 That's not a hint.
14:49 I mean, I had a look at it.
14:52 I couldn't solve this.
14:53 But some people watching the video probably can.
14:55 If you do think you've solved it, please send your answer into Jane Street.
14:59 I would love to see how many viewers
15:02 of Standup Math videos can solve this neural network challenge.
15:06 Anyone can do the puzzle.
15:08 You don't have to care at all about a career in finance.
15:11 But if you are curious about Jane Street
15:13 and the work they do with machine learning,
15:15 I will have a link below and there's like a QR code somewhere on the screen.
15:20 You can check out their open roles and available programs.
15:23 That's the video.
15:23 Thank you so much for watching.
15:25 I really appreciate it.
15:26 I also want to thank Tabitha Grove who designed these phenomenal cards.
15:29 And if you support me on Patreon, we're going to put some extra content up
15:34 on there about how Tabitha designed uh these fantastical artworks.
15:38 And uh big thanks to Matthew Franklin who is the viewer
15:42 who wrote in and mentioned this ridiculous thing in Magic the Gathering.
15:47 And if you know any other ridiculous maths things in games,
15:51 gathering related or otherwise, please do let me know.
15:54 And um yeah, we'll also link to the original Reddit post and whatnot um below.
16:00 Uh that's it.
16:01 That's the whole video.
16:03 Um if you're wondering why the table's shaking slightly,
16:06 Skyabad the dog has come through.
16:08 Skye, over here, over here.
16:09 Over here, and around.
16:11 You got to say goodbye.
16:13 Over Sky, here.
16:15 And up.
16:15 And goodbye.