Hypercups: Flipping Cups With More Than Two Sides
Micah Dykhuis, Lauren Keough, Sydney Lipton · 2026
Read Hypercups: Flipping Cups With More Than Two Sides by Micah Dykhuis, Lauren Keough, Sydney Lipton (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-27148.txt.
Introduction
Dykhuis, Keough, and Lipton generalise a classic cup-flipping puzzle. In the old trick, cups sit up or down, and flipping a fixed count per move can sort a starting layout. Their hypercups have more than two sides, cycling through several states instead of two. The question is which starting layouts can still be fully sorted, and in how many moves. A 2010 treatment by Siu and Stewart covered the ordinary two-sided case, in an extension published in 2026.
Parity, the even-odd bookkeeping behind the classic trick, no longer decides the answer. Greatest common divisors take its place, linking the side count, flip count, and cup count. Every result from the two-sided treatment carries over to hypercups under that new rule. Ordinary cups reappear as the simplest case. Playful in form, the result is a clean lesson in how number theory governs what moves can achieve.
Catalog record
| CREATOR | Micah Dykhuis, Lauren Keough, Sydney Lipton |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.27148
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Hypercups: Flipping Cups With More Than Two Sides Micah Dykhuis, Lauren Keough, Sydney Lipton Abstract In their 2010 article entitled “How to invert n n cups m m at a time” in Mathematics Today , Man-Keung Siu and Ian Stewart extend a classic trick in which 3 3 cups are flipped 2 2 at a time to any number of cups, flipped any number at a time. Here, we generalize another feature of the problem. A typical cup in our universe, as far as we know, has two “sides”, right side up and upside down. But what if the cup had k k “sides” with k ≥ 2 k\geq 2 ? We call these k − k- hypercups. The classic trick depends on parity. In this article, we show the analogous hypercups trick depends on greatest common divisors. We generalize all of Siu and Stewart’s results to k − k- hypercups. Our results imply the known results for 2 − 2- hypercups, also known as cups. A mathematician has many powers - the powers of deduction, generalization, and problem solving, to name a few. But as Uncle Ben in Spider Man said, “with great power comes great responsibility”. For example, a mathematician must always show their work, and, as we must warn you before we tell you a trick, a mathematician must never use their mathematical knowledge to trick innocent people. Perhaps you have seen this trick on social media. For example, one such video is called “THE IMPOSSIBLE 3 CUP TRICK!! WHY CAN’T I DO IT?? #wow #magic #cool #impossible” [ 2 ] . To start, you need 3 cups (or use our Desmos model at https://tinyurl.com/3cupstrick ). You begin by arranging the cups as in Figure 1 . Figure 1: The initial configuration you use in the 3 cup trick. Figure 1 alt text: Three red solo cups; the first and last are upside down and the center is right side up. Now you can perform the trick on the target. Show them that you can flip 2 cups at a time to get the cups all right side up. Of course, one can do this in one move, but for the sake of trickery, do it in 3 moves, say flipping the first two, then the outside two, and then the first two again. No problem. Now, from the finishing configuration (all cups right side up), flip the middle cup to down, as in Figure 2 and ask your target to flip all the cups to right side up by flipping 2 at a time. 1 1 1 The second author tried this trick with a middle school class early in the morning, figuring that middle schoolers would not be very observant before 8AM. She was wrong. They quickly pointed out the second initial configuration was not the same as the first. Figure 2: The configuration the target starts with in the 3 cup trick. Figure 2 alt text: Three red solo cups; the first and last are right side up and the center is upside down. If you try it for yourself, we guarantee that you will not succeed. Why? Because math! More specifically, from any configuration, if you flip two cups, one of the following three scenarios will occur (a). flip two right side up cups to upside down. (b). two upside down cups to right side up. (c). one right side up cup to down and one upside down cup to right side up. Let’s consider the effect of each of these scenarios on the net change of the number of right side up cups. In scenario (a) the net change is +2 right side up cups, in scenario (b) the net change is -2 , and in scenario (c), the net change is +0 . In the initial configuration in Figure 2 , the number of right side up cups is 2 , and one needs that number to be 3 to finish the task. But one cannot get from 2 to 3 using a combination of +2,-2, and +0 . This is an issue of parity, as 2 is even and 3 is odd. Using mathematicians’ power responsibly, Man-Keung Siu and Ian Stewart generalize the problem (not the trick) to any number of cups, flipped any number at a time [ 6 ] . In their case, start with n cups, all right side up. When can one, flipping precisely m cups at a time, get to all cups upside down, and what is the minimum number of moves to do so? Siu and Stewart show that whether a solution exists depends on the parity of m and n . More specifically, a solution exists if and only if it n is even or m is odd. In other words, the only situation where a solution does not exist is if n is odd and m is even simultaneously. Moreover, they provide the number of moves depending on the parity of m and n , and on the comparison of 2m and n , see Table 1 . Minimum number of moves for inversion. Here \lceil x\rceil is the ceiling function , the smallest integer \geq x . n even, m even n odd, m odd n odd, m even n even, m odd 2m\leq n \lceil\frac{n}{m}\rceil 2\lceil\frac{n-m}{2m}\rceil+1 no solution 2\lceil\frac{n}{2m}\rceil 2m>n 1 if m=n 1 if m=n no solution 2\lceil\frac{n}{2(n-m)}\rceil 3 if m>n 3 if m>n Table 1: Minimum number of moves for 2 -hypercups from [ 6 ] Here, we generalize the problem again, to k- hypercups, an idea first introduced at a talk [ 3 ] . For k- hypercups, instead of right side up and upside down, we have states 0,1,\dots,k-1 for k\geq 2 . Instead of flipped, we say hypercups are rotated. Definition . A k- hypercup is a mathematical object which has possible states 0,1,2,\dots,k-1 . When a hypercup is rotated, its state increases by 1 modulo k . In particular, if a hypercup in state k-1 is rotated, the hypercup returns to state 0 . For the remainder of the paper we assume n,m,k\in\mathbb{N} , with m\leq n , and k\geq 2 . For the (n,m,k) hypercup problem we start with n k- hypercups in state 0 , and rotate m hypercups at a time. Our goal is to end with all hypercups in state k-1 . Definition . The (n,m,k) hypercup problem is solvable when the n k- hypercups, all starting in state 0 , can be rotated m at a time to result in all hypercups having state k-1 . Similarly, the (n,m,k) hypercup problem is solved when s=n(k-1) , where s is the sum of the states of all n hypercups. You can try the (n,m,k) hypercup problem at https://tinyurl.com/HypercupsTrick . Choose different numbers for n , m , and k , and then starting with all n hypercups at state 0 , and rotating m at a time, get all hypercups in state k-1 . For what n,m,k can you complete the task? Spoiler alert - the result is in the next theorem! To begin exploring the problem, we used the NetworkX Python Library [ 4 ] to generate directed graphs