What problem does this paper attempt to address?

The problem that this paper attempts to solve is: **What is the minimum number of random moves required to fully scramble an already - solved standard 3x3x3 Rubik's cube?** Specifically, the scrambling process of the Rubik's cube can be regarded as a random walk on a group containing approximately 43 quintillion elements. Regarding this random walk as a Markov chain, its mixing time determines the time required to go from the initial solved state to a nearly uniform distribution. Previous research has determined that the minimum number of steps required to solve a Rubik's cube in any configuration is 20 steps (known as the "God's Number"), but the question of the minimum number of steps required to scramble the Rubik's cube remains unanswered. In this paper, the authors used a supercomputer to prove that **at least 26 random moves are required to fully scramble a solved Rubik's cube**. In other words, 25 or fewer random moves are not sufficient to ensure that the probability of each configuration appearing is approximately equal. This conclusion provides the first non - trivial lower bound. ### Formula Summary 1. **Total Variation Distance (TV)**: \[ \text{TV}(X_n, X_\infty) = \frac{1}{2} \sum_{x \in X} |P(X_n = x) - P(X_\infty = x)| \] where \( X_n \) is the state at the \( n\) - th step and \( X_\infty \) is the stationary distribution. 2. **Definition of Mixing Time**: \[ t_{\text{mix}}(\epsilon) = \min \left\{ n : \text{TV}(X_n, X_\infty) \leq \epsilon \right\} \] Usually, \( \epsilon=\frac{1}{4}\), that is, \( t_{\text{mix}} = t_{\text{mix}}\left(\frac{1}{4}\right)\). 3. **Distance Function \( d_o(x)\)**: \[ d_o(x) \in \{0, 1,\ldots, 20\} \] which represents the shortest number of steps from state \( x\) to the solved state (the origin). ### Main Results Through a large number of samples and calculations, the authors proved that: \[ t_{\text{mix}} \geq 26 \] That is, at least 26 random moves are required to fully scramble the Rubik's cube. This result is achieved by estimating the value of \( \text{TV}(Y_m, Y_\infty)\), where \( Y = d_o(X)\), and a large number of samples were generated using the Big Red 200 supercomputer for verification. ### Method Overview - **Markov Chain Modeling**: The scrambling process of the Rubik's cube is modeled as a Markov chain. - **Analysis of Distance Function**: The convergence is studied by analyzing the distance function \( d_o(X)\) from the solved state to other states. - **Supercomputer - Assisted**: Powerful computing resources are used to generate a large number of samples to estimate the total variation distance and determine the lower bound of the mixing time. This research result is not only of theoretical significance but also has an important impact on the practical applications in the speed - cubing community.

Rubik's Cube Scrambling Requires at Least 26 Random Moves

What the small Rubik’s cube taught me about data structures, information theory, and randomisation

Probabilistic estimates of the diameters of the Rubik's Cube groups

Solving Rubik’s cube via quantum mechanics and deep reinforcement learning

Solving a Rubik's Cube Using its Local Graph Structure

Solving Rubik's Cube Without Tricky Sampling

Casting Rubik’s Group into a Unitary Representation for Reinforcement Learning

Solution Numbers for Eight Blocks to Madness Puzzle

The Complexities of Random-Turn Hex, Square, and Triangle Games

The Mutando of Insanity

Entropic Scrambling Complexities

Rubik Tables and Object Rearrangement

Analysis of the picture cube puzzle

Mixing Time of the Rudvalis Shuffle

Mixing time of the Card-Cyclic-to-Random shuffle

On Algorithms for Solving the Rubik's Cube

Entanglement, Quantum Randomness, and Complexity Beyond Scrambling

Solving the Rubik’s cube with deep reinforcement learning and search

Hitting time mixing for the random transposition walk

Cube Ramsey numbers are polynomial

Towards Learning Rubik's Cube with N-tuple-based Reinforcement Learning