Learning large algsets vs learning songs

[Sue Doenim]

As an avid cuber and pianist, I have often wondered about learning algsets in comparison to learning songs. Both of of these have the two same basic steps: committing to muscle memory, and getting up to speed. Both are similar in that they will have repetitive parts, such as semi-algorithms (R U R' U' and such) and recurring melodies. For the sake of this comparison, I will be taking ZBLL, a well known, very large algset, and Fantasie-Impromptu (you should probably listen to it), a very fast, relatively well known piano song by Chopin, and comparing the two. I will be setting a single piano note equal to a single turn of the cube. ZBLL, with 493 algorithms and an average movecount of 12.08 (optimal), has 5,955 turns in the entire algset. Fantasie-Impromptu has a total of 2,970 notes. This, to me, is proof that learning similarly large algsets is a massive undertaking. For more comparisons, PLL is 248 moves, OLL is 553, COLL is 411 (optimal). I couldn't find any more relevant algsets with movecounts, and finding the number of notes in a song is rather time-consuming. I might do more later.

 

[AlphaSheep]

Learning the algorithms isn't quite like learning a song because a song always flows from start to finish in the same order. For algorithms, committing the moves to muscle memory is only a small part. The much bigger part is associating a pattern of colours with an algorithm, and then being able to pull the correct algorithm out of the hundreds you know.

A better analogy with piano would be learning the first 3 or 4 bars of a few hundred songs, and then being able to play them immediately when someone says the name of the song.

 

[tx789]

There is also the fact that algs are preformed as fast as possible.