Oh wow this is interesting. I know nothing about cubing but I assume they are all in a specific state of some sort? I am not sure if memory techniques would be used here or not.
That is just impressive. My best time with a 3x3 is 15 seconds, but this dude takes things to a whole nother level.
There are 54 stickers on the cube, but actually 20 movable pieces that we have to deal with.
Each cube has about 20 letters to be memorised (sometimes its 17 or its 25 letters or any number depending on the number of targets for both edges and corners, the number of targets can also be varied by different tracing on the scramble, or doing algorithms from different buffers). The way it is solved is used the concept of commutators, and there has been an algset called 3-style being developed for the last decade, which optimises the 3-cycle commutator for both edges and corners.
There are about 5-6 people who have done 100 or 100+ attempts, but his accuracy is the highest. An Indian cuber and former MBLD WR holder Shivam Bansal, had attempted 100 cubes 2 years back, getting accuracy of 87/100 in just over 3 hours. Another cuber from New Zealand has attempted 200 cubes a year back, and scored 165ish cubes.
In the WCA (World Cube Association), this event is recognised but there is a time limit of 1 hour. He has the current record in it, which he got in a competition (btw this attempt is just a home attempt and not official attempt), where he did 59/60 cubes just under an hour.
The only quality required to be able to pull of these attempts is extreme patience for memorising for such long periods of time without losing focus. There is no concept of savant, or hyper intelligence here. Every human brain is capable of memorising lots of information, given the right techniques, practice and lot of persistence. Also, solving a Rubik’s cube is not that hard, although solving it blindfolded requires some knowledge of commutators and precision.
This event can be compared to a high level classical chess game (classical chess game can last up to 7 hours and are energy sapping and tough) like Wijk Tata Steel , where the chess player has to play the game at the highest level, and even they make 40 moves and draw/win, they are calculating thousands of variations in their mind’s eye. Here in blindsolving, there is no calculation of variations, but we do have to trace the entire cube and make sure parity (2 corner 2 edge swap) is dealt with well. Also we have to use memory techniques to make sure all the relevant information is stored sequentially and can be retrieved sequentially. Making stories or sentences from the letters that we get from the target stickers is one way to go about it.
200 3x3s are there, in this attempt. The score is 189/200 6:21:31 (5:02:xx memo)
Congrats to Graham!
The video will be up on YT soon, which will be sped up, so that we do not have to watch all the 6 hours. Super excited to see this attempt live.
The way multi-blind works is all of the cubes are scrambled with randomly generated scrambles. Each cube scrambled differently. Then you start the timer, look through all the cubes and memorize how you are going to solve all of them (ends up being about 20 letters to memorize per cube), blindfold yourself, and then solve all of them.
I used 4 different memory palaces for this attempt. 2 with 64 rooms each, 1 with 32 rooms, and one with 40 rooms. Each room will contain 1 cube’s memo. I ended up doing something like 8 reviews total of the first 160 cubes, then roughly 6 reviews for the last 40 cubes, solved the last 40 cubes first, then the first 160.
More details are in abdyoyo’s comment.
Video link of the recent 189/200 World Best multi attempt,
Post on the speedsolving forum to gauge the reaction of the cubing community,