• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

BH Tutorial

abunickabhi

Member
Joined
Jan 9, 2014
Messages
6,687
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
Also full 5 style is not a human possibility I believe, I don’t know if you realize how huge those alg counts are and how long it would take for you to learn them. Assuming we want to only learn the edge algs without flips and breaks over the course of our life it’s going to take a while. Let’s calculate how many algs we need to learn a day assuming we start learning the day we are born and we are finish the day we die and that we live an average human life expectancy, 304,128/(365)(79)= ~10.547. So you are learning 10.547 algs per day every day of your life and that’s not even for edge cycles with flips and breaks and we haven’t even touched corners. That’s just not possible, to put that rate of learning algs into perspective you would be able to learn full 1LLL in (~3.9k algs) in just over a year.

Sorry for bumping up this post , but could not find another suitable thread to post this topic on.

If you give this method a quick thought (thought made upto a few days) , it would seem like an unrealistic pipedream when you realise the sheer number of algorithms in it , and that is the reason why I procrastinated for nearly 3 years before starting out on learning 5-cycles (from 2014 to 2016).

The doc had more content about just plain comparison , as I just wanted to address all kinds of scepticism that people might have when they hear about 5 cycles for the first time.

Few people are learning 1LLL and that itself is a massive undertaking, and by your linear scale of improvement , it should take ~1 year.

Your statement considers a linear way of learning algs , which is not the case while learning cube algorithms, generally the learning process accelerates due to common ideas between algs, symmetries and mirrors. So , I am banking on the acceleration in my learning to atleast memorise a size-able chunk of this algset.

At first , this method might always feel like worth to give up on as there is no instant reward for quite a while, but ones you start to invest time and effort into it, the learning process can gain traction , and it would not seen like a mammoth task.

(Also I plan to do the 4-cycles , and the 5-cycles having flipped edge as target in them , after i pan out the Pure edge cycles = 22x20x18x16 = 126720 algs , pure corner cycles = 21x18x15x12 = 68040).

For flipped edges case, there is already cool resources out there, which have multiple algs for multiple cases of the flipped edge.
 
U

Underwatercuber

Guest
Sorry for bumping up this post , but could not find another suitable thread to post this topic on.

If you give this method a quick thought (thought made upto a few days) , it would seem like an unrealistic pipedream when you realise the sheer number of algorithms in it , and that is the reason why I procrastinated for nearly 3 years before starting out on learning 5-cycles (from 2014 to 2016).

The doc had more content about just plain comparison , as I just wanted to address all kinds of scepticism that people might have when they hear about 5 cycles for the first time.

Few people are learning 1LLL and that itself is a massive undertaking, and by your linear scale of improvement , it should take ~1 year.

Your statement considers a linear way of learning algs , which is not the case while learning cube algorithms, generally the learning process accelerates due to common ideas between algs, symmetries and mirrors. So , I am banking on the acceleration in my learning to atleast memorise a size-able chunk of this algset.

At first , this method might always feel like worth to give up on as there is no instant reward for quite a while, but ones you start to invest time and effort into it, the learning process can gain traction , and it would not seen like a mammoth task.

(Also I plan to do the 4-cycles , and the 5-cycles having flipped edge as target in them , after i pan out the Pure edge cycles = 22x20x18x16 = 126720 algs , pure corner cycles = 21x18x15x12 = 68040).

For flipped edges case, there is already cool resources out there, which have multiple algs for multiple cases of the flipped edge.
Even with mirrors and setups it’s still thousands of algs and not humanly possible to ever learn in your lifetime, also while going through the sheet I realized lots of the algs are pretty bad and you will have to generate thousands of algs. You can fail to learn it if you want but it would be far more benificial and humanly possible to learn floating and floating parity
 

abunickabhi

Member
Joined
Jan 9, 2014
Messages
6,687
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
Even with mirrors and setups it’s still thousands of algs and not humanly possible to ever learn in your lifetime, also while going through the sheet I realized lots of the algs are pretty bad and you will have to generate thousands of algs. You can fail to learn it if you want but it would be far more benificial and humanly possible to learn floating and floating parity

Hmmm , I know the set seems humongous , and about half the algorithms pruned by me can be improved upon.

But I do not think it is humanly impossible.

The pattern recognition of our brain is even better than future supercomputers and quantum supercomputers that are going to come up.

If we focus enough , our pattern recognition will be able to digest a method like 5 cycle , in about 10 years at ease , and in 4 years in you really try hard. (It is just a rough estimate)

But , the way you are running your thought process of linearly extrapolating the time taken (10 algs per day or so) , it is not the best way to think about learning.

Again, my opinion will be of no value unless I prove it myself. (I am 2 years into this method , and almost got into the groove (15-20% confidence level))
 

Jezzular

Member
Joined
Feb 23, 2018
Messages
13
The pattern recognition of our brain is even better than future supercomputers and quantum supercomputers that are going to come up.

Thats wrong. Even today there are already lots of applications, where computers are better in pattern recognition than humans. It is estimated that in 10-20 years computers are better in almost every aspect of pattern recognition than humans.
 
U

Underwatercuber

Guest
That's extremely impressive. Do you learn the algs somewhat intuitively, or is it just straight memory? How much time do you/have you put into learning these algs?
As far as I’m aware most 5 style algs aren’t very intuitive so probably straight memory.

I would also like to know how much time you invest into this and also how you get the algs, do you gen them and have a sheet somewhere with them? How optimal are they?
 

abunickabhi

Member
Joined
Jan 9, 2014
Messages
6,687
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
That's extremely impressive. Do you learn the algs somewhat intuitively, or is it just straight memory? How much time do you/have you put into learning these algs?

I have not measured the time I have put in for learning the algs, (but I started off 2 years back)

To memorise the algorithms , I use a story making process : https://drive.google.com/open?id=1SW4sAZ3RSENqmRNAah-CzCiPZm4nmj_T

If the alg cannot be brought to the form of [C:[A,B]] , then I just have to memorise them in blocks of 4 turns each.
 

abunickabhi

Member
Joined
Jan 9, 2014
Messages
6,687
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
Edges or corners? Which sets, and how did you learn them?

For edges mainly , for corners , I am still skeptical on whether , the algs are any better than 2 3-style algs.

Cube explorer is basically weak in finding good 5 cycles for corners , and one of the reason can be due to bad orientation of one of the corner in the 4 corners that need to be cycled, the overall algorithm can be long and inefficient. To make 5 style work on corners , some addendum kind of work has to be done , where 3 style is extended out in some likely and unlikely cases , and categorised , and scaled to 5 cycles fully.

So , my focus has been on learning edge algorithms till now.

(Corners are tough)
 

abunickabhi

Member
Joined
Jan 9, 2014
Messages
6,687
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
As far as I’m aware most 5 style algs aren’t very intuitive so probably straight memory.

I would also like to know how much time you invest into this and also how you get the algs, do you gen them and have a sheet somewhere with them? How optimal are they?

I am getting the algs via cube explorer, there is no other viable and more fundamental tool I have found. (comparing to ksolve , korf solver)

They are optimal in <U,R,L,F,D> , I try to avoid B moves , and y rotations.

Sheet link : 5 style algs

Medium : https://medium.com/@abhijeetgokar/5...ger-tricky-3x3-blindfolded-method-582d05e8fcc
 
U

Underwatercuber

Guest
I have not measured the time I have put in for learning the algs, (but I started off 2 years back)

To memorise the algorithms , I use a story making process : https://drive.google.com/open?id=1SW4sAZ3RSENqmRNAah-CzCiPZm4nmj_T

If the alg cannot be brought to the form of [C:[A,B]] , then I just have to memorise them in blocks of 4 turns each.
For edges mainly , for corners , I am still skeptical on whether , the algs are any better than 2 3-style algs.

Cube explorer is basically weak in finding good 5 cycles for corners , and one of the reason can be due to bad orientation of one of the corner in the 4 corners that need to be cycled, the overall algorithm can be long and inefficient. To make 5 style work on corners , some addendum kind of work has to be done , where 3 style is extended out in some likely and unlikely cases , and categorised , and scaled to 5 cycles fully.

So , my focus has been on learning edge algorithms till now.

(Corners are tough)
Just making sure I understand, so you have to recall a story everytime you want to use an alg? Do you ever get that mixed up with images/stories in bld/mbld?
 

abunickabhi

Member
Joined
Jan 9, 2014
Messages
6,687
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
Just making sure I understand, so you have to recall a story everytime you want to use an alg? Do you ever get that mixed up with images/stories in bld/mbld?

Rather than a story , it is just a snippet or a mental note.
It does not get mixed up with MBLD images , as they are like mantras, and quite fixed into my head , as compared to the temporary sticking images of the MBLD/3BLD cube.
 

Gian21x

Member
Joined
Oct 16, 2018
Messages
6
The original videos have been taken down (On youtube) a few years back. Luckily I have downloaded it a few years back and have it on my drive. PM for the link.
Thnks mate, I don't really need It, because I'm using proto-3style, I was asking my self if the guy Who made this amazing tutorial was still cubing
 
Top