Savants and cubing

[Saintlavin]

i remembered watching a documentary on Savants and their amazing ability to memorize and calculate at amazing propotions.

i was wondering, if we can combined a savant with cubing, with him/her learning all the method for solving 3x3x3 every single one algorithms and using the calculation ability to determine which one to use at a given time, would we reach a sub 4 time for the 3x3x3? it might be possible

 

[4Chan]

You aren't the first to think so.

That's very improbable.
However, 1LLL is a close alternative, and it IS possible.

 

[LNZ]

Savants came into the news with the death of the person who's life was immortalised in the move Rainman. As for remembering algorithms and methods, these people would be awesome.

But solving a cube also depends on how fast the brain can do real time pattern matching and the agility of your hands. And the synching of your hands,eyes and brain.

The Rainman savant was hopeless at ball sports. Not a good sign.

Here, savants may not have any advantage here. To test this ablity, you'd have the savant sit a Ravens progessive and advanced matrices test (like I did in the early 1990's with score 7/36 or 2/10, no advanced test result!) and do a test involving moving ball sports.

If a savant could pass a Raven's and a moving ball sport test well, I can easily see 4 seconds for a 3x3x3.

 

[Stefan]

Oh no, not that again...

 

[LewisJ]

That bull**** again?

This.

Let's be optimistic and say that this person could solve in 30 moves every time, say they know all kinds of advanced F2L, xcrossing, and 1LLL, or maybe some incredible blockbuilding and 1LLL or something like that. They would still need 7.5 TPS to pull off 4 second solves. Looking at some of Lucas Garron's reconstructions of very fast solves, we see that only the very best have 6.xx TPS counts, and, while I didn't check every single one, the highest TPS I saw was 6.85TPS on Yu Nakajima's 6.57 second solve, and for official times, 6.07TPS on Erik's 7.08 solve.

Now let's say this savant is a really good turner and averages 6 TPS. They would still need 24 move solutions to get 4 second times - first place in 3x3 FMC is 22, 2nd is 26. This simply isn't going to happen.

 

[qqwref]

The Rainman savant was hopeless at ball sports. Not a good sign.

Funny story: I used to be hopeless at ball sports (literally... if you tossed a ball at me I couldn't catch it). Then I took up juggling for a while. This is definitely a field where natural ability plays a large role, but it's not like practice won't help, and I don't expect a savant (who may very well have no friends to play ball sports *with*) to get a lot of practice in this.

Anyway, savant-like abilities are typically limited to either very good memory or very specific things, such as calculating large primes, knowing weekdays of any day ever, being able to count hundreds of objects at a glance, etc... And note that these abilities are not always found in people with otherwise below-normal intelligence. I guess it would be very useful for someone with an extraordinary memory to take up cubing, as they could do xcross + optimal multislot + ZB or whatever, but the problem is always TPS and fingertrickiness. Someone who can average 35-40 moves in speedsolving (and this IS possible) would still need a consistent 4+ TPS to get sub-10 seconds. I agree that the best times will be had by people who use an ordinary solution with extraordinary speed, rather than an extraordinary solution with ordinary speed.

 

[Stefan]

calculating large primes

Haven't heard that before, can you tell more?

knowing weekdays of any day ever

Not "knowing". Calculating. With easy methods. And non-"Savants" are better at it.

being able to count hundreds of objects at a glance

Anywhere I can find this besides a Hollywood movie?

 

[deadalnix]

Stefan, look at Rüdiger Gamm or Alexi lemaire can do. That's pretty incredible, and this is much more than caculation day of week.

But anyway, 4s is just ********.

 

[Stefan]

Stefan, look at Rüdiger Gamm or Alexi lemaire can do. That's pretty incredible, and this is much more than caculation day of week.

Yeah, I know them. And they're just using methods and practice to become better. Do you realize you're just like the people who say "OMG he just solved a Rubik's Cube WOW that's unbelievable!11!~!OMFG"?

Btw, what do they have to do with this thread? I don't think they're regarded as "savants".

 

[LewisJ]

Haven't heard that before, can you tell more?

Not "knowing". Calculating. With easy methods. And non-"Savants" are better at it.

First, Daniel Tammet claims to "sense" whether a number is prime or not...and even if he can't sense it, he could probably use his ridiculous visual calculation abilities to figure out whether it is or not.

Second, add on telling people what was in the newspaper on that day and it's a bit more impressive.

 

[Stefan]

he could probably use his ridiculous visual calculation abilities to figure out whether it is or not.

Please don't make up myths, there are enough already.

Second, add on telling people what was in the newspaper on that day and it's a bit more impressive.

A bit, yes. And only if it was really tested and checked.

 

[Carson]

Savants typically suffer from Autism. One of the primary indicators of autism is extreme social awkwardness and withdrawal from social situations... and if you have ever been to a cubing competition, they tend to be pretty social.

Savants typically are unaware of "calculating" things. I recall watching a 20/20 interview with the guy that inspired Rainman. There was discussion of his ability to identify the day of the week of any date. His mind certainly was using some type of calculation, however he was unaware of this fact. To him, it was if he just KNEW the answer.

Another symptom of Autism is the inability to "filter out" sensory input. Where as most people can sit in a room with a radio playing soft music, traffic passing by on the street outside, the heat/air kicking on/off occasionally, the upstairs neighbors walking around occasionally, and many other "minor" things occuring, and still concentrate on what they are doing (like cubing)... someone with Autism is going to actively hear all of these things and will be unable to focus their attention away from them. This isn't just limited to hearing... every movement, sound, smell, etc etc etc is "unignorable" to someone with autism.

^^This is of course most prevalent in extreme cases, but typically those individuals identified as having savant tendencies fall into the category of extreme cases.

 

[rubixfreak]

I watched a really interesting BBC movie about Savants. THey said that they believe there's something wrong in their left brain halve. Where normal people have a kind of spam filter in it, in Savant's brain it is somehow damaged and not working properly so like Carson said above they cannot focus on one thing. Moreover they are sentenced to remember every uninteresting and unneccacery facts of their life. for example things like what did you watched on TV on 12.4.1999 ?

 

[Stefan]

Let me give you an example of how much you should trust the media about this. Here are four appearances of Ruediger Gamm in four different shows:

What is 62/109?
What is 63/109?
What is 237/109?
What is 982/109?

See the pattern? They all asked him to divide by 109. Why? Probably cause he asked them to ask for that. Cause this way he only ever had to memorize one cycle of 108 digits that he can use again and again and again to impress the fools. Just in case you really don't get it, let me try to visualize it:

62/109 = 0.568807339449541284403669724770642201834862385321100917431192660550458715596330275229357798165137614678899082(5688073394...)
63/109 = 0.577981651376146788990825688073394495412844036697247706422018348623853211009174311926605504587155963302752293(5779816513...)
237/109 = 2.174311926605504587155963302752293577981651376146788990825688073394495412844036697247706422018348623853211009(1743119266...)
982/109 = 9.009174311926605504587155963302752293577981651376146788990825688073394495412844036697247706422018348623853211(0091743119...)

He just learned a single cycle of 108 digits and then every time these shows asked him to divide a number by 109, he just determined the number before the dot and where to start in the cycle, and then recited the cycle. This is no more impressive than a child singing Jingle Bells! But of course he closes his eyes, gestures with hands and face and makes it look like he's crazily computing difficult stuff. When in reality he makes a simple calculation and then just recites. But the media (and audience!) loves it and doesn't care how it really works. They just want to show something that the average naive uneducated person will find amazing. They don't even bother to check whether he's correct! In case of the above 237/109, he actually did make a mistake! He started with "2.17432660550" which is missing "119" and they just don't care! They showed it anyway pretending he's correct because they know full well that the average idiot viewer won't check!

And look at the Wikipedia page about Kim Peek. There *is* a section titled "Scientific investigation" but what is it? A test of his capabilities? No, just scanning his head to look for biological anomalities. What the F? Since when do you study *how* something might work before verifying *that* it works?

Some general rules:
- Much of what is presented/perceived as "calculation" is really just recitation and simple tricks.
- There are very powerful techniques for memorization/recitation. And training goes a long way.
- As far as I'm aware, the top memorizers, mental calculators and speed readers are all normal healthy people using techniques and training. Not "savants" with unexplainable magic powers.
- Always compare these things to cubing, and realize you look at their stuff like non-cubers look at cubing.
- Don't f-ing believe the media about this! They'd rather tell you that someone has strange unexplainable capabilities than admit that it's just simple techniques everybody can learn.
- If you want to speculate about great memorizers and their cubing potential, check out the f-ing memory competitions and their records.

Finally, watch this: http://www.youtube.com/watch?v=YRz-yIpW3Qw
That's Kim Peek reciting US states/capitols and when they joined the union. So what? Little children can do that! He's a grown man! Like I wrote before, if Kim Peek had decided to learn to solve a Rubik's cube and then solved it in 3 minutes, media and people would go all "OOh look Rainman can solve a Rubik's cube UNBELIEVABLE!!!". And they'd probably claim he can't explain it but "senses" what moves to make.

Wake up!

 

[Me]

he could probably use his ridiculous visual calculation abilities to figure out whether it is or not.

Please don't make up myths, there are enough already.

Tammet's abilities are synesthetic, he associates colors and feelings with numbers. Sure there are those who claim to be savants, I agree with your example of Peek reciting US states/capitals order. However I still believe there are savants who posses some true ability.

Since when do you study *how* something might work before verifying *that* it works?

Is there anything that would convince you that it works?

However in relation to the topic at hand, I agree with the reasons above that savants would not be good at cubing.

 

[Stefan]

Tammet's abilities are synesthetic

And I don't care. As long as normal people using normal techniques and normal training get better results, I don't understand why people are glorifying the "savants".

Is there anything that would convince you that it works?

Proper tests of properly described stuff. All we ever get are stupid sensational unverified vague anecdotes and shows.

 

[Stefan]

Btw, my "verify that it works" was a bit short. I also mean that they should properly describe "it" first. Then verify it, then try to understand it. But in most cases even this first step is skipped. All we get is little uncomparable meaningless demonstrations. But it doesn't have to be like that!

Kim Peek did do the human calculator thing, right? That is not at all something special. Many people do it, it is included in mental calculation competitions, and there are record lists for it. We've been shown Kim Peek doing it, but I've never seen it done properly, like they do it in competitions. Why? Probably cause that would allow comparison and then people would realize that he sucked compared to the normal healthy people who are better at it. And don't give me that **** about him being autistic and not feeling comfortable being tested. He let them test his brain with an MRI and stuff like that. And he did it in shows on TV and real life. The tests he didn't do are just the tests that would expose him as being nothing special.

 

[Kyle Barry]

I agree with Stefan in that these abilities, for the most part, can be matched by non-"Savants." With very little practice, I can calculate the day of the week, just like Kim Peek, in 6-7 seconds, using a rather simple algorithm. People who are not savants can do this in about 1 second. Whether Kim Peek memorized the calendar, or has a method, he absolutely calculates it or draws it from memory, he doesn't intrinsically know the answer, of course. Specifically, I'm not positive of this, but I believe there are irregularities in the calendar that would throw him off if he didn't know the changes. I just looked at year 800, and with my algorithm, I would get those dates wrong, so it doesn't make sense, especially with a human invention, to just "know" what day it is.

 

[qqwref]

calculating large primes

Haven't heard that before, can you tell more?

I read in a reputable book somewhere about two savant twins who would tell each other large (six-digit or so) numbers back and forth, and the neuroscientist who was observing them later figured out that those numbers were all prime and gave them some larger numbers, which they agreed were prime (although it took a bit longer.

knowing weekdays of any day ever

Not "knowing". Calculating. With easy methods. And non-"Savants" are better at it.

No, not with those methods. I think the idea was something along the lines of knowing the day of every date in sequence and "flipping through" those days like a book. The reason I say it was like this was because there are savants who know the weekday of only the dates they have personally witnessed, suggesting that they are not calculating but rather remembering what weekday it was. If someone can essentially visualize time as a long calendar it is not hard to see how this would be possible. Anyway, the point is that savants who have this ability do it by knowing (or, for some, remembering) the weekday, whereas normal people would find out a weekday using a calculation method. A savant with this ability would unconsciously know what weekday a certain date was in exactly the same sense as someone with perfect pitch would be able to immediately sing a B-flat two octaves above middle C, without consciously "calculating" what note that would be.

being able to count hundreds of objects at a glance

Anywhere I can find this besides a Hollywood movie?

I saw an article about someone who unconsciously counted things and would end up just "knowing" that there were, say, 83 books in a bookshelf. This person wasn't a savant but it seems likely that there would be savants with this ability. It may also be related to memory - if someone could take a 'snapshot' of a bookshelf or whatever, they could count objects from their (stable) snapshot easily without actually looking at the things again.


PS, Stefan... don't confuse savants who have some kind of intuitive feel for doing a computation with people who have consciously trained and practiced to do it. The latter show that the human brain has the ability to do these things, but just because they have deliberately put in effort does not mean the former people also did. I would suggest that savants have not deliberately tried to hoax people by pretending an easy skill, but rather that the feats they find themselves naturally being able to do are also things that can be learned through practice. Just because it is possible to pretend to be a savant does not mean all savants are pretending.

PPS: Remember that the impressive thing is that the abilities are ingrained in someone who does not seem to normally have high mental ability. I don't care if other people can do it better with years of practice; it's still impressive to me that someone can do it without formal training. If someone came along and could solve the cube in 50 moves without knowing any algorithms or any specific method, and while being of below-average intelligence too, I think I would be impressed. It doesn't matter to me that people who know hundreds of algorithms and dozens of methods and who have practiced for years can consistently do linear FMC better than that.

Besides, many of the competition settings are designed for normal people and not savants (clearly, because normal people are competing in them). For instance the memorizing numbers competition. There may be people who can look at a field of 80 numbers for a few seconds and then write them all out later, but if the competition setting is not amenable to that then they cannot get the record. Take Solomon Shereshevsky: he could memorize huge matrices of numbers in relatively little time, and then remember them *years* later. This is an extraordinary feat but it is clearly not applicable to memory competitions. He could also memorize poems in a language he did not know, but of course they do not do this in memory competitions either because it is impossible to come up with a language everyone is guaranteed to not know, so they use a known language (such as English), but then everyone involved has the advantage of knowing the language. So I think some savants' skills are not set up to work in a competition setting, simply because competition settings are designed to give the advantage to people who have acquired the skill through practice and study.

 

[Stefan]

I read in a reputable book somewhere about two savant twins who would tell each other large (six-digit or so) numbers back and forth, and the neuroscientist who was observing them later figured out that those numbers were all prime and gave them some larger numbers, which they agreed were prime (although it took a bit longer.

I assume it's this? (hopefully that link works, let me know if not)
http://resources.metapress.com/pdf-preview.axd?code=c4p16721h27264t1&size=largest
Yeah, seems really reputable.

A savant with this ability would unconsciously know what weekday a certain date was

That's what they claim, yes. And like I said before, as long as they're not outperforming the normal people, I don't care.

an article about someone who unconsciously counted things and would end up just "knowing" that there were, say, 83 books in a bookshelf.

Great, more anecdotes...

PS, Stefan... don't confuse savants who have some kind of intuitive feel for doing a computation with people who have consciously trained and practiced to do it.

How did you get the impression I might be confusing that?

The latter show that the human brain has the ability to do these things, but just because they have deliberately put in effort does not mean the former people also did.

It also doesn't mean that they didn't.