The Future of Competitive Speedsolving (a discussion)

[qqwref]

Ok.. But why are you guys acting like roux is the legendary solution.. Its just a different form of cfop.. Build, preserve, build.. Then fixing orientation and permutation.

AHAHAHA LMFAO WOW


PS: I think Roux can be almost exactly as fast as Fridrich, at least in 2h solving. I think in 5-10 years we will see world-class Roux solvers who are on par with the top CFOPers.

 

[Stefan]

I have an extreme memory.

Can you describe that a bit more?

 

[cubersmith]

Actually it is true. If you are color neutral, thr odds of finding a CE pair during inspection is what Jorghi said, around 60%

Not after cross because then you have picked your colour lul.

 

[Cubenovice]

Can you describe that a bit more?

theZcuber
Premium Member
--------------------------------------------------------------------------------
Join Date May 2011
Location I forgot
Posts 518

 

[Jorghi]

Not after cross because then you have picked your colour lul.

... Do you understand we said during inspection? So after nothing.

 

[Phlippieskezer]

PS: I think Roux can be almost exactly as fast as Fridrich, at least in 2h solving. I think in 5-10 years we will see world-class Roux solvers who are on par with the top CFOPers.

Similarly, I think there soon (as in, 5-10 years, so not too soon, mind you ) will be a number of world-class ZZ solvers, both OH and 2H.

PS: Totally off-topic, but: I want to see the OH WR be broken by a Roux solver now. That would be so awesome.

PSS: Guys... We're straying off topic again (as am I).

 

[IamWEB]

3x3x3 still has a ways to go with limit-reaching. 2x2x2 could advance a bit more.

Can you describe that a bit more?

theZcuber
Premium Member
--------------------------------------------------------------------------------
Join Date May 2011
Location I forgot
Posts 518

XD

 

[Phlippieskezer]

3x3x3 still has a ways to go with limit-reaching. 2x2x2 could advance a bit more.

Well, in terms of 2x2 singles, I don't think it's going to go very far. There are already 3 people sharing the WR single. The average is also already very low, though I guess that could certainly improve, though probably not by much.

 

[qqwref]

Eventually someone will get a 0.5ish 2x2 single.

Also: eventually someone will get a 3-4ish sq1 single. Sufficiently easy scrambles definitely do pop up.

 

[Mike Hughey]

Well, in terms of 2x2 singles, I don't think it's going to go very far. There are already 3 people sharing the WR single. The average is also already very low, though I guess that could certainly improve, though probably not by much.

What's to stop us from getting a ridiculously lucky scramble, say 1 move from solved? Is there anything in the regulations that provides for removing such a scramble? I know that a number of years ago, a very short scramble (3 moves, if I remember right?) was removed by someone very respected (was it Ron? I don't remember), which kept the record from being ridiculously low years earlier. I would think with a 1-move-from-solved scramble, we'd easily see sub-0.5, and probably sub-0.3.

Does the random-state generator for 2x2x2 prevent a solved cube?

 

[Phlippieskezer]

What's to stop us from getting a ridiculously lucky scramble, say 1 move from solved? Is there anything in the regulations that provides for removing such a scramble? I know that a number of years ago, a very short scramble (3 moves, if I remember right?) was removed by someone very respected (was it Ron? I don't remember), which kept the record from being ridiculously low years earlier. I would think with a 1-move-from-solved scramble, we'd easily see sub-0.5, and probably sub-0.3.

Does the random-state generator for 2x2x2 prevent a solved cube?

So, it's essentially just going to be up to luck to give us a nice scramble in order to beat the world record...

 

[Mike Hughey]

So, it's essentially just going to be up to luck to give us a nice scramble in order to beat the world record...

That's why I guess I'd prefer to see some sort of arbitrary limit as to the minimum number of moves from solved. Since we're using a random state generator, it wouldn't be that hard to do that - especially with 2x2x2.

 

[Cubenovice]

That's why I guess I'd prefer to see some sort of arbitrary limit as to the minimum number of moves from solved. Since we're using a random state generator, it wouldn't be that hard to do that - especially with 2x2x2.

I suppose there is something like a minimum length included in the WCA scramblers but even set at 7 HTM this could give ridicolous scramble...
How about getting a 2x2x2 that is Sune way from being solved and how are you ever going to prevent this?

Delegates hand-picking the easy / well known alg cases out?
oh wait, delegates should not look into the scrambles as they typically are competitors...

 

[cubernya]

theZcuber
Premium Member
--------------------------------------------------------------------------------
Join Date May 2011
Location I forgot
Posts 518

LOL that's a good explaination.

I don't really know how to describe it, I can memorize 50 random digits in around 10 minutes, remember small changes others would't be able to find, etc.
If you really want more depth, PM me and I'll try my best to explain it (it's complicated)

 

[Cheese11]

Eventually someone will get a 0.5ish 2x2 single.

Also: eventually someone will get a 3-4ish sq1 single. Sufficiently easy scrambles definitely do pop up.

But they might pop up with someone that doesn't realize it.

 

[uberCuber]

I suppose there is something like a minimum length included in the WCA scramblers but even set at 7 HTM this could give ridicolous scramble...
How about getting a 2x2x2 that is Sune way from being solved and how are you ever going to prevent this?

Delegates hand-picking the easy / well known alg cases out?
oh wait, delegates should not look into the scrambles as they typically are competitors...

I think the possibility of stupid easy scrambles for 2x2 should just be accepted. It's a part of the puzzle. And really, with 2x2 there isn't an effective way to draw a line that wouldn't just result in competitors receiving nothing but the hardest positions. We are getting close enough to physical limit with the puzzle as it is. The same goes for Pyraminx.


I'm not going to bother making any statement regarding the future of 3x3 solving. Not only does it not interest me anyway, but predictions of all kinds regarding 3x3 are made so often, and any prediction I could make myself would likely end up with myself sounding dumb, either now, or in a year.
As for bigcubes, I definitely believe we have a long way to go. My feeling on this, a feeling shared by Stachu as shown in his 4x4 FMC thread, is that we just do not have the greatest speedsolving methods that we should be able to come up with. Not that I think we do for 3x3, but I think there is much farther to go with bigcubes. They are just not as well explored as the 3x3 is. I do not see improvement at bigcubes halting any time soon.

Actually it is true. If you are color neutral, thr odds of finding a CE pair during inspection is what Jorghi said, around 60%

Whether you are color neutral or not does not affect the scramble.... The probability of a scramble having a corner-edge pair is exactly the same either way. (This comment is just as much directed at Jorghi as it is you.)

I don't really know how to describe it, I can memorize 50 random digits in around 10 minutes, remember small changes others would't be able to find, etc.
If you really want more depth, PM me and I'll try my best to explain it (it's complicated)

I did that with the first 50 decimal places of pi four years ago. I still remember them perfectly today. And yet remembering memo for a single 3x3BLD is still difficult for me sometimes.... >:S

 

[cubernya]

Pi isn't just a random number. I memorized that to around 500 digits in an hour

 

[uberCuber]

Pi isn't just a random number. I memorized that to around 500 digits in an hour

It is still a string of digits that seems perfectly random to someone who hasn't looked at it before.

But anyway, I wasn't saying that trying to act like I have a better memory than you. It was more of a random vent about the fact that I fail at remembering BLD memo even though I still remember those pi digits after four years.
And regardless it would be pretty dumb to take a memory "feat"(lol) that I randomly did in a 7th grade math class, and act like it in any way represents the way my brain works right now

EDIT: K, I just helped go off topic. I won't post about this again.

 

[cubernya]

Yeah Chris warned us about going off topic. Lets stop this discussion about my memory now

 

[qqwref]

So, it's essentially just going to be up to luck to give us a nice scramble in order to beat the world record...

It's been up to luck for a long time.

But they might pop up with someone that doesn't realize it.

So we wait for another one. I said "eventually" And remember, everyone in a round gets to try the same scramble, so you just need one person with decent turning abilities to get lucky.

My feeling on this, a feeling shared by Stachu as shown in his 4x4 FMC thread, is that we just do not have the greatest speedsolving methods that we should be able to come up with. Not that I think we do for 3x3, but I think there is much farther to go with bigcubes. They are just not as well explored as the 3x3 is. I do not see improvement at bigcubes halting any time soon.

Personally I've investigated a whole bunch of methods (for speed) and I think Reduction (and Yau for 4x4 only) is pretty much near the limit of how fast the puzzle can be solved by a human. Over time people are going to discover more little tricks like they did for F2L, but I think the basic approach will remain the same.