[WCA Regulations 2014] Scramble Filtering (new poll)

[Methuselah96]

Pros of scramble filtering (more than the +2 and solved positions):

• Reduces chance of slower solvers getting ridiculously lucky and undeserved solves.

The 2014 changes to the WCA Regulations included the addition of 4b3 (https://github.com/cubing/wca-documents/issues/48):

4b3) Specification for the scramble program: An official scramble sequence must produce a random state from those that require at least 2 moves to solve. The following additions/exceptions apply:

4b3a) For blindfolded events, the scramble sequence must orient the puzzle randomly (equal probability for each orientation).
4b3b) 2x2x2 Cube: The (random) state must require at least 4 moves to solve.
4b3c) Pyraminx and Skewb: The (random) state must require at least 7 moves to solve.
4b3d) Square-1: The (random) state must require at least 11 moves to solve.
4b3e) 5x5x5 Cube, 6x6x6 Cube, 7x7x7 Cube, and Megaminx: sufficiently many random moves (instead of random state), at least two moves to solve.

Here is what would be filtered:
Rubik's Cube: 19/43,252,003,274,489,856,000 (0.000000000000000043928601%)
2x2x2 Cube: 385/3,674,160 moves (0.01047858558%)
Pyraminx: 221,233/7,558,2720 moves (0.29270314696%)
Skewb:7,1625/3,149,280 moves (2.27432937052%)
Square-1: working on that

Here is a discussion for filtering for more info: https://groups.google.com/forum/#!topic/wca-scrambler/3x_hN9Y4HCM
This regulation was added in an effort to add transparency to the process that TNoodle uses to find random scrambles. This scrambling filter (except for Skewb and Square-1) has been in effect for a long time and was documented in the readme for TNoodle but was never addressed in the Regulations or Guidelines.

This discussion was originally brought up in this thread but I thought a poll would be useful.

Should there be scramble filtering for determining scrambles?
If so, how should these limits be decided?

 

[Kirjava]

I'm resigned to the fact that this probably won't ever change, but have to register my disdain at 4b3b, 4b3c and 4b3d - I truly believe we should not be doing this.

 

[uberCuber]

I'm resigned to the fact that this probably won't ever change, but have to register my disdain at 4b3b, 4b3c and 4b3d - I truly believe we should not be doing this.

Where did they even pull the numbers 7 (for pyraminx) and 11 (for square-1) from? Did they just make up some random numbers that sounded good to them?

I also don't like these regulations.

 

[Erik]

I'm resigned to the fact that this probably won't ever change, but have to register my disdain at 4b3b, 4b3c and 4b3d - I truly believe we should not be doing this.

I strongly agree with this. Either go with a truly random position or (if you like to reason that even noobs can scramble so that there is not a 4 move solution) let random bypassers pick a position. Setting ANY filter is wrong therefore I think.

 

[Methuselah96]

Where did they even pull the numbers 7 (for pyraminx) and 11 (for square-1) from? Did they just make up some random numbers that sounded good to them?

I also don't like these regulations.

Adding those regulations changed nothing. The minimum number of moves was already in place in TNoodle which is the only official WCA scrambling program. These regulations were added to inform the cubing community how the scrambles were being generated (https://github.com/cubing/wca-documents/issues/48). Here is a conversation about scramble filtering: https://groups.google.com/forum/#!topic/wca-scrambler/3x_hN9Y4HCM

I strongly agree with this. Either go with a truly random position or (if you like to reason that even noobs can scramble so that there is not a 4 move solution) let random bypassers pick a position. Setting ANY filter is wrong therefore I think.

If a random state was one move away would you want somebody who has no idea how to cube get the WR just because they could see one move in advance or would you give it to the people who actually know how to cube?

 

[Escher]

Where did they even pull the numbers 7 (for pyraminx) and 11 (for square-1) from? Did they just make up some random numbers that sounded good to them?

I also don't like these regulations.

I strongly agree with this. Either go with a truly random position or (if you like to reason that even noobs can scramble so that there is not a 4 move solution) let random bypassers pick a position. Setting ANY filter is wrong therefore I think.

Shame that when this change was enacted, I remember me and Kir posting several times in the relevant thread (I'll try to dig it up) trying to explain our reasoning and we didn't seem to be very popular.

 

[uberCuber]

Adding those regulations changed nothing. The minimum number of moves was already in place in TNoodle which is the only official WCA scrambling program

This has nothing to do with the fact that I disagree with this scramble filtering.

If a random state was one move away would you want somebody who has no idea how to cube get the WR just because they could see one move in advance or would you give it to the people who actually know how to cube?

If a person has no idea how to cube, they wouldn't be trying to compete and would never see the scramble in the first place.

 

[Methuselah96]

If a person has no idea how to cube, they wouldn't be trying to compete and would never see the scramble in the first place.

Ok. If a person normally solves cubes in 50 seconds.

 

[uberCuber]

Ok. If a person normally solves cubes in 50 seconds.

Presumably such a person would not be able to start a stackmat timer, perform a move on their cube, and stop the timer in a faster time than every fast-ish cuber in the round. And if they did manage to, then good for them. Luck is inherent to cubing.

 

[Erik]

First of all: you know perfectly well that the chances of this are smaller than that I win the lottery 10 times on a day without rain in the Netherlands.
And to answer your question: in that case that person would be lucky. Good for him, yay, much wow. But that is part of the game, we have averages of 5 to filter the skill.

Imho a 1 move scramble and solution is no worse than a 4 move scramble and solution. Both extremely lucky and unbeatable on a 'harder' scramble.

 

[qqwref]

Adding those regulations changed nothing. The minimum number of moves was already in place in TNoodle which is the only official WCA scrambling program. These regulations were added to inform the cubing community how the scrambles were being generated

I told you guys! If the way scrambles are generated is defined by a bunch of code rather than by the regulations, changes will happen and you won't even know until someone feels like telling you, because the regulations will stay the same...

 

[Methuselah96]

Presumably such a person would not be able to start a stackmat timer, perform a move on their cube, and stop the timer in a faster time than every fast-ish cuber in the round. And if they did manage to, then good for them. Luck is inherent to cubing.

For some reason I don't think the cubing community would be very happy with the 3x3 single record becoming unbeatable and would also defeat any purpose that 3x3 single record might have.

First of all: you know perfectly well that the chances of this are smaller than that I win the lottery 10 times on a day without rain in the Netherlands.
And to answer your question: in that case that person would be lucky. Good for him, yay, much wow. But that is part of the game, we have averages of 5 to filter the skill.

Imho a 1 move scramble and solution is no worse than a 4 move scramble and solution. Both extremely lucky and unbeatable on a 'harder' scramble.

Yes, it is EXTREMELY unlikely, so why not filter it?
If averages of 5 are the real test of skill why do we even have singles?

I told you guys! If the way scrambles are generated is defined by a bunch of code rather than by the regulations, changes will happen and you won't even know until someone feels like telling you, because the regulations will stay the same...

That is why we're adding these regulations now and not later.
Although this information has been available in the readme for TNoodle if anyway ever bothered to look.

I am curious as to how they chose these numbers as previously mentioned. The exceptions I get, but having 2 moves as a standard seems odd. I don't think two moves away on a 3x3 would be any harder than one move.

 

[Erik]

For some reason I don't think the cubing community would be very happy with the 3x3 single record becoming unbeatable and would also defeat any purpose that 3x3 single record might have.

This is already reality on 2x2.

Yes, it is EXTREMELY unlikely, so why not filter it?
If averages of 5 are the real test of skill why do we even have singles?

This logic does not make sense. Seriously.

That is why we're adding these regulations now and not later.
Although this information has been available in the readme for TNoodle if anyway ever bothered to look.

I really don't like the fact that this was almost conceiled and I'm glad it ended up in the regulations.

 

[Methuselah96]

This is already reality on 2x2.


This logic does not make sense. Seriously.




I really don't like the fact that this was almost conceiled and I'm glad it ended up in the regulations.

But it doesn't have to be reality for other puzzles.

Your logic is that it's so unlikely that we won't ever have to deal with it.
My logic is that it would defeat the purpose of a single in the case that it did happen, so why we would let that position even have a remote possibility of being generated?

 

[Erik]

But it doesn't have to be reality for other puzzles.

Your logic is that it's so unlikely that we won't ever have to deal with it.
My logic is that it would defeat the purpose of a single in the case that it did happen, so why we would let that position even have a remote possibility of being generated?

You don't seem to understand what I'm saying:

you think it's a problem that there is the chance of extreme luck and try to solve this with a filter.

I say:

- There is no problem. Luck is part of the cube.
- A filter does not solve your self-created problem.

 

[Kirjava]

I don't understand how people seem to think that 2 moves is silly and compromises the integrity of the event, but 4 moves is suddenly respectable.

God's Number for 2x2x2 and Skewb is the same, the case distribution and number of states is very very similar, yet the scramble filter limits are completely different.

 

[Methuselah96]

You don't seem to understand what I'm saying:

you think it's a problem that there is the chance of extreme luck and try to solve this with a filter.

I say:

- There is no problem. Luck is part of the cube.
- A filter does not solve your self-created problem.

I understand what you're saying.
- I do not think luck should be the deciding factor of determining records.
- A filter would certainly help.

I created a poll for more discussion.

 

[Username]

That pyra filter rules out about 6.4% of all possible positions. I don't like that.

 

[uvafan]

That pyra filter rules out about 6.4% of all possible positions. I don't like that.

I agree. The pyraminx move filter is absolutely ridiculous. If someone gets a fairly lucky scramble and sets the WR single, it's just part of cubing... 7 moves is way too high.

EDIT: In general, I also disagree with scramble filtering. It is totally unnecessary. Luck is an element in any sport, and eventually it won't really be WRs that matter because they will be broken very rarely, it will be the winner of head-to-head competitions such as worlds, and all competitors will have the same scrambles so...

 

[ThomasJE]

Here is what I agree and disagree with:

4b3) Specification for the scramble program: An official scramble sequence must produce a random state from those that require at least 2 moves to solve. The following additions/exceptions apply:

4b3a) For blindfolded events, the scramble sequence must orient the puzzle randomly (equal probability for each orientation).
4b3b) 2x2x2 Cube: The (random) state must require at least 4 moves to solve.
4b3c) Pyraminx and Skewb: The (random) state must require at least 7 moves to solve.
4b3d) Square-1: The (random) state must require at least 11 moves to solve.
4b3e) 5x5x5 Cube, 6x6x6 Cube, 7x7x7 Cube, and Megaminx: sufficiently many random moves (instead of random state), at least two moves to solve.

BLD events: Adds yet another element of luck. We don't need that.
2x2: Agreed. This also means the single can still be broken.
Pyraminx: Limiting solutions to 7 moves when God's number is 11 is too high. Maybe something like 5.
Skewb: No idea
Sq-1: No idea
5x5, 6x6, 7x7: Kind of goes without saying.