Throwing out scrambles in competitions?

[cuBerBruce]

I do not see it as a waste of time. They are valid scrambles.

In a round where average matters, the time on that scramble will be thrown out when computing the average. Therefore, it will essentially have no bearing on the outcome of the competition. Hence, a waste of time. In a "best of" round, it's the non-trivial scrambles that will be a waste of everyone's time.

 

[Kirjava]

In a round where average matters, the time on that scramble will be thrown out when computing the average. Therefore, it will essentially have no bearing on the outcome of the competition. Hence, a waste of time.

Same goes for any round with an easy (and 4 or more moves) scramble.

In a "best of" round, it's the non-trivial scrambles that will be a waste of everyone's time.

Same goes for any round with an easy (and 4 or more moves) scramble.

 

[Stefan]

Your complaint is against triviality and non-cubers being able to perform those solutions. Rubik's Magic?

No, I wouldn't say "perform". Sure they can *learn* and then perform a Rubik's Magic solution easily, but how easily can they solve it on their own?

2-move or perhaps 3-move scrambles on the other hand can even be solved by non-cubers with no previous knowledge. That's also one reason why I think the opposite of Zane - I'd remove 2-movers but not 2-swappers (Jperm). The latter are still very hard for non-cubers.

You will find that it is not possible to define an easy or hard scramble. My 0.96 was 7 moves, so do me a favor and filter out every scramble solvable in 6 or less moves...

Like I said before:

"We're not going to find a perfect filter, where nobody wants to filter more or less. I'm thinking of it as at least doing *something*, filtering something that everybody agrees is easy."

My suggestion of removing 3-moves-or-less scrambles does not attempt to be perfect distinction of what's easy and what's hard. It just says that 3-moves-or-less *is* easy, not that everything else isn't. Do you disagree that 3-moves-or-less is easy? Do you think anyone else does? What about 2-moves-or-less?

 

[Mike Hughey]

I would like to say that, although I lean towards agreeing with Stefan as opposed to Thom on believing that it's reasonable to choose a minimum number of moves from solved arbitrarily, I do think that until we have a rules change, delegates should absolutely be barred from removing any scramble. The rules don't provide for it, and it seems quite unreasonable to make something like that a judgement call on the part of individual delegates.

What this means is, if we do decide we want to allow choosing a minimum number of moves from solved on an arbitrary basis (and I agree that it certainly would be arbitrary - I just see nothing wrong with being arbitrary in this case), we should do so as soon as possible, and codify that in the rules and include it in the scramblers. If next weekend someone gets a 2-move 2x2x2 scramble, then we will be reduced to either throwing out the 2x2x2 single WR or agreeing to Thom's perspective and just going with any random scramble. So I can imagine Thom is probably rooting for just such a scramble. If that were to happen, I would then agree with Thom (although perhaps for different reasons).

 

[Kirjava]

It just says that 3-moves-or-less *is* easy, not that everything else isn't. Do you disagree that 3-moves-or-less is easy? Do you think anyone else does?

I think we all agree that it's easy. We're trying to say that it doesn't matter.

 

[Stefan]

I think we all agree that it's easy. We're trying to say that it doesn't matter.

I was just countering the argument that we shouldn't do it because we can't find a reasonable definition. If a definition says "these are easy" and everybody agrees that those are easy, then that definition is reasonable.

I do think that until we have a rules change, delegates should absolutely be barred from removing any scramble. The rules don't provide for it

Yes, I agree (just in case that wasn't clear). At least when the delegate competes and thus doesn't scramble, the rules don't just not provide for it but explicitly forbid it ("must be kept secret for all but the scramblers").

If next weekend someone gets a 2-move 2x2x2 scramble, then we will be reduced to either throwing out the 2x2x2 single WR or agreeing to Thom's perspective and just going with any random scramble. So I can imagine Thom is probably rooting for just such a scramble.

Yeah, and I'm rooting against it. Right now we still have the chance to set the limit a little higher without introducing inconsistency.

 

[Erik]

Like I said before:

"We're not going to find a perfect filter, where nobody wants to filter more or less. I'm thinking of it as at least doing *something*, filtering something that everybody agrees is easy."

My suggestion of removing 3-moves-or-less scrambles does not attempt to be perfect distinction of what's easy and what's hard. It just says that 3-moves-or-less *is* easy, not that everything else isn't. Do you disagree that 3-moves-or-less is easy? Do you think anyone else does? What about 2-moves-or-less?

I dont think we can even agree on *something*. Like I said before there are people who are in favor of any scramble to be valid and there are some that think there should be filtering. The attempt of filtering anything is already a bridge too far on the 'agreeing ladder'.

 

[Stefan]

I dont think we can even agree on *something*.

You really think someone would disagree that 2-moves-or-less is easy?

 

[Kirjava]

You really think someone would disagree that 2-moves-or-less is easy?

I think he meant we can't agree on what to do about 2 and 3 move scrambles.

I'm not really rooting for a 2 move scramble or anything. It'll likely get removed anyway.

 

[Erik]

The fact that its easy doesnt mean it should be forbidden. In my opinion cubing is still solving from a random state.

 

[Stefan]

The fact that its easy doesnt mean it should be forbidden.

Like I said, I just complained about your "it is not possible to define an easy or hard scramble".

 

[Kirjava]

It is not possible to draw a line between easy and not easy.

 

[qqwref]

It'll likely get removed anyway.

And therein lies the problem. It'll LIKELY get removed. The most important reason to draw some arbitrary line is fairness. We don't want a situation where one delegate sees a 3-move 2x2x2 solution and accepts it, and another sees the same solution and gets new scrambles. Whatever you personally believe about whether it's appropriate to give arbitrarily easy non-solved positions, we definitely do need to have an official ruling so it isn't up to the delegate's opinion of how easy is too easy.

 

[Kirjava]

The most important reason to draw some arbitrary line is fairness.

You create the same fairness by not filtering scrambles. It's just right to decide either way, it's not a advantage that only filtering has.

 

[qqwref]

Yes, that's an option too, as long as we can get every delegate to do exactly that. I just thought it was necessary to remind people of why we want to make *some* choice - it's not just some pedantic discussion about whose arbitrary cutoff better matches reality.

 

[Sebastien]

I still wonder why there is no official WCA statement about delegates removing easy scrambles. I fortunately didn't have to deal with such occurances so far, but as long as there is no official guideline I would surely allow every kind of scramble.

 

[Kirjava]

it's not just some pedantic discussion about whose arbitrary cutoff better matches reality.

I don't consider my cutoff arbitrary.

 

[cmhardw]

Based on discussion in both threads I think that the best options so far are (as I read it these were made by Erik&Godmil and Stefan respectively):

1) A "scramble" is defined as a "DNF" state (or a state that is not solved).
2) A "scramble" is defined as an "non-trivial" state (a state "difficult" to solve by a non-cuber. For argument's sake, this limit is set as any scramble greater than 3 turns in length).

I feel that if a change is to be made, it should be one of these two. That is only my personal opinion.

I think allowing "+2" states should not be allowable, as that is allowing a "solved" state to be possible scramble. For the same reason the solved state of the puzzle should not be an allowable "scramble" for that puzzle.

 

[StachuK1992]

I think allowing "+2" states should not be allowable, as that is allowing a "solved" state to be possible scramble. For the same reason the solved state of the puzzle should not be an allowable "scramble" for that puzzle.

Agreed.
There's no reason to make this overly complicated - surely there are going to be some people whose idea of a 'legal scramble' are a bit different than others.

At this point, it's just up to the board to address the issue.

 

[Stefan]

Based on discussion in both threads I think that the best options so far are (as I read it these were made by Erik&Godmil and Stefan respectively):

1) A "scramble" is defined as a "DNF" state (or a state that is not solved).
2) A "scramble" is defined as an "non-trivial" state (a state "difficult" to solve by a non-cuber. For argument's sake, this limit is set as any scramble greater than 3 turns in length).

Meh, I already suggested both of these three years ago:

"new suggestion: Discard scrambles that would count as solved or solved with penalty."

"I think I'd vote for either no discardings or only trivial ones, let's say "scrambles solvable in three moves or less", which even non-cubers could solve."

I was only looking for the first of those two, I had completely forgotten about the second one and I was like "holy ****!" when I read that again

Anyway, so that's an older thread if someone's interested in previous discussion about this...