Why? When it comes to the legitimacy of what we recognize as world records, I think it should be explainable to non-cubers. If someone got a 2-move 2x2x2 world record single and that person happened to not be very experienced with 2x2x2, but enough to get a fast time on that scramble, it would probably be very difficult to convince a non-cuber that these records are legitimate and fair. It would also make non-cubers question the integrity of the WCA if we allow scrambles that are ridiculously short.
FYI: I think the current 2x2x2 WR was fair, because it was 4 moves and not 2. In the end, the faster 2x2x2 solvers out of those who received that scramble achieved the fastest times.
Why? When it comes to the legitimacy of what we recognize as world records, I think it should be explainable to non-cubers. If someone got a 2-move 2x2x2 world record single and that person happened to not be very experienced with 2x2x2, but enough to get a fast time on that scramble, it would probably be very difficult to convince a non-cuber that these records are legitimate and fair. It would also make non-cubers question the integrity of the WCA if we allow scrambles that are ridiculously short.
FYI: I think the current 2x2x2 WR was fair, because it was 4 moves and not 2. In the end, the faster 2x2x2 solvers out of those who received that scramble achieved the fastest times.
Ok, but what about the fast 2x2ers who weren't given that scramble? It's never going to be completely fair. Luck is an element in every sport. I honestly think scramble filtering is unnecessary. Single WRs for short event will be fast, but people will recognize that it is not even remotely based on who has the most skill. It has already come to that point with 2x2 and Square-1 - the person with the most skill doesn't have to have the Single WR, that is why we do averages.
Ok, but what about the fast 2x2ers who weren't given that scramble? It's never going to be completely fair. Luck is an element in every sport. I honestly think scramble filtering is unnecessary. Single WRs for short event will be fast, but people will recognize that it is not even remotely based on who has the most skill. It has already come to that point with 2x2 and Square-1 - the person with the most skill doesn't have to have the Single WR, that is why we do averages.
Whether we choose to personally acknowledge certain single WRs, they are still officially recognized as WRs, and non-cubers can view them. Regarding the fast 2x2x2 solvers who don't get lucky scrambles, it's unfortunate and not preventable since luck is an inevitable element. I know it's impossible to be completely fair so that the most skilled cubers are the only ones who get the WRs, but what we can do is prevent WRs that almost anyone can get if they're just lucky enough. All I'm trying to get at is that we should not simply have >1 move as the filters (I've repeated my arguments several times).
my point was that I doubt any 5+ second solver could execute 5 moves sub 1.36 (as Tim said). This means that if you lower the limit to 5 moves, it would still take some skill to get a WR. (as opposed to 2x2, where the WR single holder is 365th for average)
It's not just the #1. Here's a comparison of the top 100 for single along with their world rank for 2x2 and Pyraminx. Pyraminx top 9 look pretty good. Wondering whether it's because of the puzzles, the different scramble filter limits, just random, or something else.
select two.name, single222, avgRank222, pyr.name, singlePyra, avgRankPyra from
(select @ctr:=@ctr+1 ctr, name, round(s.best/100,2) single222, a.worldRank avgRank222
from (select @ctr:=0) init, RanksSingle s join RanksAverage a on s.personId = a.personId and s.eventId = a.eventId join Persons p on s.personId = p.id
where s.eventId = '222' and s.worldRank <= 100
order by s.worldRank) two,
(select @ctr2:=@ctr2+1 ctr, name, round(s.best/100,2) singlePyra, a.worldRank avgRankPyra
from (select @ctr2:=0) init, RanksSingle s join RanksAverage a on s.personId = a.personId and s.eventId = a.eventId join Persons p on s.personId = p.id
where s.eventId = 'pyram' and s.worldRank <= 100
order by s.worldRank) pyr
where two.ctr = pyr.ctr;
I don't understand the problem with just letting singles be singles? If we want to have some concept of 'adequately scrambled' then why not just say, make it God's number for any of the cubes and remove all scrambles with optimal solutions with lengths below that? We play a game that has luck built into it, why make some arbitrary rules to deny that? I really couldn't care less what the casual observer has to say about the game and what 'looks easy', and I don't see what is so unique to cubing that means that we should listen to the uninformed when thinking of regulations.
I really don't understand how it's so hard to adapt, is there some cognitive dissonance or something? Single times have always been basically meaningless; take my 7.71 which was placed 5th in the world at the time of setting and was a ranking I 'didn't deserve' if you consider it anywhere near as meaningful as average. It seriously means next to nothing and the only people for whom it is forgiveable to not know that are people who've never solved the cube before.
If we want to have some concept of 'adequately scrambled' then why not just say, make it God's number for any of the cubes and remove all scrambles with optimal solutions with lengths below that?
For a puzzle like Skewb, God's number would be 11. There are only 90 positions 11 moves away from the solved state on a Skewb. One could simply just memorize 90 algorithms.
It's not just the #1. Here's a comparison of the top 100 for single along with their world rank for 2x2 and Pyraminx. Pyraminx top 9 look pretty good. Wondering whether it's because of the puzzles, the different scramble filter limits, just random, or something else.
select two.name, single222, avgRank222, pyr.name, singlePyra, avgRankPyra from
(select @ctr:=@ctr+1 ctr, name, round(s.best/100,2) single222, a.worldRank avgRank222
from (select @ctr:=0) init, RanksSingle s join RanksAverage a on s.personId = a.personId and s.eventId = a.eventId join Persons p on s.personId = p.id
where s.eventId = '222' and s.worldRank <= 100
order by s.worldRank) two,
(select @ctr2:=@ctr2+1 ctr, name, round(s.best/100,2) singlePyra, a.worldRank avgRankPyra
from (select @ctr2:=0) init, RanksSingle s join RanksAverage a on s.personId = a.personId and s.eventId = a.eventId join Persons p on s.personId = p.id
where s.eventId = 'pyram' and s.worldRank <= 100
order by s.worldRank) pyr
where two.ctr = pyr.ctr;
Even though I am more inclined to think that an arbitrary number set for each puzzle may not be the perfect solution, I think the data you show is interesting enough to see it as a chart that compares single rank vs average rank for top 100 people in both events. The people with ranking > 100 are put at the top so the chart can remain somewhat readable (even though it didn't turn out that clear overall).
Thanks, I had actually typed much more because I made mistakes, and when I *almost* reached your number, I was paranoid enough that I wanted know whether I was still doing something wrong
If we want to have some concept of 'adequately scrambled' then why not just say, make it God's number for any of the cubes and remove all scrambles with optimal solutions with lengths below that?
Why would we want to throw out the vast majority of positions for obviously no reason?
Also, this would lead to only 32 possible scrambles for Pyraminx (ignoring tips), and even fewer if some of them are equivalent. Similar for Skewb. You could easily learn algs for all cases and recognize them during inspection, so the solve is just raw execution of a previously learned alg. Basically you're turning these puzzles into Magic.
Finally, we simply can't. For some cubes we don't even know God's number.
If anyone is denying that luck plays a role, I'd say it's those who don't want to do something about it. Those who are trying to do something about it are acknowledging luck exactly by trying to do something about it.
Also, they're not attempts to rule luck out, only to mitigate it a bit. And don't say we shouldn't be trying to mitigate luck. We're doing it with the averages, and you do seem to like those.
I don't see why you're talking about uniqueness. If you just want to ask why we should listen to the uninformed when thinking of regulations, I repeat my argument that they're still a measure for what's a puzzle. I haven't been puzzled by our regular puzzles in a long time. You?
I really don't understand how it's so hard to adapt, is there some cognitive dissonance or something? Single times have always been basically meaningless
Not sure about "cognitive dissonance", but I do think you have an odd perspective. If people ask or are asked what the Rubik's Cube world record is, I believe they're more likely thinking about single than average. Even cubers. Even the WCA site is evidence of that, as the default view shows singles.
Even though I am more inclined to think that an arbitrary number set for each puzzle may not be the perfect solution, I think the data you show is interesting enough to see it as a chart that compares single rank vs average rank for top 100 people in both events.
Pardon my english. I meant to say that even though I don't entirely agree with your view, I think some of the points you give and data you show are interesting to look into and explore.
Pardon my english. I meant to say that even though I don't entirely agree with your view, I think some of the points you give and data you show are interesting to look into and explore.
Ah, ok. I could have maybe seen that. Your English was fine. But I guess I was only looking for a direct connection between the contents of the two sentence parts (I hope this makes sense, but doesn't matter if not ).
It's not just the #1. Here's a comparison of the top 100 for single along with their world rank for 2x2 and Pyraminx. Pyraminx top 9 look pretty good. Wondering whether it's because of the puzzles, the different scramble filter limits, just random, or something else.
That's pretty interesting. However, data doesn't answer the question of whether we want/need the rankings to match.
That is, it doesn't answer whether we should be enforcing skill using a particular way of scramble filtering.
I find it interesting that this conversation is focused on filtering by number of moves when historically there have been lots of (unusable but well-intentioned) suggestions for fancier rules (e.g. number of moves for cross, number of pieces solved) -- including one for BLD that I remember posting on the Yahoo list a long while back. Has the current policy framed everyone's view, or does everyone accept that move count filtering probably the only fair way to do it?
That's pretty interesting. However, data doesn't answer the question of whether we want/need the rankings to match.
That is, it doesn't answer whether we should be enforcing skill using a particular way of scramble filtering.
I find it interesting that this conversation is focused on filtering by number of moves when historically there have been lots of (unusable but well-intentioned) suggestions for fancier rules (e.g. number of moves for cross, number of pieces solved) -- including one for BLD that I remember posting on the Yahoo list a long while back. Has the current policy framed everyone's view, or does everyone accept that move count filtering probably the only fair way to do it?
It doesn't answer that question, but is that really a question? Want? Yes. Who would not prefer rankings to reflect well how good people actually are, in the sense of the single ranking "matching" the average ranking? Need? No, we don't "need" to do anything.
It might help answer the question of how to achieve matches, though. If the pyraminx top ten match so much better than the 2x2 top ten because of something specific, it *could* maybe be used for the 2x2 as well. Of course it comes at the price of doing this something, and if that's for example scramble filtering then there are people against it, so that's a disadvantage. So I'd say the real question isn't whether we want matches, but whether we want them enough to outweigh the possible disadvantages.
I find it interesting that this conversation is focused on filtering by number of moves when historically there have been lots of (unusable but well-intentioned) suggestions for fancier rules (e.g. number of moves for cross, number of pieces solved) -- including one for BLD that I remember posting on the Yahoo list a long while back. Has the current policy framed everyone's view, or does everyone accept that move count filtering probably the only fair way to do it?
I did notice that as well. Indeed interesting. Probably mostly being framed by the current policy. And newcomers don't knowing the old discussions. I don't think it's about the kind of fairness you're thinking off, at least I simply didn't think of such other filters during this discussion and so I didn't think about their fairness implications at all. And as my stance is related to "solvability by non-cubers", a case like a single T-perm as scramble for 3x3 is simply irrelevant to me, as non-cubers usually can't solve that.
Everyone in this discussion should take a good long look in the mirror, and realise what is actually causing all the problems --> all the stupid little "puzzles" that are solvable in obscenely short periods of time. Instead of trying to deal with them, it would be so much simpler to either:
1) Remove these events
2) Stop recognising single solves as "records"
Everyone in this discussion should take a good long look in the mirror, and realise what is actually causing all the problems --> all the stupid little "puzzles" that are solvable in obscenely short periods of time. Instead of trying to deal with them, it would be so much simpler to either:
1) Remove these events
2) Stop recognising single solves as "records"
Everyone in this discussion should take a good long look in the mirror, and realise what is actually causing all the problems --> all the stupid little "puzzles" that are solvable in obscenely short periods of time. Instead of trying to deal with them, it would be so much simpler to either:
1) Remove these events
2) Stop recognising single solves as "records"
Regarding scramble filtering, there are actually two questions to discuss - preemptive filtering and retroactive filtering.
Preemptive scramble filtering:
Imagine a scramble that is widely seen as 'easy', but has a sufficiently long optimal solve to be valid under preemptive filtering. For example, superflip. Consider the response to the appearance of such a scramble at competitions in two parallel universes, one where scramble filtering occurs and another where it does not.
In the universe where there is no scramble filtering, the response need only be 'All cube states are potential scrambles, and everyone has an equal chance of encountering scramble states of varying subjective difficulty.' There needn't be controversy, as all scramble states are treated equally.
In the universe where scramble filtering does occur, the response is likely more complex. 'By filtering some scrambles, we presuppose that there are some scramble states that we find unacceptable. However, there are many, many more states that some cubers may be unhappy to see occur. Eliminating scrambles based on optimal solve length attempts to address these 'easy scrambles', but it is impossible to satisfy everyone. Any attempt to eliminate these cases will leave some people unhappy.'
Optimal solve length filtering is far from eliminating cases that would commonly be seen as trivial. It is impossible to define an exhaustive set of 'too easy' scrambles that all cubers would agree with. Scramble filtering serves to cause controversy by acknowledging that there are some scramble states that are unacceptable, yet being unable to prevent some of them from being encountered. I'd rather our position be that all scramble states are acceptable, than that some are unacceptable but unavoidable.
Which leads to the next issue. Retroactive filtering:
The problem caused by preemptive scramble filtering potentially lead to retroactive filtering. The issue here is the disqualification of solves that were done on properly generated scrambles and were conducted properly following all regulations, on the grounds that someone finds the result unsatisfactory. That is, disqualification of otherwise valid results because someone feels that a properly-generated random state scramble resulted in an undeserved result.
Consider again an 'easy' but otherwise valid scramble, like the superflip. In the case of preemptive filtering, where it is acceptable to not allow some scramble states on the ground that they are 'too easy', it is not hard to imagine some people being very unhappy with a differently 'easy' scramble, and demanding its removal. In a world where there is no scramble filtering, there is no controversy, as all potential states are valid.
The last time I brought up this issue, I asked the regulations folks to add language confirming that all solves on properly generated scrambles that otherwise follow regulations are final. This was rejected because one board member insisted that the power to invalidate previously recognized solves be reserved as a personal privilege. (I say personal, rather than board privilege, as the other board members weighing in at the time were also opposed to after the fact filtering.)
The argument that these short scrambles are so rare as to make it not worth arguing about filtering them seems to me to rather work the other way. As short scrambles are very unlikely to occur in a random-state scramble generator, filtering prevents something that will almost certainly never happen in the first place while at the same time causing problems by suggesting that some scrambles are unacceptable and opening the door for arbitrary after the fact disqualification of results.
I just read through this thread, and I'm pretty sure I don't have anything to say that hasn't already been said. I would like to say that nobody involved in making or implementing these filtering decisions has been trying to sneak anything past the community.
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...
I'm not sure what you're getting at here. If you're suggesting that changes to TNoodle drove the board/WRC to change the regulations, I want it to be clear that that was very much not the case. Filtering rules always come from the WRC/board to TNoodle, not the other way around. Originally, the WRC asked us to filter out states that are considered solved or solved without filtering, and so we did. At the time, it was a ton of work I would have rather not done (this is not a reflection on how I feel about filtering, it is just a complaint that it was a ton of work). It's unfortunate that the regulations were not immediately updated to reflect this, but I'm positive that was only because everyone involved is a volunteer and has other priorities, not because anyone was intentionally being secretive. In my experience, documentation always lags behind reality, not just for software, but for anything.
I encourage anyone who wants to stay informed on discussions about TNoodle to join the wca-scrambler group: https://groups.google.com/forum/#!forum/wca-scrambler. There's also the github repository here https://github.com/cubing/tnoodle/, which I would love to see more members of the community contribute to. I would like to emphasize that TNoodle is not necessarily the right place for debates about whether filtering should exist, or how much of it we should do. It is a place to discuss how best to fix TNoodle bugs and add new features, many of which come from the WRC or the board. If anyone is confused about TNoodle's role in the cubing community, or has any suggestions about how to better open things up to people, feel free to start a thread on the wca-scrambler group, or contact me personally.
