Proposal: Video Evidence Should NOT be used to overturn official records

[Carrot]

Wasn't the main issue that discovering misscrambles using video evidence has lead to DNFs? So we should be ensuring as close to 100% accuracy as possible on scrambles to avoid giving unfair advantages or vice versa?

discovering misscrambles AFTER the competition leads to DNFs, I am pretty sure most delegates would allow you to get an extra attempt given someone spots it during the competition.

Sure your proposal is more accurate, but it surely also gives more away of the puzzles state, you need to draw the line somewhere.

 

[Ollie]

discovering misscrambles AFTER the competition leads to DNFs, I am pretty sure most delegates would allow you to get an extra attempt given someone spots it during the competition.

Sure your proposal is more accurate, but it surely also gives more away of the puzzles state, you need to draw the line somewhere.

I'm not sure if revealing 5 stickers out of a possible 48 (not counting centers) gives a significant advantage to the competitor over 4 stickers on one face for 3x3x3 upwards. Yes, admittedly it is a much bigger problem for 2x2x2 scrambles.

The irony is that the 2x2x2 scrambles are the shortest so you'd think mistakes during scrambling are less likely, but they also need the highest accuracy so mistakes aren't as tolerable.

 

[DeeDubb]

If the scramble puts the cube under the required movecount (4 for 2x2), and it's not caught in the comp, it has to be DNF'd. If it gives them an advantage, and can be deemed as intentional (which should be obvious on video), it should also be DNF'd and an investigation must happen. If it gives them an advantage, but is deemed unintentional (like Riley's at home UWR), it should NOT be DNF'd.

 

[Kirjava]

(or even the competitor, as the visible information doesn't give away much, at least for 3x3 upwards)

that idea makes me uncomfortable

 

[cubizh]

The idea of exposing certain parts of a puzzle is a bit unfeasible given the different variety of puzzle sizes and especially of puzzle types.
It's too much hassle just to do something that the scrambler should already be doing, that is verifying if the puzzle is properly scrambled.
I think extra awareness should be raised towards people who are given the task of scrambling that it's a two step task and not just apply written moves.

 

[Kit Clement]

What would people think of having a third box on the scoresheet for the scrambler to sign? It would let us know who scrambled for each attempt, and be a way of reminding the scramblers that they're responsible for correctly scrambling.

I like this idea -- going to implement this somehow in the future. Holding individual scramblers accountable may encourage scramblers to do their job correctly, and also informs organizers as to who might be making scrambling mistakes.

 

[kcl]

I like this idea -- going to implement this somehow in the future. Holding individual scramblers accountable may encourage scramblers to do their job correctly, and also informs organizers as to who might be making scrambling mistakes.

I second this. It's a super simple idea but I can't believe I didn't think of it.

 

[Mike Hughey]

Exposing just parts of the puzzle like Stefan showed would only work if the competitor had to do the checking, since no one else would know the proper colors for unusual color schemes - it would simply look like the wrong scramble. And even if the competitor did the checking, we'd have to rely on every competitor with an unusual color scheme knowing the proper color correlation, which would be unlikely for new competitors. I don't think that idea is practical, unless we force everyone to switch to the same color scheme (which I sincerely hope we do not do, as it would mean starting over on BLD for me).

 

[qqwref]

1. If a 2x2 was scrambled wrong and a 1 move solve was done should it be allowed?
A2. If a 3x3 was scrambled wrong and a 15 move sub4 solve was done should it be allowed?
A3. What if that solve was in the WC final and would cause them to drop from 1st to 4th if DNFed?
A4. If a 3x3 was scrambled wrong and a normal solve by an average cuber was done, should it be allowed?

During a competition, if the misscramble is noticed, none of these should be allowed. It's against regulations, right? The solve should be removed and the competitor should be given a replacement solve, just like if they were given the same scramble twice, or handed a solved cube.

The problem is when, after a competition, someone gets a DNF due to a situation completely out of their control. Then there is no time to give them a replacement solve, so getting rid of the affected solve is a potentially huge penalty (especially given that we only even look at solves that are good enough to be on video, i.e. mostly regional records or official PBs). In case A1, the delegate and/or judges are idiots and should probably be "fired". In the other cases, the competitor performed a legitimate solve of a clearly scrambled cube that they had no prior knowledge of - let it stand, and if you want to punish someone, punish the scrambler.

Aside from those situations which are not the competitor's fault, we have:
B1. Regulation violations that are the competitor's fault, were accidental, and (at least) might have given an advantage
B2. Regulation violations that are the competitor's fault, were accidental, and did not give an advantage
B3. Regulation violations that are the competitors fault, were deliberate, and (at least) might have given an advantage
B4. Regulation violations that are the competitors fault, were deliberate, and did not give an advantage

B1, B2: Who cares? Accidents happen.
B3: Cheating. Disqualify an appropriate number of solves, and suspend competitor for an appropriate amount of time, depending on the actual situation.
B4: Not sure why someone would do this. But if you are sure it gives no advantage, give the competitor an official warning not to do it again, and take further action only if they keep doing it.



Re: exposing parts of the puzzle. Are you guys serious? What's the point of a cover and timed inspection then? What if someone uses a non-Fridrich method and this actually helps them out? And what do you plan to do with people with alternate color schemes (e.g. Mike Hughey scheme which swaps orange<->red and yellow<->blue from the perspective of holding white/green steady)? I guarantee this will cause more problems than it solves.

 

[BaMiao]

The problem is when, after a competition, someone gets a DNF due to a situation completely out of their control. Then there is no time to give them a replacement solve, so getting rid of the affected solve is a potentially huge penalty (especially given that we only even look at solves that are good enough to be on video, i.e. mostly regional records or official PBs). In case A1, the delegate and/or judges are idiots and should probably be "fired". In the other cases, the competitor performed a legitimate solve of a clearly scrambled cube that they had no prior knowledge of - let it stand, and if you want to punish someone, punish the scrambler.

So if Martín Telésforo had claimed that his 4.41 simply came from a misscramble, would you have allowed that "record" to stand? We would not have had any reason to suspect foul play at the time. Remember, the only justification for disqualifying that solve was because of the bogus solution to the official scramble that he presented. Allowing misscrambled times to stand might open us up to this kind of cheating in the future.

 

[TMOY]

That should maybe be done anyway (I only say maybe because I don't know how big the problem is). There were some ideas, like someone suggested adding a scramble checker person between scramblers and runners/judges. Another idea to allow rudimentary checking for the judge (or even the competitor, as the visible information doesn't give away much, at least for 3x3 upwards):

Spoiler

http://i58.tinypic.com/oh6d1h.jpg

Nµice idea, but how do you deal with nonstandard color schemes ?

Edit: ninja'd

 

[guysensei1]

How about... If a WR/CR/NR is broken and it's on video, immediately verify the scramble on the spot?

 

[tseitsei]

Exposing just parts of the puzzle like Stefan showed would only work if the competitor had to do the checking, since no one else would know the proper colors for unusual color schemes - it would simply look like the wrong scramble. And even if the competitor did the checking, we'd have to rely on every competitor with an unusual color scheme knowing the proper color correlation, which would be unlikely for new competitors. I don't think that idea is practical, unless we force everyone to switch to the same color scheme (which I sincerely hope we do not do, as it would mean starting over on BLD for me).

The competitor doing the checking is the whole point of this thing. Because if competitor does the checking and the scramble is found to be wrong after the competition, we can now say that competitor should have noticed it and we can fairly DNF the solve, because competitor had a chance to notice the wrong scramble and demand a correct one (unlike they have now).
Different color schemes don't matter since every competitor should know their scheme and regulations when competing.

Re:1. exposing parts of the puzzle. Are you guys serious? What's the point of a cover and timed inspection then?

2.What if someone uses a non-Fridrich method and this actually helps them out?

3. And what do you plan to do with people with alternate color schemes (e.g. Mike Hughey scheme which swaps orange<->red and yellow<->blue from the perspective of holding white/green steady)? I guarantee this will cause more problems than it solves.

1. The point of cover is to cover the rest of the puzzle and the point of inspection is to see the rest of the puzzle obviously.
2. I REALLY doubt that seeing 4 stickers from different pieces will help anyone significantly. (At least outside 2x2, so maybe for 2x2 this is not a good idea but for other cubid puzzles I think it could be)
3. Already answered earlier in this message.


Only real problem I see is having sufficient covers since people can have different sized cubes...

 

[Erik]

I like this idea -- going to implement this somehow in the future. Holding individual scramblers accountable may encourage scramblers to do their job correctly, and also informs organizers as to who might be making scrambling mistakes.

Please don't!

This will burden the scramblers with too little to gain.

Scrambling 2x2's or Pyraminxes or 3x3's is already very hectic now, where most of the time is actually spend on arranging the score sheet and putting the cube in the covers, checking which cube needs scrambling next etc.

Having to sign every cube as well is just too much of a hassle, I don't want to have to sign 150 2x2 scrambles for group A and then another 150 for group B (assuming I have 30 cubes).

Besides, what will you do if you find out after a competition that scrambler XY scrambled wrong? Warn him? Never allow him to scramble again? What is it that you expect to gain? Encouragement for scramblers is a bad reason to implement something like this. There are many other ways you can do this which are much more clever and less of a burden.

For instance: encourage scramblers to glance at their co-scramblers and do a quick check if they scrambled correctly (I always do that when I am scrambling) or announce that at any time a delegate/runner/random person can pick a random scrambled cube and check if it's correct or not.

The signing by the judge and competitor thing already is a mistake in the regulations. It just doesn't work. You can't read who signed it, people just make an X or a smiley or put down their autograph. In no way this can be traced back to the correct judge. You can maybe force the competitor to sign correctly since this can be checked, but not the judge. This burocracy should definitely not be expanded more.


Also: why do you say something like this before we have an idea about how big the problem actually is? If you do have numbers, please share.

 

[Logiqx]

Can someone develop a phone app which uses the camera to check the scrambled puzzle?

A single photo could verify 3 sides of a cube (e.g. FRU) which is probably going to pick up scrambling errors.

Custom colour schemes could also be handled automatically with some simple logic.

 

[tseitsei]

Please don't!

This will burden the scramblers with too little to gain.

Scrambling 2x2's or Pyraminxes or 3x3's is already very hectic now, where most of the time is actually spend on arranging the score sheet and putting the cube in the covers, checking which cube needs scrambling next etc.

Having to sign every cube as well is just too much of a hassle, I don't want to have to sign 150 2x2 scrambles for group A and then another 150 for group B (assuming I have 30 cubes).

Besides, what will you do if you find out after a competition that scrambler XY scrambled wrong? Warn him? Never allow him to scramble again? What is it that you expect to gain? Encouragement for scramblers is a bad reason to implement something like this. There are many other ways you can do this which are much more clever and less of a burden.

For instance: encourage scramblers to glance at their co-scramblers and do a quick check if they scrambled correctly (I always do that when I am scrambling) or announce that at any time a delegate/runner/random person can pick a random scrambled cube and check if it's correct or not.

The signing by the judge and competitor thing already is a mistake in the regulations. It just doesn't work. You can't read who signed it, people just make an X or a smiley or put down their autograph. In no way this can be traced back to the correct judge. You can maybe force the competitor to sign correctly since this can be checked, but not the judge. This burocracy should definitely not be expanded more.


Also: why do you say something like this before we have an idea about how big the problem actually is? If you do have numbers, please share.

+1

I agree with this

 

[kinch2002]

I think exposing parts of puzzles is a very bad idea (no offense). As well as different colour scheme problems, you also have to put the cube in a non-random orientation. Even without those problems I totally object on the grounds of...well...can't think of the word, but it just seems 'wrong' to show the puzzle

During a competition, if the misscramble is noticed, none of these should be allowed...

Agreed. My questions were posed with the stated assumption that it wasn't noticed during the competition as I don't think there's any debate what should happen if it is noticed

The problem is when, after a competition, someone gets a DNF due to a situation completely out of their control...
...let it stand, and if you want to punish someone, punish the scrambler.

Interesting...but do you have a suggestion of how to punish the scrambler?

B1. Regulation violations that are the competitor's fault, were accidental, and (at least) might have given an advantage
B2. Regulation violations that are the competitor's fault, were accidental, and did not give an advantage

B1, B2: Who cares? Accidents happen.

B1: You sure about that? What if the competitor stops the timer early by mistake and neither of them notice?

 

[Seanliu]

I think all should be filmed, so not 'Video evidence should not be used' , and 'All should be filmed and checked for errors'

 

[Dene]

I like this idea -- going to implement this somehow in the future. Holding individual scramblers accountable may encourage scramblers to do their job correctly, and also informs organizers as to who might be making scrambling mistakes.

I agree completely with Erik, but something else I wanted to add:

How often exactly do you notice a mis-scramble anyway? It's such a rare occurrence... I can't see the extra workload being nearly worth the tradeoff. That is, unless you have a significant problem with mis-scrambles at your competitions. But if that is the case, there are much better ways to address the problem.

 

[(X)]

How about having a person that sits at the scramble table and only checks if the scrambles are correct. That person could then cycle the puzzles that are misscrambled directly back to the scramblers.