3x3 BLD format in competitions

[PatrickJameson]

I...kind of accidentally permanently deleted the other thread instead of soft deleting it to reformat it...but here's a thread to continue talking about it.

*goes to hide in a corner*

(If anyone has those threads in a tab or something and can paste the info here that'd be nice)

 

[Kirjava]

tl;dr best of 5 would be cool

 

[ExoCorsair]

(If anyone has those threads in a tab or something and can paste the info here that'd be nice)

Before anyone asks, these posts are not recoverable unless an admin says otherwise.

 

[yoruichi]

get best out of 5!!!
that way we dont have to do like safety solves or w.e

 

[Toad]

Why are thresholds not possible...?

Just go for something like:

Best of 3.
Best 5 if 1 successful solve sub2 in first 3 solves or 2 DNFs sub2.

I dunno, I think it could work...?

 

[amostay2004]

Why are thresholds not possible...?

Just go for something like:

Best of 3.
Best 5 if 1 successful solve sub2 in first 3 solves or 2 DNFs sub2.

I dunno, I think it could work...?

That would definitely not work lol.
"oshi I can't sub2 this I better DNF and get more tries..."

 

[Toad]

Why are thresholds not possible...?

Just go for something like:

Best of 3.
Best 5 if 1 successful solve sub2 in first 3 solves or 2 DNFs sub2.

I dunno, I think it could work...?

That would definitely not work lol.
"oshi I can't sub2 this I better DNF and get more tries..."

Ok how about extra tries for a single sub2 success?

 

[JeffDelucia]

Beforehand I apologize because I don't know what conversation came before this.

I think that best of 5 would be great because it gives more opportunities for a success (duh). However, there would definitely need to be some cutoff system because some people (including myself) can take up to 10 minutes for just a single blind solve. I think maybe a system where you are only allowed like 30 minutes of blind solving total instead of a cutoff per solve. This would avoid confusions with DNFs when deciding who to cut off. Also it would allow some slower cubers to go over 10 minutes. For instance say there are 3 cubers at a competition. Cuber #1 averages 3 minutes blind. Cuber #2 averages 9 minutes blind. Cuber #3 averages 20 minutes blind. #1 would get 5 solves and only take ~15 minutes. #2 would get 3 solves and take ~27 minutes. #3 would get 1 solve and take 20 minutes. A setup like this would benefit both the faster cubers by giving them more solves and the slower cubers by giving them solves in the first place.

 

[Gunnar]

Beforehand I apologize because I don't know what conversation came before this.

I think that best of 5 would be great because it gives more opportunities for a success (duh). However, there would definitely need to be some cutoff system because some people (including myself) can take up to 10 minutes for just a single blind solve. I think maybe a system where you are only allowed like 30 minutes of blind solving total instead of a cutoff per solve. This would avoid confusions with DNFs when deciding who to cut off. Also it would allow some slower cubers to go over 10 minutes. For instance say there are 3 cubers at a competition. Cuber #1 averages 3 minutes blind. Cuber #2 averages 9 minutes blind. Cuber #3 averages 20 minutes blind. #1 would get 5 solves and only take ~15 minutes. #2 would get 3 solves and take ~27 minutes. #3 would get 1 solve and take 20 minutes. A setup like this would benefit both the faster cubers by giving them more solves and the slower cubers by giving them solves in the first place.

Can't we just treat BLD as any other event and allow for combined finals? Then you may have a threshold, either a time limit or a number of competitors, for whom will get a 2nd, 3rd, 4th or 5th attempt. The exact thresholds should be up to the organizer of the competition as it is in other events.

I think that allowing for up to best of 5 is a good idea, since other events taking roughly the same ammount of time (~1-3min) may have 5 attempts in a round (average of 5). This would, I think, give a more equal opportunity for BLDers as people focusing on other events, to perform in competitions.

Again, the details should be up to the organizer, depending of how much time they are willing to set off for the event.

 

[qqwref]

I think it would be nice to have up to 10 minutes, total, to do up to 5 blindfold solves. You could round times to the nearest 5 or 10 seconds to make it easier to calculate. For a DNF you would use the time when the timer was stopped. The scorecard would have to have an extra space for time remaining. When you reach the remaining time on a solve, you would be automatically cut off (DNF'd).

 

[amostay2004]

I think a total time limit for 5 solves is a great idea, though 10 minutes might be a bit harsh for beginners. I have little competition experience but in Malaysian Open people have like 7-9 minutes successes so in this format they would've run out of time by the second solve. Probably 20 to 30 minutes could be allocated, but then again the time limit should be a flexible thing, like how cut off times are decided.

 

[keemy]

I think it would be nice to have up to 10 minutes, total, to do up to 5 blindfold solves. You could round times to the nearest 5 or 10 seconds to make it easier to calculate. For a DNF you would use the time when the timer was stopped. The scorecard would have to have an extra space for time remaining. When you reach the remaining time on a solve, you would be automatically cut off (DNF'd).

This is allowed by the rules already (as with most of the other purposed formats people have said). Err an interesting format I don't think is explicitly allowed by the regs is "up to [time] rounded down" which would be similar to what you said but instead of cutting them off during their last solve when they go over limit letting them finish.

I think this would be nicer logistically because then you don't have to have the judge remember to stop them after a specific amount of time which they need to calculate and remember. This on the other just needs to be checked at the end of the solve(you can even do the estimate to nearest 5 seconds or w/e for easier addition) though it will be about equivalent for most competitors (if like instead of "10 min stage time" you have "7 min stage time rounded down").

 

[Tim Reynolds]

I think it would be nice to have up to 10 minutes, total, to do up to 5 blindfold solves. You could round times to the nearest 5 or 10 seconds to make it easier to calculate. For a DNF you would use the time when the timer was stopped. The scorecard would have to have an extra space for time remaining. When you reach the remaining time on a solve, you would be automatically cut off (DNF'd).

Here's one problem I see with that: Suppose someone's a 3-minute solver. They're not going to get more than 3 full attempts, right? Say they don't like parity. They can (and should) quickly count the parity on the first solve, or look for some other indication of easiness, and then skip it if it's a "hard" solve. To the extreme, an 8-minute solver could make the scramblers solve and scramble his cube 4 times when he had no intention of doing more than one solve. It's in your best interest to choose the easiest solve. In my opinion (and I am aware that this is just that, an opinion), a person should not have incentive to get a worse result on any earlier solve in order to get a better overall result. Whether this goal is obtainable, I'm not sure, unless we do a straight-up combined final, where the top x successes get to keep going.

 

[qqwref]

Well, some currently used BLD formats already have that property. I've been at a competition where you would get a 2nd and 3rd solve if your first was under a certain time, DNF or not. So if you were slow, it would be in your best interest to DNF quickly, because then you could have 2 full attempts instead of 1.

To be honest, I'm not really sure I see a problem with either (a) rewarding people for a bad result, or (b) asking scramblers to do extra work - at least not in the 3BLD round. Issue (a) isn't a problem because nobody will get 100% successes at full speed, and so if you want to do any kind of combined-final type round but not introduce too much luck you basically have to look at the time of the solve and ignore whether it was a success. Of course, by that metric a 2-3 minute DNF is definitely better than a 6 minute success. Issue (b) isn't a problem because BLD solves take so long (2+ minutes for a typical competitor) compared to the scramble time (<10 seconds - negligible for all intents and purposes) that at all but the largest competitions you can have a single scrambler and they won't even be very busy. In fact, as someone who often helps out at the scrambling table, I'd rather apply a lot of extra 3x3 scrambles than wait around for the last solves of the inevitable slow solver who didn't start until the round was pretty much already over.

 

[Tim Reynolds]

Well, some currently used BLD formats already have that property. I've been at a competition where you would get a 2nd and 3rd solve if your first was under a certain time, DNF or not. So if you were slow, it would be in your best interest to DNF quickly, because then you could have 2 full attempts instead of 1.

To be honest, I'm not really sure I see a problem with either (a) rewarding people for a bad result, or (b) asking scramblers to do extra work - at least not in the 3BLD round. Issue (a) isn't a problem because nobody will get 100% successes at full speed, and so if you want to do any kind of combined-final type round but not introduce too much luck you basically have to look at the time of the solve and ignore whether it was a success. Of course, by that metric a 2-3 minute DNF is definitely better than a 6 minute success. Issue (b) isn't a problem because BLD solves take so long (2+ minutes for a typical competitor) compared to the scramble time (<10 seconds - negligible for all intents and purposes) that at all but the largest competitions you can have a single scrambler and they won't even be very busy. In fact, as someone who often helps out at the scrambling table, I'd rather apply a lot of extra 3x3 scrambles than wait around for the last solves of the inevitable slow solver who didn't start until the round was pretty much already over.

Right, I definitely don't like the formats you mention. I also would argue they're not legal by WCA regulations, as you're not letting people continue based on time (you had a time of DNF) or ranking.

As for (a), my issue is that people can choose to voluntarily take a DNF, which is a better option for them than attempting a solve. I don't like that a strategy that is in all cases inferior to another (in terms of results) should be preferred. This is different from taking risks--when you take a risk, there's a chance of failing and ending up with a worse result than the "safe" option, so the safe strategy is not in all cases inferior.

As for (b), I suppose, but I don't like the idea of someone not attempting a solve and then expecting the scramblers to do another scramble. Maybe it's just the morality of it, I don't know.

 

[KJiptner]

I would also agree on having more attempts! 2 is just tooo stressful and frustrating (see recent events). 3 is okay I think. What about the regular best of 2 and you get 3 attempts extra if you meet the timelimit (just like we do for many other events).

 

[Mike Hughey]

What about the regular best of 2 and you get 3 attempts extra if you meet the timelimit (just like we do for many other events).

That's okay, but unfortunately it requires that you get a successful solve in the first 2 attempts. It would definitely be unjustified to allow the extra attempts if you get DNFs for the first two attempts, regardless of how fast they go. So then I don't see why this would be any less stressful or frustrating. I know I was scared silly in Toronto when we did a combined final - we had to get a success on the first solve or we wouldn't get the other two. It was the most stressed I ever felt on a BLD solve. The funny thing is that the first solve was my fastest of the three - I think I was so relieved (and therefore relaxed) after getting that first solve that I wasn't able to focus enough to get a good solve on the other two.

I think I like the idea of giving a specified amount of time for all your solves. Ten minutes would be best for big competitions, like US Nationals, but for local competitions you could give more time like 20 or even 30 minutes if you as an organizer want to. That would allow some beginners to have a better chance, and still give good people the opportunity for up to 5 solves.

Tim's objections make sense, and since I hate DNFs so much, I empathize with them a bit, but I think there are a lot of advantages to this too, and for me, they outweigh the disadvantages. I seriously doubt I would ever take a DNF intentionally, though.

The only rules change required to implement qqwref's suggestion is to allow best of 5 format as well as best of 3 format for 3x3x3 BLD. Having a total time requirement is already in place and is often used for big cubes BLD. So it really wouldn't be much of a change. But we would probably want to add details indicating that times should be recorded with DNFs, and rules for calculating whether or not another attempt is allowed.

 

[amostay2004]

So what if people intentionally take DNFs to get more attempts.

The dilemma here is that best of 5 format is good, however it takes up too much time. By setting a maximum time for all attempts, organizers can be rest assured that the event will end within a reasonable period of time.

It is also easier to stick with the schedule and plan the time for BLD rounds more accurately. I believe this arrangement increases the efficiency of BLD rounds - it doesn't really affect the slow people, and fast people are able to get more tries.

 

[Sakarie]

Yeah, I think too that a "maximum 10 minutes or 5 attempts, but minimum 2 attempts" would be a good rule/alternative. I think there should be a minimum in the comps with the possibility.

I don't think there will be a problem with people intentionally DNF:ing. And those aren't so fast that they're the most competitive persons. And also, you could put a small clausule to avoid it.

 

[Tim Reynolds]

And those aren't so fast that they're the most competitive persons.

Anyone who averages ~2 minutes has incentive to DNF. And they're in contention to do reasonably well at a competition. Also, why that matter?

And also, you could put a small clausule to avoid it.

I don't see how.