3x3x3 Blind, changing to 3/5?

[AvGalen]

For almost all events (if time allows it) the preferred format is now 3/5 (remove best and worst, average of the remaining 3 solves)

In 2009 the rules for multi-blind were changed to allow for DNF's (although the basic idea is still to stimulate perfection over speed)

Regular blindfolded solving is currently best 1 out of 3. This has resulted in competitors specialising in methods that can give great results, but also allows for many DNF's. The basic idea for most competitors seems to be "go faster then you are really capable of and accept the DNF's if they occur. A "safety" solve after 2 DNF's can always be tried to get a ranking.

I am in no way a good blindfolder, but I think it is not in the interest of the audience (and very annoying for judges) to see so many DNF's happen. Especially after you see (other) competitors do successful multi-blinds of > 5 cubes.

All of this leads me to the question: Should blind become a "3/5" event as well or is the goal for blind really "do 1 succesful and fast, plork the rest"?

Advantages of 3/5:
- Many more successful solves
- More attempts (for the more successful solvers)
- Mostly compatible with the current system because of "single" solves
- Removing lucky solves from the average to better represent skill while still allowing for a DNF

Disadvantage of 3/5:
- Might force competitors to practice and compete differently
- Not entirely compatible with current system
- Might require more time in competitions if competitors are allowed to do solves after DNF-ing twice
- More safety solves, leading to more time used and less spectacular results

Just throwing it out here to hear your opinions about this

 

[Mike Hughey]

Most people probably already know I agree with Arnaud on this one.

But I suspect it really doesn't matter how we go in the near future - I suspect we'll get to the point someday where this just makes sense to do, although it may take quite a few years yet.

The fact is that cubing is in its infancy, and BLD cubing even more so. When the best BLD solvers have practiced regularly for many years, there really will be no reason to expect them to miss very many of their BLD solves. Practice will make perfect. I know that I can squeeze out 5 or 10 extra seconds by really rushing, with my accuracy going down correspondingly, but that time will decrease as I get better, and should be even lower for the really fast people. No one will want to waste a good attempt anyway, so all the good people will eventually really want to get every solve correct.

So I believe that our accuracy will eventually go up to where this is not a particularly big change.

And I agree with Arnaud's opinions about spectators. It's very disappointing for spectators to see so many DNFs. If we imagine this on TV someday, we would want a format where at least most of the solves are successful, and this format would encourage that. I think it would be better for spectators.

And while I really admire Haiyan (well, his blindfold solving, at least) and Ville (period), I must admit my BLD hero is Clément Gallet, which is why I keep bringing him up every time this issue comes up. 31 successful solves in a row - simply amazing!

 

[Kidstardust]

That are your thoughts while traveling through Asia????? :confused:

 

[Slash]

All my reasons against it are because I suck at bld accuracy. I think wit hthis format, we'll be able to see who really is a good blindsolver. For example, maybe Mike could win a competition (which is on the one hand because of the format)

And yes, the blind event would get harder. But that would "filter out" those competitors who just go and try a blindfolded solve "just coz I solved it once at home"

 

[Mike Hughey]

All my reasons against it are because I suck at bld accuracy. I think wit hthis format, we'll be able to see who really is a good blindsolver. For example, maybe Mike could win a competition (which is on the one hand because of the format)

And yes, the blind event would get harder. But that would "filter out" those competitors who just go and try a blindfolded solve "just coz I solved it once at home"

Please note that competitors could still do this - results of a 3/5 average are ranked such that the top people are those who get a full average, and then ranked according to single solves. So if you just want a single success, you can still go for one, and still have a ranking in the results. In fact, for a while, I suspect you wouldn't need a successful average at most competitions to make the top 3, and we'd probably often have a winner who didn't get a successful average. And there's nothing wrong with that. But if we made this change, the good people would start working harder to improve their accuracy, so eventually we'd probably almost always have a full podium of averages.

I'm in no hurry to make this change, by the way; I'd like to see it happen in 5 years or so. Give the event some more time to mature first. But if we did it earlier, it would give me an advantage.

 

[FatBoyXPC]

Well then I vote we do it earlier! I think BLD is one of the more fun events to watch. I think stiffening up the event a bit would be a good thing, and it would especially make for more another set of WR for BLD. I think the mere fact that if HaiYan could use the 15s inspection time we got do it sighted, he'd set a much lower WR, which is so crazy to think about.

My question is: how long until we end up doing this for 4bld, 5bld, etc?

 

[AvGalen]

That are your thoughts while traveling through Asia????? :confused:

For the first time in a long time I did the weekly competition and got all my solves for 222_blind and 333_blind correct. I also saw Haiyan compete and get 3 DNF's. Then http://www.speedsolving.com/forum/showthread.php?p=453479#post453479 happened and I thought it was time to post this. Anything wrong with my thoughts?

All my reasons against it are because I suck at bld accuracy. I think wit hthis format, we'll be able to see who really is a good blindsolver. For example, maybe Mike could win a competition (which is on the one hand because of the format)

And yes, the blind event would get harder. But that would "filter out" those competitors who just go and try a blindfolded solve "just coz I solved it once at home"

Please note that competitors could still do this - results of a 3/5 average are ranked such that the top people are those who get a full average, and then ranked according to single solves. So if you just want a single success, you can still go for one, and still have a ranking in the results. In fact, for a while, I suspect you wouldn't need a successful average at most competitions to make the top 3, and we'd probably often have a winner who didn't get a successful average. And there's nothing wrong with that. But if we made this change, the good people would start working harder to improve their accuracy, so eventually we'd probably almost always have a full podium of averages.

I'm in no hurry to make this change, by the way; I'd like to see it happen in 5 years or so. Give the event some more time to mature first. But if we did it earlier, it would give me an advantage.

In the past it has happened that I was DNF-fing my 2nd blindfolded attempt (quite slowly, as always) and everyone else had already gotten 3 DNF's. When I went up for my last attempt I knew I only had to do a succesful solve to win, time wasn't important.
And yes, Clément Gallet is my blindfolded hero as well, proving that "slow" but steady can be a perfectly fine tactic for blind.

I must say that I am surprised that fast blindfolders aren't shooting this idea down en masse :tu

Well then I vote we do it earlier! I think BLD is one of the more fun events to watch. I think stiffening up the event a bit would be a good thing, and it would especially make for more another set of WR for BLD. I think the mere fact that if HaiYan could use the 15s inspection time we got do it sighted, he'd set a much lower WR, which is so crazy to think about.

Just to be clear: I am not suggesting to use 15 seconds inspection time for blind!
So far, only 60 people have ever succeeded at 444bld (15 below 10 minutes) and for 555 there are only 22 (nobody below 10 minutes). Never say never, but there just won't be enough time at a competition to do it, or people capable of doing it

 

[cubedude7]

Awesome idea! You have my vote!!

 

[amostay2004]

I don't see why fast blindfolders (or any blindfolders) would disagree with this. After all, they can just not go for accuracy and they still get 2 more attempts than before. It's a win-win situation for people who aim for speed, and those who go for accuracy.

Of course, the introduction of a format like this would definitely encourage fast BLD-ers to get a good average in comp as well So I definitely agree with this.

 

[Mike Hughey]

So far, only 60 people have ever succeeded at 444bld (15 below 10 minutes) and for 555 there are only 22 (nobody below 10 minutes). Never say never, but there just won't be enough time at a competition to do it, or people capable of doing it

Yes, never say never. It might be a while, but we really might get there someday.

As for an example of someone who already can, look at Gabriel Alejandro Orozco Casillas. Get him to a couple more competitions, and he'll be my new hero.

4x4 Cube: Blindfolded
Guadalajara Open 2010 Final 1 5:54.56 NR 5:54.56 6:33.44 6:00.16

 

[AvGalen]

So far, only 60 people have ever succeeded at 444bld (15 below 10 minutes) and for 555 there are only 22 (nobody below 10 minutes). Never say never, but there just won't be enough time at a competition to do it, or people capable of doing it

Yes, never say never. It might be a while, but we really might get there someday.

As for an example of someone who already can, look at Gabriel Alejandro Orozco Casillas. Get him to a couple more competitions, and he'll be my new hero.

4x4 Cube: Blindfolded
Guadalajara Open 2010 Final 1 5:54.56 NR 5:54.56 6:33.44 6:00.16

Good example Mike. I must admit that I hadn't heard his name before.
But I think you proved that doing several continuous bigcubes blind without DNF'ing "years ago" in the weekly. Credit where credit is due

 

[jms_gears1]

So far, only 60 people have ever succeeded at 444bld (15 below 10 minutes) and for 555 there are only 22 (nobody below 10 minutes). Never say never, but there just won't be enough time at a competition to do it, or people capable of doing it

Yes, never say never. It might be a while, but we really might get there someday.

As for an example of someone who already can, look at Gabriel Alejandro Orozco Casillas. Get him to a couple more competitions, and he'll be my new hero.

4x4 Cube: Blindfolded
Guadalajara Open 2010 Final 1 5:54.56 NR 5:54.56 6:33.44 6:00.16

Good example Mike. I must admit that I hadn't heard his name before.
But I think you proved that doing several continuous bigcubes blind without DNF'ing "years ago" in the weekly. Credit where credit is due

The Famous Blindfold rubiks

 

[FatBoyXPC]

AvG: I wasn't saying you were proposing inspection, I was saying that HaiYan can memo in almost the 15 seconds we get for inspection, just imagine what that would do to WRs (especially if he was still able to compete in the WCA right now). I do agree though with the 3/5 format

 

[qqwref]

I don't like this idea, and here's why.

I've been to a lot of small competitions in my time. I'm sure for all your BLD masters there are people at every competition (that is, YOU) who are not only fast but also can get at least 4 out of every 5 solves correct. But from my point of view this is not the case. I've gone to a lot of more local competitions where the only people who are likely to do that well are the very slow people who take their time. You'll get 7 minute people winning small competitions, I guarantee it. You will also get fast people getting very angry when they realize they've made a mistake on another solve. The reason is that the fast people often go quickly to get a good time, whereas the slow people tend to care mainly about consistency because of how long it already takes to do an attempt.

The thing about BLD is that getting a DNF is very easy. It's like Clock, but even worse. Make a small mistake and don't notice it, and your solve is forfeit. To me, asking for a maximum of one DNF in 3BLD is just as strict as asking for 3 of 3 non-DNFs in clock: even for the best people you cannot go fast without some chance of a DNF, and that means there is always some probability that you automatically lose. Unlike speed events, DNFs are final: you can rarely just fix it before the timer stops. This proposal just adds more stress to every BLD round. You have to remember, also, that competitions are not optimal environments to do something which requires great and constant concentration. You might have people running around, other cubers solving stuff in the background, people talking, and so on. This never happens in the big/multi BLD room, but 3BLD is usually on the main stage. So making BLD a 3/5 event takes control completely out of the hands of all but the most skilled or slow-and-steady competitors. Instead of being a contest of speed, it becomes a game of luck, where a bad roll will completely disqualify you no matter how much practice you have done or how skilled you are.

tl;dr: DNFs are luck, and asking for so few of them means that even very fast people will get randomly DQ'd. The event will be more stressful and the winners will occasionally be extremely slow.

 

[Escher]

Instead of being a contest of speed, it becomes a game of luck, where a bad roll will completely disqualify you no matter how much practice you have done or how skilled you are.

Agreed ~ in my mind speedcubing has enough problems with variance in and of itself, let alone introducing new formats that are punishing for it in an event that is one of the worst for it.

 

[Slash]

I don't like this idea, and here's why.

I've been to a lot of small competitions in my time. I'm sure for all your BLD masters there are people at every competition (that is, YOU) who are not only fast but also can get at least 4 out of every 5 solves correct. But from my point of view this is not the case. I've gone to a lot of more local competitions where the only people who are likely to do that well are the very slow people who take their time. You'll get 7 minute people winning small competitions, I guarantee it. You will also get fast people getting very angry when they realize they've made a mistake on another solve. The reason is that the fast people often go quickly to get a good time, whereas the slow people tend to care mainly about consistency because of how long it already takes to do an attempt.
...

You're right in some way, but those slower people at smaller competitions would also win with their singles. so, the ranks would look like this:
first: best average with best single
second: worse average with (maybe) worse single
third: DNF average with good enough single
...

so, at a small competition where there are no that good blindsolvers, the DNF average with good single can easily win. (It's like at Czech Open 2010 Charlie was the third in clock with DNF avg.)

 

[Mike Hughey]

You'll get 7 minute people winning small competitions, I guarantee it.

Are there really many 7 minute people who are also that accurate? And besides, you would most likely run something like this as a combined final - one of your first two solves in under 2 or 3 minutes, for instance. So a 7 minute solver would never have the chance to get an average.

 

[AvGalen]

Instead of being a contest of speed, it becomes a game of luck, where a bad roll will completely disqualify you no matter how much practice you have done or how skilled you are.

Agreed ~ in my mind speedcubing has enough problems with variance in and of itself, let alone introducing new formats that are punishing for it in an event that is one of the worst for it.

...and I have no problem with disqualifying people that don't solve their cubes blindfolded during a blindfolded event. It doesn't matter how much you have trained or how skilled you are, if you don't solve the puzzle you don't deserve to win.

I also think that you are wrong in assuming that fast people would DNF more often than slow people. I have judged many fast people that solved cubes blindfolded in 55, 60 and 65 seconds while a slow persons just DNF'ed 3 times after 9 minutes.

Mostly, I have listed all your points already in my startpost under advantages/disadvantages. The only point you added was that circumstances for 3x3x3 blind can be bad during competitions. For that I have to say: The organisers and delegates should take care of those problems, "earplugs" are allowed and Tim did 24/24 inside the main hall (not in a special quiet blind room)

 

[Tim Reynolds]

tl;dr: DNFs are luck, and asking for so few of them means that even very fast people will get randomly DQ'd.

I disagree.

The event will be more stressful and the winners will occasionally be extremely slow.

I agree.

My main objection is that it would take considerably longer to do a blindfold round at a normal competition. Not only would two more solves need to be done, each solve would take longer as people would be more careful. 3-4 minute solvers, who usually aren't holding up the competition too much, might take 5-6 minutes to make sure they get each solve. Sure, for competitions like worlds at a time when people have gotten used to this format, people would still be able to go quickly, but at small competitions it makes BLD a much harder event to run.

Also I don't think we need to worry about the media getting upset about too many DNFs--if they see people in the finals at worlds DNFing, they'll think it's very difficult, not that the competitors suck at it.

As for combined finals, if you cut people off after 2 solves who are over (say) 5 minutes, then people with 2 DNFs have to be cut off as well. I still don't know how to effectively run a combined round for BLD that follows the regulations.

 

[FatBoyXPC]

Not to mention, cubers can train themselves in loud areas. In the "Frustrated while cubing" thread, I mentioned training while under different types of pressure. Chris Hardwick says he likes to train where he can get distracted and tells his family to try to distract him while he cubes. He does that to get better. My point is that if you train yourself to only work in a quiet environment, that is your fault as a cuber.

A good comparison can be made to the sports teams that aren't used to playing in cold conditions, real grass vs astroturf, etc. It's much harder to train in snow in Florida (american football) than it is to train in a very loud distracting area (no matter where you live).