[WCA Regulations 2012] Proposed: Alternative formats for 3x3x3 BLD

[keyan]

RE Dene and others suggesting an expanded format with five solves with 'combined round' formats:

I'm a BLD solver averaging five minutes. Format is best of five, with a successful solve under three minutes in the first two to complete the remaining three. I can try to blitz the first two, trying for a chance at the remaining three, or just go for regular solves on the two. Old format I get three solves, now I either get two likely DNFs with a chance of five solves, or only two solves.

I'm a BLD solver averaging five minutes. Format is best of five, with a solve (success or DNF) under three minutes in the first two to complete the remaining three. My first solve, I just pretend to do a full solve, for example only memorizing corners. Oh, a DNF, but look, I had part of it, and look only 2:30. Now I get four more solves, at about five minutes each. Old format I get three solves, fifteen minutes. New format I get four solves, twenty minutes, plus the added time of me faking a solve.

I'm a BLD solver averaging five minutes. Format is best of five, with a total allowed solve time (success and DNF) of fifteen minutes. Before I start competing, someone gets four 1:00 solves, then on their fifth solve gets a 0:20 world record. My turn to solve. First three solves I do a lazy fake attempt, two minutes each. Fourth solve is a safety solve, seven minutes, I at least get a ranking. Final solve, I go all out knowing a world record has already been set on this scramble, three minute solve. Old format, three solves fifteen minutes. New format, two solves ten minutes, plus three fake solves and the chance to game the system and choose which scramble I want to compete with.

 

[qqwref]

I'd like to see an "average of 5" competition, with the single ranking taking precedence. So the winner is the person with the best single time, not the best average. It's equivalent to "best of 5" except that averages are also recorded, and average records can be set. This is better than "best of 5" because of the possibility of average records; it's also better than a pure "average of 5" because consistency is not required to do well, and a few DNFs won't destroy your chances at winning.

EDIT: Oh, and also, knocking someone out after one or two solves with this format wouldn't destroy their chances of getting ranked.

 

[aronpm]

RE Dene and others suggesting an expanded format with five solves with 'combined round' formats:

I'm a BLD solver averaging five minutes. Format is best of five, with a successful solve under three minutes in the first two to complete the remaining three. I can try to blitz the first two, trying for a chance at the remaining three, or just go for regular solves on the two. Old format I get three solves, now I either get two likely DNFs with a chance of five solves, or only two solves.

I'm a BLD solver averaging five minutes. Format is best of three, with a successful solve under three minutes in the first to complete the remaining three. I can try to blitz the first, trying for a chance at the remaining two, or just go for a regular solve on the first. I either get a likely DNF with a chance of three solves, or only 1 solve.

I don't see a difference.

I'm a BLD solver averaging five minutes. Format is best of five, with a solve (success or DNF) under three minutes in the first two to complete the remaining three. My first solve, I just pretend to do a full solve, for example only memorizing corners. Oh, a DNF, but look, I had part of it, and look only 2:30. Now I get four more solves, at about five minutes each. Old format I get three solves, fifteen minutes. New format I get four solves, twenty minutes, plus the added time of me faking a solve.

This isn't a problem, you just allow an x minute total solve time

I'm a BLD solver averaging five minutes. Format is best of five, with a total allowed solve time (success and DNF) of fifteen minutes. Before I start competing, someone gets four 1:00 solves, then on their fifth solve gets a 0:20 world record. My turn to solve. First three solves I do a lazy fake attempt, two minutes each. Fourth solve is a safety solve, seven minutes, I at least get a ranking. Final solve, I go all out knowing a world record has already been set on this scramble, three minute solve. Old format, three solves fifteen minutes. New format, two solves ten minutes, plus three fake solves and the chance to game the system and choose which scramble I want to compete with.

I'm a BLD solver averaging five minutes. Format is best of three, with a solve (success and DNF) of twelve minutes. Before I start competing, someone gets two 1:00 solves, then on their third solve gets a 0:20 world record. My turn to solve. First solve I do a lazy fake attempt, two minutes. Fourth solve is a safety solve, six minutes, at least I get a ranking. Final solve, I go all out knowing a world record has already been set on this scramble, three minute solve. Two solves nine minutes, plus a fake solve and the chance to game the system and choose which scramble I want to compete with.

None of your points are valid or don't already apply to best of 3.

The event takes more 5 or 10 more minutes? Keep the total time limit the same if it's such a big deal.

And for competitors who average 5 minutes and want 5 solves? Try practising more.

 

[keyan]

None of your points are valid or don't already apply to best of 3.

I don't support those formats in the current best of three, either. I was specifically responding to people who said that changing to a best of five format wouldn't be a problem because the time can be limited using various methods. So far, none of those methods work.

 

[aronpm]

I disagree, having a total time limit has worked at every competition I've attended that has used it.

 

[keyan]

I disagree, having a total time limit has worked at every competition I've attended that has used it.

...all four?

I think my third example covers this. The 'honest' five minute competitor gets three solves, and doesn't get the potentially easy scramble. Someone out to game the system gets to skip whichever of the first four solves they want, and specifically choose the final 'easy' scramble.
Even better, like this: A 'good' BLD cuber (averages 1:00) goes up, gets 2:00 on the first three solves, then 1:00 and 0:20. The 'dishonest' competitor then knows he wants to skip the first three scrambles and focus on the fifth. Whereas the 'honest' competitor only solves on the 'harder' scrambles.

 

[Dene]

I'm with aronpm. I don't see how this suddenly becomes an issue now that we allow 2 more solves. Just have a strict cutoff time so that only those competitors which are truly fast (or reasonably fast) are able to complete all 5 solves. Using a format where a DNF under a certain time limit meets the cutoff is silly. A DNF should not meet a cutoff time, as the cube was never solved.

Now that I've thought about it more I am actually on the side of allowing for averages and not just best-of. Those that only care about single can go hard for that and also get more attempts (5 instead of 3), and those that care about consistency can go for the best average. If you are fast and consistent, all the better for you.

 

[aronpm]

...all four?

Do you have a problem with that? They've all been busy competitions with 2 rounds of 3BLD.

I think my third example covers this. The 'honest' five minute competitor gets three solves, and doesn't get the potentially easy scramble. Someone out to game the system gets to skip whichever of the first four solves they want, and specifically choose the final 'easy' scramble.
Even better, like this: A 'good' BLD cuber (averages 1:00) goes up, gets 2:00 on the first three solves, then 1:00 and 0:20. The 'dishonest' competitor then knows he wants to skip the first three scrambles and focus on the fifth. Whereas the 'honest' competitor only solves on the 'harder' scrambles.

Again, this already applies to best of 3. If best of 3 works, then best of 5 works. I also don't see why that situation is a problem.

But try proposing an alternative for best of 3/5 instead of saying it just doesn't work.

 

[Dene]

...all four?

I think my third example covers this. The 'honest' five minute competitor gets three solves, and doesn't get the potentially easy scramble. Someone out to game the system gets to skip whichever of the first four solves they want, and specifically choose the final 'easy' scramble.
Even better, like this: A 'good' BLD cuber (averages 1:00) goes up, gets 2:00 on the first three solves, then 1:00 and 0:20. The 'dishonest' competitor then knows he wants to skip the first three scrambles and focus on the fifth. Whereas the 'honest' competitor only solves on the 'harder' scrambles.

This same issue applies to any event. For example, I sit back and notice that several people have gotten good times on a particular 3x3 solve, and because I know most people use white cross I ensure that when I come to that solve I check out the white cross and look hard for what was so easy that everyone else noticed.

 

[keyan]

Do you have a problem with that? They've all been busy competitions with 2 rounds of 3BLD.

'scuse me, that was a bit rude. Simply saying that just because you think it worked in your experience, doesn't mean it works always.

To both of you, I'm not saying that this "suddenly" becomes an issue. It's a problem that exists if it's used with best of three, absolutely. And as such I don't think this competition format should be used in current competitions either. I'm addressing those people that are saying a change to five solves instead of three won't be a problem because time can be limited with these, broken, methods. Competition time is limited, changing to five solves will take more time, don't use broken examples to try to offset that time.

And as for Dene's example, the difference is there you're still solving all the scrambles, whereas between the two BLD solvers, they didn't.

 

[Dene]

And as for Dene's example, the difference is there you're still solving all the scrambles, whereas between the two BLD solvers, they didn't.

So? The only way to prevent this is to force them to solve the puzzle (i.e. not DNF) within a strict time limit (e.g. 1 minute). Then they have to really solve it, and really solve it fast to get a chance at the scramble they know is very good.

Also, I believe the proposition is to have up to best of 5. A competition organiser could still choose best of 1 if they wanted.

 

[keyan]

So? The only way to prevent this is to force them to solve the puzzle (i.e. not DNF) within a strict time limit (e.g. 1 minute). Then they have to really solve it, and really solve it fast to get a chance at the scramble they know is very good.

Also, I believe the proposition is to have up to best of 5. A competition organiser could still choose best of 1 if they wanted.

Or, much easier, you prevent it by not having a limit on total solve time. That way everyone solves all the scrambles without having to fake solves or whatever.

Note I never said I oppose five solves. I'm just saying that the people supporting five solves and saying it won't cause a time problem are using bad arguments. People saying five solves isn't a problem because you can use a broken system to limit the time it takes aren't actually helping their position. Better to mention benefits, more representative results with average three of five or median of five or whatever.

 

[TimMc]

As far as averages for BLD go;

BLD and speedsolving are seperate things. You're attempting to treat BLD like speedsolving when it is entirely different.

Given how fast 333bf is now would it be fair to regard it as speedsolving? However, I can understand your point when applied to 555bf or mbf...

I don't really see the benefits of increasing 'x' to 5 for BLD events especially. Why not just have more rounds for more attempts?

Introducing an 'average' would bring a familiar measure into things and separate the elite from the lucky/inconsistent solvers.
+ it would just default back to what's essentially 'Best of 5' if everyone DNF's more than once.

Tim.

 

[Kirjava]

Given how fast 333bf is now would it be fair to regard it as speedsolving? However, I can understand your point when applied to 555bf or mbf...

No. It is a very different event that is much easier to DNF in.

I don't really see the benefits of increasing 'x' to 5 for BLD events especially. Why not just have more rounds for more attempts?

A lot of the time there are not enough competitors for this.

Introducing an 'average' would bring a familiar measure into things and separate the elite from the lucky/inconsistent solvers.

Consistent != elite.

 

[Dene]

Consistent != elite.

Of course consistent = elite; the elite of being consistent. That is what a bld average would be all about.

 

[Kirjava]

Of course consistent = elite; the elite of being consistent. That is what a bld average would be all about.

I would rather test the elite of BLD, not the elite of consistency.

 

[aronpm]

Of course consistent = elite; the elite of being consistent. That is what a bld average would be all about.

The goal of speedcubing/blindsolving is not consistancy.

 

[Dene]

Neither is the goal of speedcubing picking up and putting down a magic as fast as possible, but we still do it.

But why shouldn't consistency be a goal of speedcubing? It would make for an interesting and new challenge for people. Also, without consistency it would be pretty difficult to beat the current WR average from any event.

 

[TimMc]

An Average of 5 format would simply default back to what's essentially Best of 5 when everyone gets a DNF. Anyone who's accurate and fast enough would take the average.

Tim.

 

[Kirjava]

But why shouldn't consistency be a goal of speedcubing?

It should, but I don't see blindsolving as that much like speedsolving. The very nature of the event causes it to produce very different types of results to other nonBLD events time wise.