• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Proposal: Combined Best of 3/Best of 5 Rounds for 3BLD

What format do you prefer for 3BLD round?

  • Best of 3 (keep it as it is)

    Votes: 18 50.0%
  • Combined Best of 3/Best of 5

    Votes: 18 50.0%
  • Other (please specify)

    Votes: 0 0.0%

  • Total voters
    36

porkynator

Member
Joined
Oct 27, 2010
Messages
1,322
Location
Belluno, Italy
WCA
2011TRON02
YouTube
Visit Channel
I have original posted this on github a couple of weeks ago. Please read the whole Pros/Cons list before voting, especially if you would like to pick "Other".

I would like to propose the following new format for 3x3 Blindfolded rounds in WCA competitions: combined Best of 3/Best of 5.
  • After the first 3 solves, competitors meeting a specified cutoff time can procede and do two more attempts.
  • In competitions, the final ranking is still based on best single.
  • "Average of 5" would replace "Mean of 3" in the average rankings for 3BLD, as has happened recently for 3x3 With Feet and for other events in the past. For 3BLD, this has no effect during the competition itself.
  • The ideal cutoff should be somewhat low, like 1:30 or 1:00, to prevent the round from lasting too long.
Reasons
At high level, 3BLD times are comparable to 4x4 and 3x3 with feet. These events are already held in "Average of 5" format (the latter having been changed lately), and even longer events (5x5 and Megaminx) get 5 solves per round. Thus, it seems reasonable to allow 5 solves for 3BLD as well. This means more fun and more chances of setting new records.

Pros
  • More attempts = more fun!
  • Bigger chances of setting new records
    There is quite a big gap between the the best known unofficial results and the official world records. This is true for any event and I think the main reason for this is that people can do many more solves at home than in competition. This is in particular true for 3BLD, according to the speedsolving UWR page: in the original post on github you can see a table showing this gap for 3x3 Cube, 3x3 Blindfolded and events with solving time comparable to the latter (4x4 Cube and 3x3 With Feet). I think this new format can help reducing this gap.
  • It's easier to get a succesful Average of 5 than a succesful Mean of 3
    If one's success rate is better than 25%, see attachment. This is a matter of basic probability computations, assuming the success rate is the same for all attempts. This doesn't take into account the fact that one might get tired after some solves, or that one can influence their own success rate at will.
  • The competition strategy doesn't change
    In a Best of 3 round one has to get at least one good solve out of 3, so the "minimum success rate" is 1/3. In a Bo5 round this is 1/5, but using the combined Bo3/Bo5 format we can allow for more solves, while keeping this minimum success rate at 1/3. The only thing that changes is how to get an average, but this doesn't matter for the competition, and in any case it would become easier (see previous point).
Cons
  • Slower cubers might not get an average
    This is true for any "combined" round. Since the ranking would still be based on single, this is less revelant with the proposed format. There may be a bigger disproportion between beginners' and world-class times than what we have in other events, but it doesn't take an insane amount of practice to get down to, say, 1:30 average.
  • It is more tiring to do 5 solves
    I believe for anyone faster than, say, 1 minute, doing 5 solves in a row is not a problem. This is consistent with the cutoff time suggested above. Moreover, if one finds it tiring to complete the last two solves, they can always DNS them and have no big disadvantage (the format is still "Best of X").
  • It will slow competitions down
    That's what cutoffs and time limits are for. If they are low enough, this will be no problem. And in many cases faster competitors can complete 5 solves faster than slower competitors doing 3 solves, so the extra solves for them would not be the bottleneck. It is also possible to have a cutoff time hardwired in the regulations, but I would rather specify it in the guidelines or in some other way. Moreover, it is still possbile to use cumulative time limits to ensure the round doesn't last too long.
  • With low cutoffs, very few competitors may be able to complete the round
    This is true, but as I mentioned before it is not as bad as in the case of Bo2/Avg5 or Bo1/Mo3 combined rounds: competitors are ranked by single in any case. Missing the average doesn't change anything from the point of view of the competition. Even if nobody meets the cutoff it shouldn't be a problem.
  • Inconsistency: there is no other event with such a format
    Personally, I don't think this is a good reason to keep things like this. The point of the regulations is to make competitors have fun and compete under fair conditions, not to give a satisfactory sense of order when read.
Conclusions
I think this new format has the advantage of giving more opportunities for records and more fun to faster solves, while not giving any true disadvantage and not changing anything about the event competition-wise.
 

TDM

Member
Joined
Mar 7, 2013
Messages
7,006
Location
Oxfordshire, UK
WCA
2013MEND03
YouTube
Visit Channel
I've been thinking Ao5 would be a better format for 3BLD for a while, and I'd be happy if this was changed.

However, there is one thing which I would do differently to what you've proposed. For all other Ao5 events, the default is Bo2/Ao5 for combined rounds, not Bo3/Ao5. I know that DNFing is more likely than other Ao5 events, but I can't see why having that extra solve before the cutoff matters. After all, if you only get one success in the first three solves, the average is already a DNF. Using the standard Bo2/Ao5 format would not only make it more consistent with other events (and less confusing to newcomers), but it could also speed up comps in comparison to the proposed Bo3/Ao5. As you mentioned, the slower cubers are usually the last to finish.

E: Out of interest, I graphed the probability of an Ao5 success (blue) vs a Mo3 success (red) against a cuber's success probability (x-axis), assuming each attempt is independent
6UMHjuY.png

https://www.desmos.com/calculator/xxqgrl1ddr
 
Last edited:

xyzzy

Member
Joined
Dec 24, 2015
Messages
2,876
However, there is one thing which I would do differently to what you've proposed. For all other Ao5 events, the default is Bo2/Ao5 for combined rounds, not Bo3/Ao5.
Just a quick note: the proposal is Bo3/Bo5, not Bo3/Ao5. The two extra solves are there to help fast people get better singles (especially since BLD is primarily ranked by singles).
 

mark49152

Premium Member
Joined
Oct 29, 2012
Messages
4,719
Location
UK
WCA
2015RIVE05
YouTube
Visit Channel
The main thing I wouldn't like about this is that the best solve would be excluded from ao5, which seems inconsistent for an event ranked on single. It's like saying your best solve is legitimate to win the comp but "too lucky" to count towards your average.

On the bo2 versus bo3 point, bo2 for sighted events is a reasonable measure of whether the competitor can reach a certain level, and they will solve as fast as they can to reach that target. For BLD it's different, and I think it would degrade comps to have people sitting down to do a safety on their first solve.

Like everyone, I'd like to have more attempts in comp, but personally I'd prefer to have that via a second round of bo3 with 75% going through.
 

TDM

Member
Joined
Mar 7, 2013
Messages
7,006
Location
Oxfordshire, UK
WCA
2013MEND03
YouTube
Visit Channel
Just a quick note: the proposal is Bo3/Bo5, not Bo3/Ao5. The two extra solves are there to help fast people get better singles (especially since BLD is primarily ranked by singles).
You're right, sorry. I read him talking about Ao5 a couple of times, and assumed that was what the round format was as well.

In that case, I'm not so sure about this. It just seems like giving extra attempts to faster people, to make the records faster. Having an average result which is easier to achieve isn't as important if it's not being used in competitions. Since averages are easier to get, and if they're going to be ranked anyway, why not have them as the format in comps?
 

porkynator

Member
Joined
Oct 27, 2010
Messages
1,322
Location
Belluno, Italy
WCA
2011TRON02
YouTube
Visit Channel
Since averages are easier to get, and if they're going to be ranked anyway, why not have them as the format in comps?

Ao5 is easier to get than Mo3, but still pretty hard. For example, with 75% accuracy you have around 63% of chance of getting an Ao5. It's too easy to mess up even for accurate people. For this reason I don't think it's feasible to use Ao5 to determine winners.
 
Joined
Apr 23, 2010
Messages
1,391
Location
Scotland, UK
WCA
2009SHEE01
YouTube
Visit Channel
The main thing I wouldn't like about this is that the best solve would be excluded from ao5, which seems inconsistent for an event ranked on single. It's like saying your best solve is legitimate to win the comp but "too lucky" to count towards your average.
I was thinking this might potentially work, until you pointed this out, I think it's a pretty big catch. I agree with other concerns pointed out by other people too, so voting to keep current system.
 

porkynator

Member
Joined
Oct 27, 2010
Messages
1,322
Location
Belluno, Italy
WCA
2011TRON02
YouTube
Visit Channel
My general opinion about the problems/inconsistencies with Ao5 replacing Mo3 is that it shouldn't be too important after all. The "real" records for the event are the single records (because you win competitions with single results), while the average is just recorded as an extra achievement. So it is not a big deal.

This fits in the point of view that "competitions are more important than records", for example World Champion title > World Record. Not everybody agrees with this, but I think this is the direction we are going (for example, history of world/continental/national championship podiums was added to WCA profiles recently).

anyway, thanks everyone for the feedback!
 

Loser

Member
Joined
Sep 4, 2017
Messages
306
Location
Massachusetts, USA
WCA
2016QUIN01
One of the major arguments for changing the feet event to average of 5 was that the record feet times were below that of megaminx and 5x5. As the record times for blind are also lower than 5x5 and megaminx, and compareable to feet. Obviously havjng blind be an average of 5 doesn't make sense, but does changing it to best of 4 or 5 make sense? Or taking the second best of 5?
I'd never heard of this topic being brought up, and I decided to post it here to hear people's thoughts.

Edit with personal thoughts: while I'm definitely fine with bo3, I would be interested in bo4 or bo5
 

One Wheel

Member
Joined
Feb 24, 2016
Messages
2,883
Location
Wisconsin
WCA
2016BAIR04
Porky maybe a post on this I believe. I'll edit this comment when I find it.

EDIT: Here it is. https://github.com/thewca/wca-regulations/issues/571

After reading through that proposal I find it highly compelling. I have a very long way to go before I could ever hope to come close to making anything like a 1:30 cutoff, but I fully support allowing people who invest the time to get really good a few more chances.
 

Mike Hughey

Administrator
Staff member
Joined
Jun 7, 2007
Messages
11,303
Location
Indianapolis
WCA
2007HUGH01
SS Competition Results
YouTube
Visit Channel
I really believe it's better to just stick with the current approach. In almost all cases where this could be used, it will be just as easy to run another round. And I'd prefer to get another round, rather than get 2 more chances on the same round - to me, another round is better in nearly every way than 2 more solves on the same round.
 

TheGrayCuber

Member
Joined
Feb 21, 2012
Messages
188
WCA
2012GRAY01
YouTube
Visit Channel
I really believe it's better to just stick with the current approach. In almost all cases where this could be used, it will be just as easy to run another round. And I'd prefer to get another round, rather than get 2 more chances on the same round - to me, another round is better in nearly every way than 2 more solves on the same round.

This is a good point, but there is the problem of limited number of rounds. If we had 5 attempts a round, you could get 15 in one competition, where the max is 9 right now.
 

abunickabhi

Member
Joined
Jan 9, 2014
Messages
6,687
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
The Bo3/Bo5 model is quite different from Bo2/Ao5 that is followed in combined first rounds of sighted events.
I am for Bo3/Bo5 , as the 2 solves , are very much needed statistically and psychologically for the competitor . By analysis from statistics , we can reduce the UWR - WR gap by having 2 extra solves , which can give a better single because of more chances.

And psychologically because , AFAIK the competitor can get really nervous after 2 DNFs in a Bo3 , which makes the 3rd solve unnatural and stressful and so result-orientated for him/her, so the attempt gets reduced to a 2-proper solve attempt. And another nasty case that happens , a bit in 3BLD and more in bigblind , is when the first solve is DNFed. When the first solve is DNFed , then the other 2 solves are performed under different mindset giving us an inferior performance because of nerves.

If we have a Bo3/Bo5 system , both these bad scenarios will be avoided. The only thing that remains a problem is new comers will not be able to get a mean , and get the 5 complete solves , which can demotivate them. And also the competition schedule may get delayed due to Bo5 , if there are many 1-5 minute solvers. I think this issues are bearable issues , and do not outweigh , the benefits that Bo3/Bo5 will bring.
 

porkynator

Member
Joined
Oct 27, 2010
Messages
1,322
Location
Belluno, Italy
WCA
2011TRON02
YouTube
Visit Channel
I really believe it's better to just stick with the current approach. In almost all cases where this could be used, it will be just as easy to run another round. And I'd prefer to get another round, rather than get 2 more chances on the same round - to me, another round is better in nearly every way than 2 more solves on the same round.

I disagree, for at least two reasons:

1. If you have more rounds, you want to be sure to pass to the next one every time. This often means that some of the solves (one or two per round) have to be safe solves. With less solves per round, many of one's solve are going to be safe attempts. Extra solves each round means that one can probably do more solves knowing that they will pass to the next round anyway, and so without pressure.
2. I think Avg5 is a better format than Mo3 for the average (see also my remarks in the original post). This is less important than the previous argument in my opinion, but it is one more reason for changing.
 

mark49152

Premium Member
Joined
Oct 29, 2012
Messages
4,719
Location
UK
WCA
2015RIVE05
YouTube
Visit Channel
1. If you have more rounds, you want to be sure to pass to the next one every time. This often means that some of the solves (one or two per round) have to be safe solves.
How is that different to your proposal? One round of bo3/bo5 is comparable to two rounds of bo3. In both cases competitors may need to make a safety solve in the first bo3. (And with two rounds, you'd have one extra all-out attempt in the second round.)
 

porkynator

Member
Joined
Oct 27, 2010
Messages
1,322
Location
Belluno, Italy
WCA
2011TRON02
YouTube
Visit Channel
How is that different to your proposal? One round of bo3/bo5 is comparable to two rounds of bo3. In both cases competitors may need to make a safety solve in the first bo3. (And with two rounds, you'd have one extra all-out attempt in the second round.)

With Bo3/Bo5 you have the same pressure in the first three attempts, but once you get a good time in one of them you get two more solves to try and get a good time.
 

mark49152

Premium Member
Joined
Oct 29, 2012
Messages
4,719
Location
UK
WCA
2015RIVE05
YouTube
Visit Channel
With Bo3/Bo5 you have the same pressure in the first three attempts, but once you get a good time in one of them you get two more solves to try and get a good time.
With two rounds of Bo3 you have the same pressure in the first three attempts, but once you get a good time in one of them you get three more solves to try and get a good time.
 
Top