Alternative Multiblind Scoring System

[TheSpicyBlueFlamingo]

So recently, I've gotten into blind solving and more recently, multiblind. But here's my problem: I completely disagree with the scoring/ranking system for multiblind. Here's the current system:
Regulation 9f12c) ) For 3x3x3 Multi-Blind, rankings are assessed based on the number of puzzles solved minus the number of puzzles not solved, where a greater difference is better. If the difference is less than 0, or if only 1 puzzle is solved, the attempt is considered unsolved (DNF). If competitors achieve the same result, rankings are assessed based on total time, where the shorter recorded time is better. If competitors achieve the same result and the same time, rankings are assessed based on the number of puzzles the competitors failed to solve, where fewer unsolved puzzles are better.

In my opinion, multiblind rankings should not be affected whatsoever by the number of cubes unsolved. In any given situation where individuals or teams are ranked, the objective is to assess a certain skill. For example, rankings for the 100m dash are determined my how quickly one can run 100 meters. In 3x3 speedsolving, rankings are determined based on how fast an individual can solve a 3x3 cube. Based on Regulation 9f12c), we can gather that the skill that's being assessed for multiblind is as follows: Who can obtain the greatest difference of cubes solved minus cubes unsolved. In other words, a competitor who solves 8 cubes out of 13 would get a score of 5 points. This makes no sense! In the term "Multiple Blindfolded Solving", the skill being assessed should be fairly evident: Who can solve the most cubes blindfolded (time limit notwithstanding). Therefore, I believe that if a competitor solves 9 cubes out of 13, their score should be 9, because they successfully solved 9 cubes blindfolded.
The only downside to this proposal would be that someone could attempt 6 cubes for the sole purpose of getting more time, and only actually try to solve 3 of them.
Let me know if you have any thoughts on this.

 

[tx789]

NO

n/n attempts are really hard and not considering accuracy in the score is removing a important part of the metric.

A 45/45 losing to a 46/60 just isn't right. The 46/60 cuber would be much more annoyed about their attempt.

Also having people attempt 6 so they can get 2 point is a huge waste of competition time and judges.



While people have issues with the current system the proposal is worse.

MBLDers wouldn't like this. Auccuray is important having high auccuray is a skill.

 

[TheSpicyBlueFlamingo]

I disagree; if someone got a 45/45 that means they also attempted 45 cubes. Therefore, that person has the confidence and the ability to solve 45 cubes in one hour. Meanwhile, Someone who got a 46/60 would not only have solved more cubes, but would also have the confidence to be able to solve 60 cubes in an hour, thus solving more cubes and solving those cubes faster.
That being said, I agree with you on your other points, and most people will be more experienced than I am (I've been cubing for 2 years).

 

[Samuel Baird]

At the end of the day it boils down to opinion. I think accuracy should play a big role as well as number of cubes solved which is why I personally favor the current system which values both accuracy and number solved.

This system also forces competitors to not waste competition time by doing things like filtering scrambles. For example, if my memo was bad I could submit 1000 cubes, trace all of them, determine the 2 cubes with the shortest memo, and then solve those cubes

 

[brododragon]

I disagree; if someone got a 45/45 that means they also attempted 45 cubes. Therefore, that person has the confidence and the ability to solve 45 cubes in one hour. Meanwhile, Someone who got a 46/60 would not only have solved more cubes, but would also have the confidence to be able to solve 60 cubes in an hour, thus solving more cubes and solving those cubes faster.
That being said, I agree with you on your other points, and most people will be more experienced than I am (I've been cubing for 2 years).

So if I'm confident I can get a sub-10 solve 3x3 solve (but get a minute solve), I should be win to someone who doesn’t have confidence, but gets sub-10?

This scoring system is SOOOO broken.

The only downside to this proposal would be that someone could attempt 6 cubes for the sole purpose of getting more time, and only actually try to solve 3 of them.

The only downside is someone could infinitely expand there time? Say I wanted to beat the WR. he had to do all the cubes sub-hour. I could just attempt a thousand and I've got way more time for 60.

This system gives absolutely no punishment for messing up. Does that mean in every other event DNFs shouldn't hurt my time?

 

[TipsterTrickster]

The only downside is someone could infinitely expand there time? Say I wanted to beat the WR. he had to do all the cubes sub-hour. I could just attempt a thousand and I've got way more time for 60.

You can’t infinitely expand your time in multi. It’s 10 min per cube up to 1hr.

anyway, The best scoring system I’ve seen so far is (number solved - number unsolved) + (number attempted/100).
This would score a 7/8 as 6.08 which would rank it higher than a 6/6 which would score 6.06. However, this gets kinda murky when we get to 100+ cubes attempted, so I would say put it in as x.100 and have that rank higher than x.10-x.99. Or even just keep the same scoring system we currently have and break ties with cubes attempted then time taken rather than just going straight to time taken for tie breakers.

 

[brododragon]

You can’t infinitely expand your time in multi. It’s 10 min per cube up to 1hr.

anyway, The best scoring system I’ve seen so far is (number solved - number unsolved) + (number attempted/100).
This would score a 7/8 as 6.08 which would rank it higher than a 6/6 which would score 6.06. However, this gets kinda murky when we get to 100+ cubes attempted, so I would say put it in as x.100 and have that rank higher than x.10-x.99. Or even just keep the same scoring system we currently have and break ties with cubes attempted then time taken rather than just going straight to time taken for tie breakers.

I think the problem is that sometimes you have to sacrifice accuracy for simplicity. thanks

 

[m0nkiem0nkie]

Although not particular a comment on the scoring system, but rather on the subject of the skills being assessed from a MBLD-er, quoting OP’s “Who can obtain the greatest difference of cubes solved minus cubes unsolved” versus “Who can solve the most cubes blindfolded”, I propose to reassess first what kind of skills we would like to see showcased by a MBLD-er during a competition:

Although I have much respect for the skills to blind solve a multitude of 3x3’s in an hour, I would rather see this as an occasional or special event when for instance a local, national or WR is being attempted.

In its place, I would prefer a regular MBLD-competition to look like this:
Solve a scrambled group of a 2x2, 3x3, 4x4 and 5x5 cube in the fastest way possible; blindfolded and memorization time included, of course. It could even be expandable to include a 6x6 and a 7x7 cube. One unsolved cube results in a total DNF to maintain the level of accuracy.

In this way, the competitive level and balance of speed and accuracy remains and would most likely expand the participation field of MBLD-competitors; not to mention that it would also be much more attractive for the audiences as well to witness these types of competitions.

 

[Dylan Swarts]

So recently, I've gotten into blind solving and more recently, multiblind. But here's my problem: I completely disagree with the scoring/ranking system for multiblind. Here's the current system:
Regulation 9f12c) ) For 3x3x3 Multi-Blind, rankings are assessed based on the number of puzzles solved minus the number of puzzles not solved, where a greater difference is better. If the difference is less than 0, or if only 1 puzzle is solved, the attempt is considered unsolved (DNF). If competitors achieve the same result, rankings are assessed based on total time, where the shorter recorded time is better. If competitors achieve the same result and the same time, rankings are assessed based on the number of puzzles the competitors failed to solve, where fewer unsolved puzzles are better.

In my opinion, multiblind rankings should not be affected whatsoever by the number of cubes unsolved. In any given situation where individuals or teams are ranked, the objective is to assess a certain skill. For example, rankings for the 100m dash are determined my how quickly one can run 100 meters. In 3x3 speedsolving, rankings are determined based on how fast an individual can solve a 3x3 cube. Based on Regulation 9f12c), we can gather that the skill that's being assessed for multiblind is as follows: Who can obtain the greatest difference of cubes solved minus cubes unsolved. In other words, a competitor who solves 8 cubes out of 13 would get a score of 5 points. This makes no sense! In the term "Multiple Blindfolded Solving", the skill being assessed should be fairly evident: Who can solve the most cubes blindfolded (time limit notwithstanding). Therefore, I believe that if a competitor solves 9 cubes out of 13, their score should be 9, because they successfully solved 9 cubes blindfolded.
The only downside to this proposal would be that someone could attempt 6 cubes for the sole purpose of getting more time, and only actually try to solve 3 of them.
Let me know if you have any thoughts on this.

I strongly disagree with this. Although someone brought this up at our nationals last year where a guy won multi-blind with a 2/2 and I came 2nd with 5/9. It was quite a sight to see someone do 9!? cubes in South Africa, as the largest amount attempted the previous year was 3 or 4, so I understand why that might have played a role in wanting to declare my attempt better than the 2/2. I disagreed, for basically the explanation that follows:

If someone went and attempted 10 cubes, and only managed to solve 7, it is fair to now have 4 points. The reason why: Well, I think it can be seen as a way of 'punishing' the person for trying to do 10 and failing to solve all of them/more of them. Sure, you mess up sometimes when you could have gotten 10/10, but so be it.
To look at it in another way: Attempting 10 because you think you can probably get 8 or 9 on 10 cubes, and knowing that attempting say 13 will probably result in more dnfs and then a much lower point. If scoring were to work in the way you suggest, you can just go ahead and do 13, because there is a chance of getting maybe 10/13, but also still the possibility of getting 8 or 9 out of 13, but it won't have such a negative effect on you point as with the current point calculation would. This way the current system sort of prevents you from just picking a large number and having a bigger chance of getting a larger amount of solved cubes.

 

[AlphaCuber is awesome]

I could just "attempt" 60 cubes and go through and look for the easiest 10 scrambles and have an easier 10 cube multi

 

[brododragon]

Although not particular a comment on the scoring system, but rather on the subject of the skills being assessed from a MBLD-er, quoting OP’s “Who can obtain the greatest difference of cubes solved minus cubes unsolved” versus “Who can solve the most cubes blindfolded”, I propose to reassess first what kind of skills we would like to see showcased by a MBLD-er during a competition:

Although I have much respect for the skills to blind solve a multitude of 3x3’s in an hour, I would rather see this as an occasional or special event when for instance a local, national or WR is being attempted.

In its place, I would prefer a regular MBLD-competition to look like this:
Solve a scrambled group of a 2x2, 3x3, 4x4 and 5x5 cube in the fastest way possible; blindfolded and memorization time included, of course. It could even be expandable to include a 6x6 and a 7x7 cube. One unsolved cube results in a total DNF to maintain the level of accuracy.

In this way, the competitive level and balance of speed and accuracy remains and would most likely expand the participation field of MBLD-competitors; not to mention that it would also be much more attractive for the audiences as well to witness these types of competitions.

So replace MBLD with BLD Relay? No. There are many people who love it and would hate to have few opportunities at it. It's an entirely different event. MBLD is how much can you solve in the allotted time, BLD Relay is just how fast. You're replacing a 1-of-a-kind event with just the normal event format. It would not expand the competitor number; more people can do 2/4 MBLD (minimum score) then 4x4/5x5/6x6/7x7 BLD. Also, have you watched Big BLD? It's not more entertaining than MBLD. In fact, I would rather watch MBLD than Relay BLD, and I'm sure many people can agree with me.

 

[dudefaceguy]

There is already a blind event that prioritizes speed first and accuracy second: 3BLD. As long as you get one success, you can get 2 DNFs and it doesn't matter.

Big blind events prioritize accuracy more than 3BLD but you still get 3 chances, so you can have 40% accuracy and still get a world record.

MBLD prioritizes accuracy in 3x3 blindsolving. It is a great event for people who like to practice accurate solving. After all, it's much more impressive and satisfying to successfully solve a cube blindfolded every time, rather than failing twice and succeeding once. The scoring system hits exactly the right balance.

Changing to the scoring system you propose would make MBLD too similar to 3BLD. Competitors would just focus on speed without worrying much about accuracy. Doing many cubes in multiblind is not much more difficult than doing a single cube, since memorization techniques work so well. Top solvers could easily memorize and execute 150 cubes in one hour if they didn't review memo at all - this is 25 seconds per cube. Graham Siggins has multiple small attempts on YouTube in which he solves with 100% accuracy at about 30 seconds per cube. Throwing out accuracy could easily get that down to 25 seconds per cube, which is 144 cubes in an hour. This would lead to everyone attempting 150 cubes, because why not? As others have said, in addition to being silly, this is a colossal waste of time.

Switching to your proposal would lead to records looking like 80/150, which is just silly. I find it much more impressive to solve accuractly than to solve quickly, and I think that Graham Siggins should have the world record for 59/60 cubes, rather than someone who solves 62/150 cubes. If we switch to your scoring system, striving for 100% accuracy in blind solving will no longer be rewarded in competition, which would be a shame.

 

[Nmile7300]

My preferences is 1 pt for XP, 3 pt for all FG regardless of distance, and -2 for misses within 40 yards. Keep it simple.

Uh what?

 

[Zubin Park]

He's talking about football, I assume, and its definitely not right, especially while bumping a thread...

 

[Nmile7300]

He's talking about football, I assume, and its definitely not right, especially while bumping a thread...

Well yeah, but like... why?

 

[VIBE_ZT]

I could just "attempt" 60 cubes and go through and look for the easiest 10 scrambles and have an easier 10 cube multi

Exactly. End of discussion. You get to pick and choose your scrambles more. The way it should be is: You have to memo every cube you have. Unfortunately, this isn't the case. That's exactly why 2-cube MBLD attempts are DNF if 1/2 isn't solved. There should be a relatively similar punishment for >2 cubes. Like there is.

 

[AlphaCuber is awesome]

Exactly. End of discussion. You get to pick and choose your scrambles more. The way it should be is: You have to memo every cube you have. Unfortunately, this isn't the case. That's exactly why 2-cube MBLD attempts are DNF if 1/2 isn't solved. There should be a relatively similar punishment for >2 cubes. Like there is.

It actually isn’t about having to memo every cube you solve as the reason a 1/2 is a DNF is because you haven’t solved multiple cubes blindfolded which is the point of the event.

 

[dudefaceguy]

Well yeah, but like... why?

I believe the intention was to create a humorous effect.

 

[abunickabhi]

I find the current way of scoring good enough. But it demands a more careful approach towards selecting the number of cubes you want to attempt.

Another proposal I saw somewhere on the SS, which I liked very much is that MBLD points will be calculated as no. of cubes solved - no.of unsolved cubes/2. So a 8/13 attempt in an hour will be 5.5 points which seems a reasonable score accounting the effort and accuracy.

 

[Rosh Abin Abraham]

hi
I am a starter. I just started a few months ago. I feel your a member in this . Please guide me and write a reply.