[Help Thread] Multi-Blind Discussion

[Torch]

1/5 in 16:23
I was really confident with my memo as well, 3 of the 4 unsolved were off by twisted corners and solved edges

Do you still use audio?

 

[YouCubing]

Do you still use audio?

yeahhh

 

[G2013]

Official 3/6 19:21 >.<
Heh, but I've got the WR for the fastest 3/6 attempt xD

DNFs were:

1: flipped edge
2: 2 twisted corners
3: asdfasdfasdfasdf idk what happened, probably I did some extra move or something.

xd the NR is 6/6 in 40something minutes lol, I went for that but failed like a pro

 

[kake123]

25/25 MBLD in 1hr 1min 33.04s+

I plus 2ed one of the cubes, and 40min memo

 

[the super cuber]

37/49 1:19:41 [51:50] yesterday
41/49 1:16:11 [49:50] today
3 pops + 1 corner twist god why, my cubes suck
without that it would be 44/49 or 45/49 (3 cubes are completly solved with one twisted piece and 1 is messed up due to pop in the middle of algorithm)
other 4 mistakes are small and avoidable
there's a lot of potential to be faster and better, if only I'll keep such concentration as today

IMG_20170326_213317

• kamilprzyb
• Mar 26, 2017
• 1

wow you got really fast! nice job

 

[Fábio De'Rose]

Going for a 3/3 Multi in a while.

Will get some coffee first. Wish me luck fam.


Edit: 2/3 in 46:26.66

Memorized the 3rd cube in a wrong orientation and realized it too late. Decided to keep going instead of redo it, then fix centers. Got anxious, which ruined my concentration and I ended up flipping an edge twice. The other 2 cubes were alright.

Multi is pretty fun. I'm taking it slow so I can organize my rooms and optimize their space.

 

[Fábio De'Rose]

Another shot at 3 points, 2/3 in 47.xx

Third cube (again) was off by 2 edges. I celebrated too early, lol. My third room is a bit messy, that's probably what caused the DNF. Overall I'm happy to know I am at least passable at the event. Plenty of improvements to be made until Nationals in July.

 

[Keroma12]

After over 30 attempts, I finally got a 7/7.
Pretty depressing since I got 6/6 on the first try.
On to 8 now I guess.

 

[newtonbase]

After over 30 attempts, I finally got a 7/7.
Pretty depressing since I got 6/6 on the first try.
On to 8 now I guess.

Well done.
It's funny how that happens. I took 11 attempts to get 4 but only 2 to get 5. I've had 4/6 on both attempts at 6.
What sort of errors have you been making?

 

[Keroma12]

Well done.
It's funny how that happens. I took 11 attempts to get 4 but only 2 to get 5. I've had 4/6 on both attempts at 6.
What sort of errors have you been making?

Thanks.

Getting the wrong letter from the correct piece.
For example, if I am looking at the red-blue edge, and in the red-blue spot there is the green-red edge, then I sometimes pick the red-green letter instead of the green-red letter (I don't use the standard letter scheme, and the problem usually (but not always) arises from the matching colours). This also happens with corners.

Occasionally during edges, I'll pick the letter which is mirrored across the M slice. So if I need the FL letter, I might pick the FR letter instead. This is much less common, and only on edges.

Next most common mistakes are forgetting to undo a setup, then forgetting memo (more so in 19 cubes attempts, but still not very frequent).

(I have a spreadsheet keeping track of my mistakes in each attempt.)

 

[newtonbase]

Thanks.

Getting the wrong letter from the correct piece.
For example, if I am looking at the red-blue edge, and in the red-blue spot there is the green-red edge, then I sometimes pick the red-green letter instead of the green-red letter (I don't use the standard letter scheme, and the problem usually (but not always) arises from the matching colours). This also happens with corners.

Occasionally during edges, I'll pick the letter which is mirrored across the M slice. So if I need the FL letter, I might pick the FR letter instead. This is much less common, and only on edges.

Next most common mistakes are forgetting to undo a setup, then forgetting memo (more so in 19 cubes attempts, but still not very frequent).

(I have a spreadsheet keeping track of my mistakes in each attempt.)

A bit of everything then. I'm pretty much the same, memo mostly then execution then recall. I wasn't checking that the memo was correct in reviews, just making sure I could recall it. For my 4 cube success I did full reviews and solved very carefully. It added 50% to the time but was worth it.

 

[newtonbase]

Does anyone else use magnetised cubes for MBLD? I have a slight issue with some cubes sticking to each other and some repelling. It hasn't affected a result yet but I do have to be extra careful picking them up.

 

[bobthegiraffemonkey]

Does anyone else use magnetised cubes for MBLD? I have a slight issue with some cubes sticking to each other and some repelling. It hasn't affected a result yet but I do have to be extra careful picking them up.

I wouldn't have expected the magnets from neighbouring cubes to be able to interact much, that's interesting. Did you use particularly strong magnets?

 

[newtonbase]

I wouldn't have expected the magnets from neighbouring cubes to be able to interact much, that's interesting. Did you use particularly strong magnets?

I use N35 4mm x 3mm. They are about 30% stronger than the ones used most often. There's sufficient force that when you pick up a cube it can drag another just enough to change the orientation. It's just something to be wary of.

 

[kake123]

23/25 MBLD in 1hr 30.07s

Roughly 40min memo and execution was very fast for me, so quite good

 

[Fábio De'Rose]

2/3 in 28:08. Off by two corners and one Edge cycle.

Huge improvement since the last attempt, cut a whooping 19 minutes and finally managed to do it under 30.

Gotta do some more attempts. My goal is steady sub-40min, 4 points, until Nationals in July.

 

[KAINOS]

Proposal of a new MBLD scoring system

There are 2 major 'flaws' (in my opinion, at least) in current MBLD scoring scheme:
1) Time does not matter at all unless it is over time limit
You can say 10/10 in 30:00 is much better result than 11/11 in 60:00, right? But according to current rules it's 10 vs 11 and the second one wins the game, regardless of the solving time.
2) Points are determined by (number of solved cubes)-(number of unsolved cubes), not number of solved cubes only
I know this could be kind of controversial, but to me being penaltied by unfinished cubes seems to be unfair. For example, why should 5/10 attempt counted as 0 point and then DNFed when the competitor succesfully solved 5 cubes? He/She did complete exactly same amount of cubes with the one who did 5/5, but the result is nothing but DNF only because of the rest of the cubes, remaining unsolved. That's kinda nonsensical, I should say.

So, to fix these problems I've thought of new scoring system. The formula is:

(Number of solved cubes)/(Time (in hour))^(1/2)

To briefly explain how I came up with this, it is geometric mean of (number of solved cubes) and (time per cube). In this way both quantity and speed are equally evaluated.

Examples:
10/10 in 30:00 - 10/0.5^(1/2) = 14.14 points (You can see it beats 11/11 in 60:00.)
41/41 in 54:14 (Yup, it's Maskow's WR result) - 41/0.90389^(1/2) = 43.12 points

Let me know if you have any questions or opinions on this - I'm pretty eager to hear those

 

[Christopher Cabrera]

Proposal of a new MBLD scoring system

There are 2 major 'flaws' (in my opinion, at least) in current MBLD scoring scheme:
1) Time does not matter at all unless it is over time limit
You can say 10/10 in 30:00 is much better result than 11/11 in 60:00, right? But according to current rules it's 10 vs 11 and the second one wins the game, regardless of the solving time.
2) Points are determined by (number of solved cubes)-(number of unsolved cubes), not number of solved cubes only
I know this could be kind of controversial, but to me being penaltied by unfinished cubes seems to be unfair. For example, why should 5/10 attempt counted as 0 point and then DNFed when the competitor succesfully solved 5 cubes? He/She did complete exactly same amount of cubes with the one who did 5/5, but the result is nothing but DNF only because of the rest of the cubes, remaining unsolved. That's kinda nonsensical, I should say.

So, to fix these problems I've thought of new scoring system. The formula is:

(Number of solved cubes)/(Time (in hour))^(1/2)

To briefly explain how I came up with this, it is geometric mean of (number of solved cubes) and (time per cube). In this way both quantity and speed are equally evaluated.

Examples:
10/10 in 30:00 - 10/0.5^(1/2) = 14.14 points (You can see it beats 11/11 in 60:00.)
41/41 in 54:14 (Yup, it's Maskow's WR result) - 41/0.90389^(1/2) = 43.12 points

Let me know if you have any questions or opinions on this - I'm pretty eager to hear those

Multi-blind is a much different event than all the other ones in cubing. In most other events time is the main objective and is almost the sole criteria for determining a winner. However, I feel like the main point of multi blind is to solve as many cubes blindfolded as accurately as you can.Even though there is a time limit, time it should come third to these two criteria (Volume and Accuracy).

1.) Since the whole point of multi blind is to solve as many cubes as you can blindfolded it makes sense that the number of cubes solved should take priority over time taken to solve them. (provided they don't go over the time limit of course)

2.) Accuracy and precision are two of the most fundamental principles, not just for Multi blind , but for blind solving in general. People should be rewarded for knowing there limits rather than just brute forcing a bunch of cubes and not worring bout their mistakes.

Also, a problem with your current formula is that a competitor can show up and present 100 cubes to the judges with the intention of only solving 10. This would become very irritating to judges and scramblers. There is nothing in your system that discourages this kind of behavior.

 

[kamilprzyb]

Proposal of a new MBLD scoring system

There are 2 major 'flaws' (in my opinion, at least) in current MBLD scoring scheme:
1) Time does not matter at all unless it is over time limit
You can say 10/10 in 30:00 is much better result than 11/11 in 60:00, right? But according to current rules it's 10 vs 11 and the second one wins the game, regardless of the solving time.
2) Points are determined by (number of solved cubes)-(number of unsolved cubes), not number of solved cubes only
I know this could be kind of controversial, but to me being penaltied by unfinished cubes seems to be unfair. For example, why should 5/10 attempt counted as 0 point and then DNFed when the competitor succesfully solved 5 cubes? He/She did complete exactly same amount of cubes with the one who did 5/5, but the result is nothing but DNF only because of the rest of the cubes, remaining unsolved. That's kinda nonsensical, I should say.

So, to fix these problems I've thought of new scoring system. The formula is:

(Number of solved cubes)/(Time (in hour))^(1/2)

To briefly explain how I came up with this, it is geometric mean of (number of solved cubes) and (time per cube). In this way both quantity and speed are equally evaluated.

Examples:
10/10 in 30:00 - 10/0.5^(1/2) = 14.14 points (You can see it beats 11/11 in 60:00.)
41/41 in 54:14 (Yup, it's Maskow's WR result) - 41/0.90389^(1/2) = 43.12 points

Let me know if you have any questions or opinions on this - I'm pretty eager to hear those

No

 

[KAINOS]

Multi-blind is a much different event than all the other ones in cubing. In most other events time is the main objective and is almost the sole criteria for determining a winner. However, I feel like the main point of multi blind is to solve as many cubes blindfolded as accurately as you can.Even though there is a time limit, time it should come third to these two criteria (Volume and Accuracy).

1.) Since the whole point of multi blind is to solve as many cubes as you can blindfolded it makes sense that the number of cubes solved should take priority over time taken to solve them. (provided they don't go over the time limit of course)

2.) Accuracy and precision are two of the most fundamental principles, not just for Multi blind , but for blind solving in general. People should be rewarded for knowing there limits rather than just brute forcing a bunch of cubes and not worring bout their mistakes.

Also, a problem with your current formula is that a competitor can show up and present 100 cubes to the judges with the intention of only solving 10. This would become very irritating to judges and scramblers. There is nothing in your system that discourages this kind of behavior.

I perfectly understand what you mean. Especially 2) showed my logic was far from perfection. Yet, I still insist on 2 things: 1. Time should be counted as a factor of scoring. 2. With current rules, penalties with unsolved cubes are a little too harsh.
From that, I've tweaked my formula a little bit:

((Number of solved cubes)-(Number of unsolved cubes)*x)/(Time (in hour))^y

Both x and y are constants that I'm unsure about their appropriate value, but I think x should be around 0.5~0.8 (to soften the penalty) and y be around 0.1~0.3 (to apply weaker 'speed bonus'). This should make the formula more moderate.

And to be honest, I'm a total MBLD noob - my PB is only 2/2 in 9:39. I've started to think if posting my idea was a mistake at the first place since I do not know very well about the event...