Blindsolving Accomplishment Thread

[Cale S]

first 5 attempts today:

5:15.62, 5:37.22, 5:06.34 5:05.00, 5:53.93 = 5:19.73 5BLD avg5

not even pb

 

[Daniel Lin]

this is my first time ever getting 2 sub 20's in a row

Mean of 3: 22.15
1. 19.67 B' U L U' B' R' L' D' R U' R2 F2 D F2 B2 R2 U B2 D' F2
2. 18.05 U' L F2 D R2 F D L F D2 B2 L2 U F2 R2 U R2 D' F2 D
3. 28.72 D F2 U2 F2 U B2 L2 U B2 D U R U' R D L2 F2 L' U' F'

 

[T1_M0]

I could start a 3bld race thread, would somebody be interested? Don't know yet which would be a good format though. I was considering something like 3 or 5 mo3:s per week. Graduate when a sub-x single 3 or 5 rounds in a row. Suggestions?

 

[Christopher Cabrera]

I could start a 3bld race thread, would somebody be interested? Don't know yet which would be a good format though. I was considering something like 3 or 5 mo3:s per week. Graduate when a sub-x single 3 or 5 rounds in a row. Suggestions?

I was planning on starting a general blindfolded race thread with 3 BLD, 4 BLD, 5 BLD, and Multi
my thoughts on format were that 4, 5, and Multi should be held by WCA standards (best attempt out of 3 for 4 and 5 BLD and best of 1 attempts for multi. My thought for 3 bld was to give around 12 scrambles where you take the average of the fastest half your attempts. If more than half of your results are dnf then you receive a DNF

 

[Fábio De'Rose]

PB Mo3

Generated By csTimer on 2017-5-4

mean of 3: 1:47.53

Time List:

1. 1:31.79 [36.84] F2 R2 U' F2 L2 R2 F2 D2 L2 D U L' U' L2 R F2 L B2 F L2 Rw Uw2
   
   
2. 2:01.62 [57.32] B2 U2 L' F2 R U2 F2 U2 R' D2 F2 D' L2 R F2 L B F' L' F2 L2 Fw Uw'
   
   
3. 1:49.16 [54.63] U' L U' L F L' F2 U' L D' F2 U' F2 U F2 U L2 U2 B2 L2 Fw Uw'
   

 

[Fábio De'Rose]

PB Ao12!

Generated By csTimer on 2017-5-4
avg of 12: 2:06.50

Spoiler: Scrambles

Time List:
1. 1:50.52 B F2 D R2 D' F2 D' R2 U B2 L2 F D' F2 U' R' B F' D' U2 Fw'
2. 1:56.85 L2 F2 U2 F D2 F U2 R2 B' R2 F2 R F' U' F2 U' L2 B2 R B' F Fw' Uw2
3. 2:36.96 R' D2 B2 R2 F D2 L2 F R2 F L U F U2 F D R2 D2 R2 Uw
4. 2:21.21 R2 F2 L' F' U R' B L U' R' F2 D2 F2 R B2 R B2 L2 U2 L' Fw'
5. 1:47.63 L R2 U' F2 U2 F2 U R2 U B2 L2 F2 B' U' F' D' B L2 R B' F' Fw Uw
6. 2:05.62 B2 R2 F R2 F' U2 F R2 B' F' U2 R' B' R' D' L F' L B L' D2 Fw' Uw'
7. 2:22.70 L' F U2 B F2 R2 B L2 R2 F L2 U2 L U F R' F L U' Rw' Uw
8. 1:50.59 L2 B2 R2 U' F2 L2 R2 U' L2 D2 U L' R' F2 D B2 R' B F R' B Fw' Uw2
9. 2:09.69 D B L2 U2 B' D2 R' U B' L2 U2 L2 U B2 U2 R2 B2 D2 R2 U' Rw
10. 2:03.20 B2 D2 R2 B' U2 F D2 F U2 R2 F2 D F2 R B' R2 D' B2 L B' D' Uw2
11. (3:14.96) R2 F' U2 R2 B F2 U2 R2 D2 R2 U2 L' D' F' D' R' F R' B' F2 U Rw Uw
12. (1:29.18) U2 B2 D' F2 R2 U2 B2 R2 D' L2 U2 L' D2 B2 L2 R D R2 F' U2 L' Rw Uw2

 

[T1_M0]

I was planning on starting a general blindfolded race thread with 3 BLD, 4 BLD, 5 BLD, and Multi
my thoughts on format were that 4, 5, and Multi should be held by WCA standards (best attempt out of 3 for 4 and 5 BLD and best of 1 attempts for multi. My thought for 3 bld was to give around 12 scrambles where you take the average of the fastest half your attempts. If more than half of your results are dnf then you receive a DNF

I don't know because just single really matters in blind. If you get one good solve out of three, that is your real level rather than some average.

 

[Christopher Cabrera]

I don't know because just single really matters in blind. If you get one good solve out of three, that is your real level rather than some average.

This is true for official competitions. However, if you are trying to gauge your actual skill in the event you would want to base it off of an average rather than just getting a lucky single. If you look at other race threads such as 3x3, 4x4, or pyraminx they base their results off of an average of 12 rather than averages of 5 because 12 gives you a better idea of where you actually stand.

 

[sqAree]

This is true for official competitions. However, if you are trying to gauge your actual skill in the event you would want to base it off of an average rather than just getting a lucky single. If you look at other race threads such as 3x3, 4x4, or pyraminx they base their results off of an average of 12 rather than averages of 5 because 12 gives you a better idea of where you actually stand.

However, for 3BLD ao5s and ao12s are significantly worse than what you would get (as a lucky single or even mo3) at a comp as we have to go really safe to pull it off.
So a better estimate would be something like mean of all the bo3s of x solves or something (just an idea).

 

[mark49152]

However, for 3BLD ao5s and ao12s are significantly worse than what you would get (as a lucky single or even mo3) at a comp as we have to go really safe to pull it off.

This is the reason for taking the mean of the best 50% of attempts. If you go safe and achieve a much higher success rate you are compensated by deletion of your slower successes.

 

[sqAree]

This is the reason for taking the mean of the best 50% of attempts. If you go safe and achieve a much higher success rate you are compensated by deletion of your slower successes.

At least we agree that an ao12 doesn't make much sense to consider.
But also about this idea that might make sense when comparing home averages: If I go safe with a high success rate let's say I will never get a sub2 single, thus no matter how many solves are deleted, I'm not sub2 by measurement. However, when I compete and do 3 fast attempts I will almost always have a sub2 single in there (just an example).
So I guess your system works fine for comparison purposes in the race thread, but as T1_M0 pointed out not for estimating one's real level (in respect to official attempts).

 

[T1_M0]

To all of you, could it work to take like 3 rounds per week, all rounds have three solves and the best attempt counts. It would represent official solves fairly well and if you need to get the sub-x single like 5 rounds in a row, you still always have to have the success to not get a dnf. That would represent your real level much better than any average.

 

[Rcuber123]

To all of you, could it work to take like 3 rounds per week, all rounds have three solves and the best attempt counts. It would represent official solves fairly well and if you need to get the sub-x single like 5 rounds in a row, you still always have to have the success to not get a dnf. That would represent your real level much better than any average.

Maybe average of 5 best of 3s?

 

[Christopher Cabrera]

To all of you, could it work to take like 3 rounds per week, all rounds have three solves and the best attempt counts. It would represent official solves fairly well and if you need to get the sub-x single like 5 rounds in a row, you still always have to have the success to not get a dnf. That would represent your real level much better than any average.

How would these results be ranked in a race-like format?

Maybe average of 5 best of 3s?

Is this average of 5 like in 3x3 or is it a mean of the 5 solves? If it's the latter, that is essentially what I proposed except you would be removing 66% of the bad solves as opposed to removing 50%. I'd personally like to set the standard at 50 % success rate, but I could get behind this ideology. 15 solves with a mean of the best 5.

 

[T1_M0]

How would these results be ranked in a race-like format?

Same as for example 4x4 race. That has 2 rounds per week, this would have 3 rounds. If you get a sub x single in round 1a, 1b, 1c, 2a and 2c, you graduate. What is unclear, I hope we can find out as good race format as possible

 

[Christopher Cabrera]

Same as for example 4x4 race. That has 2 rounds per week, this would have 3 rounds. If you get a sub x single in round 1a, 1b, 1c, 2a and 2c, you graduate. What is unclear, I hope we can find out as good race format as possible

I would prefer to consolidate all a person's results into one number. That's why I keep proposing the format of an average. I feel like if we take each result from all your "best of 3" rounds and average those together it would give you a good idea of what you could expect from an official competition in the long run. Under these conditions you would get 1 result every week but you would be required to do 3 successful results in order to graduate.

 

[T1_M0]

I would prefer to consolidate all a person's results into one number. That's why I keep proposing the format of an average. I feel like if we take each result from all your "best of 3" rounds and average those together it would give you a good idea of what you could expect from an official competition in the long run. Under these conditions you would get 1 result every week but you would be required to do 3 successful results in order to graduate.

That could work, you mean like 3x mo3 per week and average of their best singles? Nice idea

 

[Christopher Cabrera]

That could work, you mean like 3x mo3 per week and average of their best singles? Nice idea

Not sure if it would be better to do 3 means of 3 or we could do 5 means of 3 like Rcuber said. I think 3 would appeal a bit more to people of a lower skill level.

 

[mark49152]

So suggestion #1 was average the best 50%, for example do 12 solves and average the best 6.

Suggestion #2 is do 15 solves and average the best 5, so the best 33.333% instead of 50%. Except those 5 must be taken one each from 5 groups of 3.

I don't understand how organising it as 5 x bo3 gives a better measure of performance than just doing 15 attempts and picking the best 5.

 

[sigalig]

Maybe this could be taken to the BLD discussion thread?