Jakube
Veteran
Maybe. It definitely feels a bit unfair when someone with only 6 cubes beat you, although you solved 12 (of 18), like it happened to me last year.
Blindsolving Discussion and Help
- Thread starter Kenneth
- Start date
-
- Tags
- 2x2 bld 2x2 blind 3bld 3x3 m2 4x4 r2 advice begginers bld bld soleves blind blindfold blindfolded blindsolve buffer canoda centers corners cycle edges execution fold helgp images improving jackson letter m2method mbld memo memorisation memorization memos method multi bld multi blind multiple blindfolded old pochmann opa pair parity piece pochmann popop problem problems rubber rubiks solve solved speedsolver spheffz techniques turbo twisted pieces visual
Maybe. It definitely feels a bit unfair when someone with only 6 cubes beat you, although you solved 12 (of 18), like it happened to me last year.
kinch2002
Premium Member
It might feel like that initially, because you know you have the greater ability. But when you think about it, they did solve all of their cubes, while you missed 6, which isn't great at all
AbstractAlg
Veteran
:'|
Yea, it seams a bit unfair to me, aslo.
Where do you get all the cubes for home training? Just go and buy 20 cubes?
Last edited:
Jakube
Veteran
It might feel like that initially, because you know you have the greater ability. But when you think about it, they did solve all of their cubes, while you missed 6, which isn't great at all
Sure, your right. Now I also see, that his performance was way better. Being accurate is way harder than being fast. Nevertheless at my next comp I will try as many as possible again.
Jakube
Veteran
Yes, I started with 5-8 cubes I had at home. Because I was bying always the next better cube (W-Cube, Speed Cube Ultimate, AV, C3, Dayan, ...), and when I got faster and serious about multi, I started bying better more cubes, and also replaced some of the older crappier ones. I now have 29 3x3x3 in my room.
AbstractAlg
Veteran
dear lord. and I though that 2 cubes of mine was a success. respect!
now I need to earn 20x 15$ ~ 300$. nice.. -_-'
- Joined
- Jun 7, 2007
- Messages
- 11,923
- Location
- Indianapolis
- WCA ID
- 2007HUGH01
- YouTube
- Visit Channel
It might feel like that initially, because you know you have the greater ability. But when you think about it, they did solve all of their cubes, while you missed 6, which isn't great at all
One thing I have noticed is that, for me at least, the best strategy for winning at multiBLD in competition is in fact to go for as many cubes as you can comfortably solve in an hour. It doesn't pay to go for the absolute limit (although that might be the way to get world records, it doesn't pay for winning competitions), but it also doesn't pay to go smaller than the comfortable number. With the highest comfortable number, you give yourself a little leeway for an accidental missed cube or two, and still keep the small numbers people from really even having a chance.
To me, the fact that this strategy works well shows that our current scoring system is pretty well-tuned. It encourages people to go for high numbers of cubes, but doesn't encourage them to go so high that they fail hard.
DrKorbin
Veteran
There was a time when I asked myself: how many cubes I have to try in competition. Then I did some math.
Suppose that the probability of solving 1 cube is equal to p (i.e. among 100 cubes you solve 100p cubes successfully). Then if you try to solve N cubes then you solve Np cubes, you don't solve N(1-p) cubes and you get N(2p-1) points (in terms of expected value). What does it mean? If p < 1/2 then your result will be most likely DNF. If p = 1/2 then you can try any amount of the cubes and you are expected to get 0 points. If p > 1/2 then you must try as many cubes as you can memorize and solve within an hour, because the number of points you get is in proportion to N.
Suppose that the probability of solving 1 cube is equal to p (i.e. among 100 cubes you solve 100p cubes successfully). Then if you try to solve N cubes then you solve Np cubes, you don't solve N(1-p) cubes and you get N(2p-1) points (in terms of expected value). What does it mean? If p < 1/2 then your result will be most likely DNF. If p = 1/2 then you can try any amount of the cubes and you are expected to get 0 points. If p > 1/2 then you must try as many cubes as you can memorize and solve within an hour, because the number of points you get is in proportion to N.
kinch2002
Premium Member
If you can't get p>1/2 then you're never going to be particularly good at multi - that's fairly obvious.
The problem is that as N gets higher, p gets lower. You still have to try to find the optimal N for yourself.
Here's how I find my optimal N:
1. Just before the comp, do an attempt with an estimation of N
2. See how long I take for that attempt and figure out how many I would have done in 55 minutes by scaling.
3. Find an odd number near it - that's my N.
The problem is that as N gets higher, p gets lower. You still have to try to find the optimal N for yourself.
Here's how I find my optimal N:
1. Just before the comp, do an attempt with an estimation of N
2. See how long I take for that attempt and figure out how many I would have done in 55 minutes by scaling.
3. Find an odd number near it - that's my N.
AbstractAlg
Veteran
Why it has to be odd?
Also, how do you do mbld? You memorize in a row and repeat memo twice for example, or review every 4 cubes..? Or it is just up to personal preferences?
I think Daniel memorises each cube carefully and slowly (not meaning that in a bad way...) so that he doesn't need to review the cubes. Other people memorise in sets, and review the sets individually. For example I do sets of either 2 or 4 cubes. Depending on how fast you can memorise, the sets should be bigger. I think spaced repetition works well for multibld which is why I tried to make use of itWhy it has to be odd?
Also, how do you do mbld? You memorize in a row and repeat memo twice for example, or review every 4 cubes..? Or it is just up to personal preferences?
(Spaced repetition is the idea of revising something shortly before you would forget it. There's a lot of good reading material online about Spaced Repetition Systems. Programs like SuperMemo or Anki, or audio lessons like Pimsleur)
AbstractAlg
Veteran
I'm changing corner memo method. From tap+side letter (Eric Limeback type) to letter for each of 21 stickers. All because of mbld, because visual memo is not good for mbld, for obvious reasons.
I know that kinch002 uses 4 letters in CW direction per face, going through letter in alphabet. I got used to cell phone's buttons so I assigned for one corner orientations ABC, for second DEF, ...you get the point...all the way to TUV.
Do you have any ideas that are more simpler and/or some suggestions?
Thanks.
I know that kinch002 uses 4 letters in CW direction per face, going through letter in alphabet. I got used to cell phone's buttons so I assigned for one corner orientations ABC, for second DEF, ...you get the point...all the way to TUV.
Do you have any ideas that are more simpler and/or some suggestions?
Thanks.
kinch2002
Premium Member
That is pretty much correct. I usually memo sets of 4 slowly, then review the 4. Then move on and never come back to them until I'm solving. Most people seem to do around 4 cubes at a time, but going through them faster than me and hence also more reviews
Jakube
Veteran
You can setup the pieces with L and R:
f.I.: RD: R' (u R u' M2 u R' u') R
The question is, if this is faster than just doing a cube rotation. I tried this some days ago, and didn't liked it very much.
(You can also do LB this way: u R' u' (M2) u R u'
yeee707
Regular
Youtube isn't loading very well for me right now, so I was wondering if someone could answer my question. In Eric Limeback's M2/Pochmann tutorial, he solves corners first, then edges. If there is an odd number of edge shoots, the parity algorithm must be applied. If we do corners first, do we have to do the R PLL after finishing the corners, then solve edges and do the parity and R PLL again?
emolover
Legend
Why is the record for megaminx blind something 15 minutes when I has 55 pieces where as the record for 4 blind is 3 minutes when it has 56 pieces?
Last edited:
Mollerz
Swag Overlord
Because in 4BLD you have corners, 24 letters (8 pieces, 3 orientations for each piece), edges, 24 pieces (Fixed orientation), and centres, 24 pieces (Assuming none are solved which is never). So one can simple apply a lettering system for them.
In Megaminx... 30 edges, with 2 orientations, so 60 pieces. 20 corners, with 3 orientations, so another 60 pieces. You cannot apply this simple lettering system to it. It is thus much harder to memorise.
Also megaminx commutators are poop.
wontolla
Veteran
You deal with parity only once. If you do it twice, you will get once again two edges and two corners switched.
I would do the R PLL at the very end of the solve, no matter if you solve edges and then corners or the other way around.
1) memo is 1000000x harder
2) comms are **** (especially edges)
so basically what james said
I would like to do master kilominx BLD though (whatever the one is that is like a 4x4), because you can do real edge comms
emolover
Legend
What system do people use?
If I were to do a megaminx blind(long time from now) there would be 3 colored sections. Blue would represent the U F L and R faces. Red would represent the D Dfr Dfl and Dl face. Yellow would represent the Br Bl B and Dr faces. I would use a similar anti-spefz lettering scheme that I normally use. If a letter pair was D on the U face and K on the Br side I would memorise a green Donkey Kong. Green because the the primaries Blue and Yellow make green.
I feel as if that is not that hard. I am not saying I could actually do it because my blind ability is fairly poor but it is something I will be sure to try when I have the ability.
Similar threads
