• 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!

TheMachanga

Member
Joined
Jun 26, 2009
Messages
1,441
Location
Chicago, Illinois
WCA
2010LECH01
YouTube
Visit Channel
I just tap the locations in order and it gets forced into temporary muscle memory. I think I'll label all 24 locations soon and then learn M2 shortly afterwords. As for my BLD times, honestly, I only timed my first DNFs and my very first success, other than that, I have yet to time myself.
Not a very efficient memo method. I recommend using the color method for the edges (make a story or scene out of the words) in Old Pochmann. Then for M2 you label the spots where the edges belong. That way you won't have to learn whole new labels. Remember you can use J perms along with T perms too.
 

Innocence

Member
Joined
Aug 22, 2009
Messages
567
Location
Australia, mate.
anytime you want to. M2 is so much faster than T perm edges and really does not take that much effort to get used to.

Actually, my first BLD success was with M2. :D

@TheMachanga @Tyjet66:

I disagree. From what I've heard, visual memo is one of the fastest memo methods around for single cubes. It just depends on how much your brain can handle. I personally do visual for corners and letter pairs for edges, as these two I believe to have the most potential for speed. Theoretically, the less "translation" to your memory, the faster, but you need to balance that with ease of use.
 

cmhardw

Premium Member
Joined
Apr 5, 2006
Messages
4,115
Location
Orlando, Florida
WCA
2003HARD01
YouTube
Visit Channel
I suppose I should just go ahead and put my foot in my mouth now while I'm ahead, but here goes at a couple replies.

Do "people" say that you can't get sub-60 with move optimal BH edges? A lot of people saying that this isn't the second best edge method, while speed optimal is the best? How many claims M2 is faster than pure BH-edges? And what would you have "proved" by doing sub-60?

I've never heard anyone say explicitly that BH edges (move optimal) can't be sub-60, but I sometimes get the "vibe" that it's "not advisable" to use move optimal over speed optimal, which I would tend to agree with if you're trying to fully maximize your speed for 3x3x3 BLD actually. I sort of wish this wasn't the case, but on the other hand it makes perfect sense. BH edges is intended as a 5x5x5 central edges method, so it makes sense that something else would be used for 3x3x3 (speed optimal, M2, etc. etc.)

As to what would be "proved" by achieving sub-60 using move optimal edges, I'll take a line from my first post:

I'm just being weird and trying to prove a strange and mostly irrelevant point that move optimal BH edges can be fast enough to get a sub-60 single, even if non-optimal algs are faster.

So I do acknowledge that in all likelihood this proves absolutely nothing of consequence. If nothing else it would just be interesting to me to know that it can be done. I think the best comparison would be that I'm trying to get sub-20 using a 4LLL instead of a 2LLL, just to see if it can be done.

I couldn't imagine M2 would be faster, but maybe that's just me. And please do not give responses only with "This cuber uses M2, and he's so good", because you could probably be sub-60 with any commonly used method.

This is a tough call, because it depends on whether we mean faster single or faster average, or faster turns per second etc. I think it's quite difficult to compare two methods, as probably all factors need to be taken into account. It's entirely possible that each method beats the other in some respects but not others. I think M2 is an extremely fast method by the way, and has already been proven to allow the solver to achieve incredibly fast times using it.

Anyway; My recommends is that you start with move optimal center safe edges, except given U-perms and M'U2MU2, and then expand to non-center safe, and then non-move optimal.

I know this is what I should do, and I suppose that I will have to start working on my edges at some point anyway. I'm trying to pay attention to using half slice-planes in my 3x3x3 solving, but I often do the supercube safe commutator instead just because it pops into my head immediately, and I tend to view 3x3x3 practice as practice for my 3x3x3 parts of the 5x5x5. I think I will try to take it in stages as you suggest though.

I haven't kept statistics, so this is purely anecdotal, but generally, it feels like most of my sub-1:10 solves are ones where I use a lot of BH edge cases. If I have to do many extra moves, I generally can't sub-1:10. Or at least, it feels that way.

Mike, to be honest I've only gotten maybe 3-4 successful sub-1:10's ever, they're not that common for me. I did use BH edges on my most recent two sub-1:10's, though they were probably slightly more turns than move optimal just because I was trying to go quickly and may have missed some of the stranger "wide setup into a XXXX" cases, or even an A9 or something.

So I guess, to sum it all up, my point here is that I don't really have a point. *Commencing with the putting of the foot into the mouth*
 
Last edited:

cmhardw

Premium Member
Joined
Apr 5, 2006
Messages
4,115
Location
Orlando, Florida
WCA
2003HARD01
YouTube
Visit Channel
Dude! How many journeys do you have?

I'm currently trying to learn one journey of 100 places (which is 10 groups of 10 places, based on the city where I live and places I go). Am considering using various "parts" of this when doing memo so that you get used to using all of it and don't over-use any.

Wow man, you must have a lot of journeys. Can you give a brief overview?

Double post ftw! I feel these two replies are different enough, and my first post was already really long.

Toastman, I think having your 100 location journey will certainly be very helpful to you, and will help a lot for you being able to do solves without having to repeat locations often. If you find that you do sometimes still get memorization interference, then just make more journeys.

As for my journeys, I recently made a table of contents of all my journey locations to help me better keep track of what I need for the number of solves I do per week. My journeys are cube specific. As an example I would never do a 5x5x5 solve using a 4x4x4 journey. The memorization simply wouldn't fit into the journey, there'd be too much information.

3x3x3: Enough locations for about 25 solves before I need to repeat, plus some spillover locations like Mike put it.
4x4x4: Enough for 11 solves before I need to repeat a location. This includes some spillover locations.
5x5x5: Enough for 6 solves before I need to repeat a journey (w/ spillover)
6x6x6: Enough for 1 solve, plus a number of spillover locations. I don't need all of the spillovers here, so I use them for 3x3x3 sometimes.
7x7x7: Enough for 1 solve, plus a couple spillover locations.

The exact number is 280 locations if you count how many are in each journey. If it's close to a competition I will rest enough locations to account for the cubes I will be solving, so I don't use those locations in practice. Boris Konrad I think has more than this. He told me that he memorized cards by putting only one image per location, and having 52 locations per journey. He never told me how many journeys he had, only that he had "a lot" of them.

My goal right now is to create two more 5x5x5 journeys (enough for 2 solves). I have a basic layout for a journey that will let me do 1 solve, but it's not set in stone yet. I'd like to be able to do at least 8 solves before having to reuse a location (enough to do the weekly BLD event and speed event done BLD without having to reuse any locations). This would let me truly rest journeys for at least a week for all cubes for every weekly competition, which would be ideal for me.

I'm definitely closer to Mike on the spectrum of how many locations I have prepared. Mondo is insane, I love the idea of how he memorizes, but I'm not sure how he pulls it off to be honest.
 

RyanReese09

Premium Member
Joined
May 16, 2010
Messages
4,033
Location
Whiteford, MD, USA
WCA
2010REES01
YouTube
Visit Channel
For single BLD solves, making images is too slow, so if I wanted to still use my letter pair list would sentences be the way to go? Really annoyed at the fact I can only do 3 BLD solves per day at the moment (3 rooms) (Not that I am doing 3 BLD solves per day at the moment :()
 

Cubenovice

Forever Slow
Joined
Feb 6, 2010
Messages
2,291
Location
Vlaams-Brabant (Belgium)
Sentences based on letter pairs work just fine for me.
Two sentences, one for edges, one for corners.
I memo edges first, then corners.
Then I solve corners first, this allows me to do the first few corners on visual memo.

I have not yet prepared any rooms / journey's. This will probably be the next step as I want to move to multi or 4x4x4 after the 3x3x3 get faster and more reliable.
 

Innocence

Member
Joined
Aug 22, 2009
Messages
567
Location
Australia, mate.
I think letters are faster.

Source: I'm fast at memo.

I said later in my post that letters and visual were probably potentially the two fastest, so I don't disagree there. But as visual is also very fast, I think that it's silly to change from it if you prefer it as a memo method.
 

toastman

Member
Joined
Dec 6, 2010
Messages
204
Location
Sydney, Australia
Here's an old Pochmann (or general 2-cycle method) question. In a solve, I see that I have one corner twisted. Let's say UFR needs a clockwise twist. Assume my buffer is UBL

To fix this, I could shoot the buffer to this corner, twice (different stickers each time).

Or, theoretically, I could leave it until the very end, do "F D F' D' F D F' D'" move the buffer to UFR (U2), do "D F D' F' D F D' F'" and do U2 and the cube should be solved.

My assumption is "If there is one unsolved corner at the end and it needs a clockwise twist, this means your buffer corner will need an anti-clockwork twist". This thinking is correct, yes? I just had it in a solve, and worked out this whole approch while blindfolded and with the clock ticking. I think I got it but messed up somewhere else in the solve (DNF)

Any better approaches/algorithms for doing this?
 

cmhardw

Premium Member
Joined
Apr 5, 2006
Messages
4,115
Location
Orlando, Florida
WCA
2003HARD01
YouTube
Visit Channel
My assumption is "If there is one unsolved corner at the end and it needs a clockwise twist, this means your buffer corner will need an anti-clockwork twist". This thinking is correct, yes?

Yes this always works, for any method. You can do the same for edges too. I always flip the permuted but disoriented pieces right at the beginning of the solve, using the buffer as one of the pieces if necessary. The buffer piece balances out the orientation in a sense. If you have two corners that both flip clockwise, use the buffer as a third corner that also flips the same direction. If you have 3 flipped edges, then use the buffer edge as a 4th flipped edge, etc.
 

cmhardw

Premium Member
Joined
Apr 5, 2006
Messages
4,115
Location
Orlando, Florida
WCA
2003HARD01
YouTube
Visit Channel
@Chris and Mike: what are your average movecounts for 3bld? And if you've reconstructed fast solves, what's your lowest known movecount?

I don't reconstruct my solves very often, so I don't have very much to compare to. I did try some solves today. I did 6 solves, the first 4 were DNFs and then these two successes. The second scramble was a ridiculous scramble with lots solved, so I suppose this is a good stand in for a low move count solve.

Solve #1
Scramble: R' B' U' B L R' U R F D L D R2 F' B2 L' D U R2 B D U' L U' R2
Time: 1:28.18

My execution:

orient the cube to my color scheme: z2

flip edges: R' y E' R E' R E' R2 E R E R E R2 y' R
flip corners: B' L U2 L' U' L U' L' R' U2 R U R' U R B

Speffz in use below

cycle edges:
A -> G -> D : [L E L', U']
A -> O -> P : [M', D' R' D]
A -> C -> H : y' [U2, R E2 R'] y
A -> B -> L : y' [L E' L', U'] y
A -> K -> N : F' [R U R', E'] F *This is not move optimal. If I were doing optimal moves I should have done: y [E' L' E, R'] y'

cycle corners:
A -> K -> W : I actually did: R [L' D2 L, U2] R' which is optimal with x [D2, L U2 L'] x'
A -> J -> N : I actually did R [U, R D R'] which cancels to an A9. This case is optimal with 8 moves. I probably would have done x' [R' D' R, U'] x

Fix parity:
D' R D R' y (Right handed J perm) y' R D' R' D

Total is 114 moves.

----------------------

Solve #2
Scramble: L' F L2 R2 F' L2 U2 F2 B' L' U D' L' R D2 L' D R F2 U R' D2 R' F2 R'
Time: 1:18.11

Warning: This is such a ridiculous scramble, but it was a fast time for me.

Speffz in use below

(This is in the exact order I executed)

orient the cube to my color scheme: z2

flip edges: U2 x' E' R E' R E' R2 E R E R E R2 x U2

cycle edges:
A -> J -> I : y [E R' E', L] y'
A -> K -> E : y M2 D M' U M D' M' U' M' y'
A -> B -> R : [U, R' E2 R]
A -> L -> M: y' M U L E' L' U' L E l' y (that's a small l' at the end, this is a wide setup into a drop and catch case)

twist corners: F' x2 L' U' L U' L' U2 L R U R' U R U2 R' x2 F

cycle corners:
A -> D -> V : x' [L' U' L, D2] x
A -> G -> C : [R' D R, U2]
A -> X -> M : y D' R2 D' L D R2 D' L' D2 *solved*

Total is 89 turns. Again, this is a ridiculous scramble, and I really doubt this is anywhere close to my average. This would probably represent closer to my lower limit on a solve to be honest.
 

Sakarie

Member
Joined
Oct 19, 2008
Messages
459
Location
Alingsås, Sweden
WCA
2008SKAR01
YouTube
Visit Channel
Any better approaches/algorithms for doing this?
Is there any worse?

An easy variation would be doing it with <RD> instead of <FD>. Dan Brown can teach you that, but it's practically the same thing as FD. Or you can get all of the corners in F and use (UR'U'R)x2.

A probably faster way is to setup into headlights or chameleont and use 2xSune R U R' U R U2 R' + L' U' L U' L' U2 L or the inverse of that one.

I recently wrote down what I think is the fastest algorithms for orienting any corner with UBR, and if there's an interest I could upload it?
 

Mike Hughey

Administrator
Staff member
Joined
Jun 7, 2007
Messages
11,303
Location
Indianapolis
WCA
2007HUGH01
SS Competition Results
YouTube
Visit Channel
Solve #1
Total is 114 moves.
I got 121. Your edge flip and corner twist were a little more efficient, and your parity was significantly better than mine.

Solve #2
Total is 89 turns.
I got 95. You twisted corners and flipped edges more efficiently than me, and you saved a few moves with pure BH edges over me.

This would probably represent closer to my lower limit on a solve to be honest.
I bet it's not. You had flipped edges and twisted corners. Add a couple of 8-move 3-cycles and remove the twisted corners and flipped edges, and you easily go sub-80, with no solved pieces.
 

aronpm

Member
Joined
Sep 9, 2009
Messages
2,010
I got 123 and 85 moves for those scrambles.

Try this scramble: D' B2 R2 U B2 R2 U B2 R2 B2 D F D L2 U2 R D2 U F2 U F'
Corners:
x' RUR' D' RU'R' D x (8/8)
D2 L'U2L D2 L'U2L (8/16)
y2z' U2 R'U2R' D RU2R' D' R2 U2 zy2 (11/27)

Edges:
R U'M2U R' U'M2U (8/8)
L2 UM2U' L' UM2U' L' (9/17)
M'U2MU2 (4/21)
UR'U M' U'RU M U2 (9/30)
R2 UM'U2MU R2 (7/37)
uM'u2M'u (5/42)

Moves: 69

I've got a bunch of example solves written down somewhere, and the average of their move count is about 90-95.
 

Mike Hughey

Administrator
Staff member
Joined
Jun 7, 2007
Messages
11,303
Location
Indianapolis
WCA
2007HUGH01
SS Competition Results
YouTube
Visit Channel
I got 123 and 85 moves for those scrambles.

Try this scramble: D' B2 R2 U B2 R2 U B2 R2 B2 D F D L2 U2 R D2 U F2 U F'

I got 82 on that one. The edges weren't all that good for me; corners were almost the same as you.

Edit: When I tried it using your orientation, I got 78 moves.

Looks like you're at least 10 moves better than me on average. It's probably due to a few bad edge algs I use, flips and twists, and parity.
 
Last edited:
Top