# SpeedBLD

#### kid who cubes

##### Member
How is speed blind memorised?
Do you literally just sit there and try to one look a whole solve? Because one looking a whole solve seems impossible.
Are normally blind methods ever used and just revised more?
How do people one look an entire F2L solve?
Any tips?

#### Mike Hughey

##### Super Moderator
Staff member
Yes, you try to one look an entire solve. The way it's done is to track the pieces. So for instance, with the most basic and uneducated of methods using CFOP, first you one-look the cross. Then you pick an F2L pair to try to insert, and you track those two pieces as they would move as you make the cross. It's not too hard to track a couple of pieces as you mentally execute the cross. So then you know where that pair is, and can add the moves to your memorized solution, so that you have a solution for an extended cross. Now pick a second pair and do it again. And so on until the last layer, whereupon you do the same with the OLL and PLL. My first time trying it, it took about an hour to memorize.

#### Thom S.

##### Member
All of the SpeedBLD solves I have seen/done are in CFOP, so

Yes, you use your Speedsolving method. People who do 2x2, Skewb or Pyraminx have some advantage as one-looking their Puzzle is relatively easy.
My way of doing it was just going through and memorising your solution Move-by-Move, then track a piece while going through your solution again to see, where it ends up and repeating this some times. I found it fairly easy until the third pair.(I've done SpeedBLD once and I got Left Sune-U-Y Perm-U' as a Last Layer)

Edit: Mike Hughey and me wrote the basically the same thing at the same time. I feel very special now.

#### TDM

##### Member
Staff member
Although the site no longer works properly (iirc it used to have images as well as better formatting), Lucas Garron proposed "compound OLL", which is similar to 2-look OLL except you use different algorithms to help preserve pieces. This makes LL much easier. However, tracking pieces later on into F2L and OLL is still a challenge. Although you can use tricks like Compound OLL to help you, you will still need to do a lot of tracking with any method.

All of the SpeedBLD solves I have seen/done are in CFOP
Alex Lau used to hold the UWR for SpeedBLD with Roux.
E: looks like Waffo did an attempt too

#### Lucas Garron

##### Moderator
Staff member
Do you literally just sit there and try to one look a whole solve? Because one looking a whole solve seems impossible.
It kind of feels like that at first. But if you try to solve the cross and one slot, and then the next slot, and then the next... it's actually quite doable without learning anything new. In particular, CFOP keeps a lot of pieces steady during F2L, and moves the others in ways that are easy to track. (I don't know about other methods' I haven't tried those.)

I wrote a guide here back in 2006, although the applets no longer work.

#### Underwatercuber

##### Member
Although the site no longer works properly (iirc it used to have images as well as better formatting), Lucas Garron proposed "compound OLL", which is similar to 2-look OLL except you use different algorithms to help preserve pieces. This makes LL much easier. However, tracking pieces later on into F2L and OLL is still a challenge. Although you can use tricks like Compound OLL to help you, you will still need to do a lot of tracking with any method.

Alex Lau used to hold the UWR for SpeedBLD with Roux.
E: looks like Waffo did an attempt too
I think using pure OLL would probably be easier since you would just need to trace your pieces to where they would be for LL (for oll) but you could trace for pll at the same time.

##### Member
This is how I do it:

1. Figure out and memorize an EO cross
This will take more moves than a regular cross, but will make figuring out the rest of the solution a lot easier.

2. Track the 4 F2L slots
EO will help here since the amount of cases you'll run into will be reduced, no cube rotations will be necessary for the insertions, and you'll only need to track edges' positions without having to worry about their orientation.

3. Track the orientation of 3 of the corners for OLL
This is the hardest/most annoying part for me. What I do is always track the same 3 corners: UBL, UBR, and UFR (A, B, and C in Speffz). Once you figure out how those 3 will be oriented after F2L, you'll know the OLL case (the 4th corner's orientation can be deduced by the orientation of the first 3). I solve OLL by rotating the corners in place, so I won't have to worry about OLL affecting PLL. There's no need to do anything about the edges since you know they'll be oriented.

4. Track the position of 2 corners for PLL phase 1
In step 3, you track the orientation of 3 corners (A, B, and C), so for PLL you can track the position of just two of them (A, and B) and figure out where C and D end up after F2L (C will be the 3rd piece you tracked in step 3, D will be the remaining corner). In case you're not familiar with Speffz, let's label the corners of a solved top face like this:
A - B
D - C

Now, let's say after step 3 you know A, B, and C will end up in the positions marked with an X, the remaining corner will be D:
X - D
X - X

Then, in step 4 you figure out the positions for A and B:
B - D
X - A

Based on the information from step 3 you'll be able to deduce the position of the 4 corners:
B - D
C - A

5. Track the position of 3 edges for PLL phase 2
I always track UB, UR, and UF for this (A, B, and C in Speffz). It's a lot easier than the corners because once again, you don't have to worry about their orientation.

6. Figure out PLL and solve
Using Speffz, a solved top face will be any of the following:

A - A - B
D - X - B
D - C - C (Fully solved)

D - D - A
C - X - A
C - B - B (U' away from solved)

C - C - D
B - X - D
B - A - A (U2 away from solved)

B - B - C
A - X - C
A - D - D (U away from solved)

So let's say you get this after tracking the corners and edges:
B - B - D
A - X - C
C - D - A

Then you'll have to solve this very simple puzzle in your head to figure out which PLL it corresponds to. You can notice that the edges are solved relative to each other, so you'll only need to permute the corners to finish the solve. That B - B on top is a block that's already solved, meaning you only need to permute D, C and A (an A perm), to reach this position:
B - B - C
A - X - C
A - D - D

And then perform a single U turn to finish the solve.

Example solve:

Plz ignore the smug attitude at the end. It was a video response to a kid claiming he could memorize a cube and solve it blindfolded by tracing all the pieces as he solved it in his "mental cube". He posted a video of him staring at a cube for 45 min then solving it for 2 min as proof, and I told him I could replicate his video by just doing a standard SpeedBLD solve.

Scramble:
F2 U' B2 L2 D' L2 U R2 F2 L2 F' L D2 B L' U2 R F2 U R

Solution:
F' L U2 B D2 L2 D' L2 // EO Cross
U2 R U' R' // 1st pair
U' L R' U2 R L' // 2nd pair
L U L' U2 L U' L' // 3rd pair
U2 L' U2 L U2 L' U L // 4th pair
R U R' U R U2 R' L' U' L U' L' U2 L // OLL
y' l' U R' D2 R U' R' D2 R2 x' // PLL
U // AUF

alg.cubing.net