• 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!

The Rice Method

Joined
Jul 7, 2008
Messages
627
Location
Saskatchewan, Canada
YouTube
Visit Channel
Having tried CF methods, I agree, but...

...what if you could do them quickly, somehow? I mean, for the best people, the corners can be done in 2 seconds, and so can L6E. If all we have to do is solve 6 edges and the centers in 6 seconds (to get sub-10), maybe there's a way.


Realistically, you're not going to have a 2 second solve for the corners IF you want to guarantee decent recognition for step 2a. Considering that the 2x2 average world record is just over 2 seconds, I think you'd be lucky to have 3 seconds for the corners (for an elite 2x2 solver). Anyway, with that aside, I don't think it's impossible to get sub-10 with this method. It just requires more refined strategies for the F6E. With the experimenting I've done with CFL, I've actually had some decent solves, but I don't think the potential is anywhere near CFOP or Roux.

What I like about this method is that it gives elegant and short solutions for casual solving.
 

rice

Member
Joined
Mar 8, 2012
Messages
29
Inspection time plays a huge role here. I will make a list of which edges CLL messes up and will expand to EG if there is demand. Tracking also plays a huge role as well. I think sub-10 is possible if you can plan out the solve far enough that you can track pieces. One obstacle on the learning curve is being able to rapidly find the desired edge pair(s). If you aren't tracking, can you infer where your pair(s) are most likely to be in 2 glances max? The first glance should pick out edges in 3 adjacent faces (like U,F,L) and make a 2nd glance, if necessary, based on that. While this method is more involved, mentally, I believe it has potential. As with anything, practice makes perfect.

@qqwref simultaneously pairing and inserting the first 2 pairs sounds like it would make for many cases.
 

rice

Member
Joined
Mar 8, 2012
Messages
29
When first solving with misaligned layers, I would often solve two edge pairs only to find out that they were oriented the wrong way. A quick fix would be F B M2 F' B' or anything that is comfortable for you. The rice method is very flexible!
 

rice

Member
Joined
Mar 8, 2012
Messages
29
Welcome to the 2nd Annual TRIM (The RIce Method) Revival!
This past year has given way to a significant innovation in the method, yielding a much more interesting way to solve the cube, as well as paving the path for faster times. Below is an updated guide with the details.

Step One: Solve the corners
Nothing new here; use whatever 2x2 method you fancy.

Step Two: First Six Edges, ignoring position/orientation [See OP for how to pair edges]
Opposite colors are equivalent, meaning red = orange and blue = green for the purpose of pairing edges.
Make any two pairs and insert them in any orientation, ignoring the centers.
This means you could have orange and red edges in F and blue edges in B. Only pay attention to the E-layer edges in the following example. M U F B U' B

Then, pair and insert DL and DR with the correct position and orientation.

Step Three: LSE^2
Solve LSE like normal, ignoring the E slice. Familiar cases will look distorted, but solve them like normal. Often, the position of opposite edges will be swapped, ie UF <-> UB, but worry not. It will get dealt with shortly.

Align the corners and center, then do a z/z' cube rotation. Depending on the case, it may be advantageous to do an
additional x/x' cube rotation. One of the side colors is now the new 'U.' Solve LSE again. Some cases will appear to be distorted but solve them like normal.

End cases:
http://alg.cubing.net/?setup=R2_U2_R2_U2_R2_U2&alg=R2_U2_R2_U2_R2_U2

http://alg.cubing.net/?setup=xz_U_R2_U2_R2_U2_R2_U&alg=U_R2_U2_R2_U2_R2_U

http://alg.cubing.net/?setup=U2_R2_U2_R2_U2&alg=U2_R2_U2_R2_U2

Final Notes:
A faster way to insert pairs in FSE is x/x' U M2 U' x'/x (or some variation of that), rather than F/B M2 F'/B'.
This also allows for a glimpse at the D edges. In step three, adjusting to a new 'U' will not take long and is overshadowed by the much greater flexibility in edge pairing/insertion, as well as moving some of the bottleneck in looking for the E edges to the second LSE.

Example solves:
Solve 1

Solve 2
 
Last edited:

rice

Member
Joined
Mar 8, 2012
Messages
29
Quite often, you'll have 4 misoriented edges and/or DF is swpped with DB. I'm not familiar with commutators, though, so it might be easy.

edit: I guess you could use BH edges?
 
Last edited:

rice

Member
Joined
Mar 8, 2012
Messages
29
Growth does not occur linearly, but is characterized by jumps and leaps amidst lengthy plateaus. And so it was with this little method of mine. After picking up cubing again, I made a number of conceptual breakthroughs that removed the over-reliance on look-ahead and vastly increased the number of options one has for optimally solving each sub-step. While there are a number of similarities with the old way of doing things, this latest iteration of the rice method will look very different from before.

The post will include a brief synopsis of the steps, define some terms, followed by an in-depth explanation of each part, discuss the advantages and disadvantages of using the method, have some example solves and videos, then end with advanced material. So let's get to it.....

Synopsis
0. Don't get confused by the fact that we will start with blue as D, then end up with white as D.
1. Make a cross using middle layer edges with blue and green edge pairs on the bottom
2. Solve blue corners in the D layer, then green corners in the U layer
3. Insert green edge pair into U layer
4. AUF to make columns and rotate the cube so that white is bottom and yellow is top like normal
5. Insert a white edge pair into the DL/DR slot with the correct centers
6. LSE

Definitions
Edge pair: Two edges of the same primary color that have opposite secondary colors that are in the correct position relative to each other. For example, the blue/red edge and blue/orange edge is considered an edge pair when you can form a blue line with them, such as when they are in the UF and UB positions, respectively.

Corner pair: Two adjacent corners that are in correct position relative to each other. For example, the UFR and UFL corners. Discussed further in the last section.

Middle layer edges: Assuming the traditional "white is always down" philosophy, the middle layer edges would consist of the blue/red, blue/orange, green/red, and green/orange edges.

Layer: What you think of when referring to a given side/slice of the cube. Note that the centers won't be relevant until Step 5, so they can be ignored until then.

The 6 Step Program

Step 0: Essentially, we are solving for the middle layer edges first so we don't have to hunt for them mid-solve, building the corners around that, then making some adjustments so that we end up in the same place as what would be accomplished by block-building and CMLL in Roux.

Step 1: Let blue be the D color and green be the U color for now. Make a cross in the D layer with two edge pairs using the middle layer edges, one vertical and one horizontal. An example would be a blue/red+blue/orange edge pair in the DF/DB position and a green/red+green/orange edge pair in the DR/DL position. Note that you don't have to pair up these edges before putting them in place; they can just be put in one at a time. Pick the side with an easy cross to be your D layer.

Step 2: Now that we have a cross, we want to put the blue corners in the D layer. Make sure to place them correctly relative to the blue edge pair. Then, solve for the green corners. While the vast majority of 2x2 or CMLL algs will keep the cross or at most dislodge an edge that can easily be inserted, I would consider using a different alg if two or more edges are moved around.

Step 3: At this point, AUF so that doing an M2 move will insert the green edge pair of the cross into the correct position in the U layer. Line the columns up.

Step 4: Rotate the cube so that white is on the bottom and yellow is on the top.

Step 5: Using only U, D, and slice moves, make a white/blue+white/green edge pair and put it in the U layer. Then, use slice moves to line up the blue center with the green columns, AUF the edge pair, and do an M2 to insert the pair with the correct center.

Step 6: LSE

Advantages of using the rice method:
1. No algs needed, very intuitive
2. Allow for a range of solving styles, from rigid speedsolves to CN and FMC solves
3. Don't have to block-build

Disadvantages of using the rice method:
1. Requires knowing LSE
2. Colors can be confusing
3. Solving for the corners can eat up a lot of moves if you're not careful, but knowing algs will help.

Example solves:
Example 1: uses corner pairing
Example 2: also uses corner pairing
Example 3: uses block building to solve the first layer
Example 4: also uses block building


A poor attempt at a tutorial video

Intermediate
1. The cross and corners can be made with red and orange being the U and D layers. Even with a blue/green cross, you can have the D corners be green.
2. Inserting D corners one at a time can be inefficient, so making a corner pair and inserting that will speed things up. Lets say that we are solving for a blue DFR/DFL corner pair. If blue is point up, place the pair above where it needs to go and do xMF2M' to insert it. If blue is facing the right side, then either perform xrFM'FRF' or xMSR'FS'M' to insert the corner pair. Mirror this if blue is facing the left side.
3. When inserting the DL/DR edge pair, you can solve for a yellow edge pair instead of a white one and make yellow be your new D layer.

Advanced
4. Blockbuilding the D layer is very move-efficient, since you can build a "first block" and tack on a row to that. Another option is to build three rows; the sky is the limit.
5. Color-neutrality: One thing I've noticed is that I've been able to take advantage of situations where a number of 1st layer pieces are already in position or easy to solve for that would not be an option even for Roux solvers.

Here are some of the best solutions I've found that illustrate this:
Example 5
Example 6
Example 7
Example 8
Example 9
 
Last edited:

rice

Member
Joined
Mar 8, 2012
Messages
29
Another year, another post. It seems like every new method proposal comes with a corresponding set of 10's to 100's algs to learn, and you know what? It doesn't have to be that way. That is why I have always been thrilled by how far I can push the more intuitive, puzzle-like aspects of cubing every time I dust off the rice method and look at it with fresh eyes.

This latest, self-contained post brings with it an overall conceptual framework of maximizing freedom of available moves and minimizing "breaking/undoing things that had previously been built up in order to get to the next step." Contrast that with CFOP, which forces you to pick a color, solve the cross, forces you to solve F2L without breaking anything, then solving LL breaks things in the process of trying to solve everything at once anyway, which is why LL algs are so long. This is just one example, but the same general critique applies to other methods as well. Now, these aren't bad methods, just different ones, and a well-rounded cuber will absorb all of them and make them their own.

With that said, let's get to business. Select your main method:
  1. Ignore the centers and choose a D color
  2. Solve the four CE pairs. See PCMS for tips. Rouxers can blockbuild.
  3. Orient and permute U corners
  4. Solve DL and DR, along with their centers. Example
  5. LSE
  1. Ignore the centers and choose a D color
  2. Build the first block, but ignore LF, LB, and the center
  3. Build the second block, but ignore RF, RB, and the center. Check the CFOP Spoiler for another blockbuilding option.
  4. CMLL, or CMSLL if you know PCMS
  5. Solve UL and UR, then do x('). Note that this step can be done before CMLL. Just pair them up and place them in the DF and DB position, solve CMLL, then do (U) M2.
  6. Solve your new DL and DR, along with their centers. Example
  7. LSE
I'm sorry. Your method is just too restrictive. Pick another option and try again.
I mean, I can't really fault ZZ for EOLine because that's the whole point of ZZ. Nor can I fault Petrus for the 2x2x3 block because that's the whole point of Petrus. Like I said, there are no bad methods, just different ones.

One potential disadvantage, at least at first, is that the look-ahead for DL/DR/centers will be the primary roadblock. However, this can be mitigated with practice.

So, what are the advantages of using the rice method?
  1. No new algs to learn! And a complete beginner would only need to learn 2-look corner algs to get started.
  2. Good use of inspection time + good look ahead potential between each step + color neutrality = fast solves
 
Last edited:
Top