• 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!

Immune System (solving method+ potential 1LLL)

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
Immune System ((M)-CELL)

Ok, this is a set of methods with increasing difficulty that is based around the idea of an M-CELL. They are 2OP, T-CELL and B-CELL.

But first, what is an M-CELL? Simply, M-CELL stands for Missing Corner and Edge Last Layer. So it means that 1 edge and 1 corner in the D-face are not solved and will be solved at the same time as the last layer in a similar method to LS. The name also reflects how there is a "cell" of caged unsolved pieces in the D-layer. At the time of writing, CELL is solved using variants of TCLL and L5E

1) CELL looks cooler
2) CELL is more efficient (as it doesn't need any PLL, OLL, ELL etc afterwards)
3) In the most advanced form, CELL can be a 1LLL
4) It can be used in conjunction with any method equally well.
5) You can use CELL after solving less of the cube than is required by LS
6) They are more intuitive than most LS methods.

So what are the methods and how do they differ?
The simplest method is 2-OP which stands for 2 (double) Orient and Permute. This is very similar to a VOP 3x3x3 solve where first the L5C are oriented then permuted, followed by L5EOP (the same thing but with edges). Thus the steps are:
1) Solve F2L-2 (I will go into detail on the various ways in which this can be done later in the post)
2) Orient corners (perhaps MGLS?)
3) Permute corners
4) Orient edges
5) Permute edges
NOTE: to be honest, steps 2-5 are completely interchangeable so long as you first orient the orbit, then permute it (technically you can do this the other way round but the algs are much worse and the movecount is much higher)
This is essentially a 1-look variant of 2-OP and is thus the natural progression there are 96 algs for the corners but you can do the edges intuitively on average in <10 moves STM. The steps are as follows

1) build to F2L-2- with an M-CELL in the base (see below)
2) L5C (TCLL or similar)
3) L5E (straight)


This method is where you can create a 1LLL-recognise L5E, solve using L5C, solve L5E. This method will give an LL that is <24 moves long that can be executed all in one go with no pauses.
With this variation, the CELL does not have to be adjacent (as the others had to be- hence the name Twinned-CELL). While this variation is no more efficient than T-CELL when it come to the CELL stage, it does provide a significant advantage when it comes to the F2L as there are a lot more options for what block you can build/place. The downside of this is that the locations of the CELLs can be very awkward sometimes so this variation may not be for speedcubing but it can still give some crazy good efficiency and low movecounts.

The steps:
1) F2L-2 (but the cells don't have to be adjacent)
2) L5C
3) L5E (maybe combine the last 2 so this could be a 2 step method)
This is the section where I will go through the various ways to build to an F2L-2. First I will present the way to do it for 2OP and T-CELL then give my suggestions for modification to increase efficiency for B-CELL

1) solve 3 cross pieces and place on left
2) do the CE pairs like ZZF2L using the R-face to match up and then insert
3) last two pairs (one with a misoriented corners, insert)
Modifications for B-CELL: insert one of the corners for the F2P (first two pairs) of the case for that is much easier/better, otherwise, you can't really change much. It may also be worth learning VHLS/ZBLS to orient the edges if the opportunity presents itself although this is not exclusive to B-CELL and anyway, the orientation in L5E is still quite efficient and there is no guarentee that you can. In addition, the TCLL cases are quite ergonomic and just as fast, if not faster, than standard COLL and the ZBLS will not really reduce alg count
Essentially the same as CFOP style but with the EoLine first.
1) Eoline
2) ZZ FB
3) L2P (one misoriented)
Modification: see above, in addition, you could also only place one edge in the EoLine and use the other one as the CELL
1) 2x2x3
2) form one pair+insert+orient?
3) 2nd pair
Modification: perhaps a misoriented corner in the 2x2x3 or a missing edge in it. Also, see above.
1) Roux block on LD
2) form a 2x2x2 above it either UFL or UBL
3) insert CE

My F2L style that I feel goes quite well with the method. I can usually finish this in <20 moves, often ~15.
Modification for B-CELL: for step 2, the M-layer and the cap do not have to match up so long as both are on L. This is about 3-4 moves more efficient for me on average.
Do I need to explain this one?
1) 3 piece cross
2) edges
3) corners
4) last pair.
Modifications: see 1,2 and 3.
The most advanced and complicated way although it is also far more efficient than 1,2,3 and 5 and a bit more efficient that SS style. Unlike the other ones this style is only really more useful when using B-CELL so that is what i will present.
1) 3 heise blocks, not the Centre, Edge, Centre one
2) move to correct places and orient edges simultaneously.
This variant is rotationless (in contrast to the above detailed variant)
1) FB (Roux block
2) 2x2x2 pair in BDR using mainly U, R and Rw turns.
3) Last F2L pair (corner misoriented so still easy)
4) TCLL
5) L5E.
1) Ergonomic (mostly R, Rw, M and U turns)
2) Quickly eliminates hidden pieces
3) Low movecount
4) No abstract concepts like the EoLine (so is direct solving)
5) Good lookahead (links into point 2)
6) Not too high an alg count (<100- around the same as CFOP, perhaps less if you mirror some of the cases and exclude things like 3-cycles that can be done intuitively)
7) Fast turning is definately possible with the method
8) No rotations
9) There are many options for the various ways the F2L can be constructed as it is a relatively easy shape that does not have too many pieces getting in the way of others (such as in a full F2L)
1) New cubers may find it difficult to optimise block building (but this is the same with many intuitive things such as cross, F2L etc).
2) 100 algs may still be more than some people may want to learn (but it's not much for a 1LLL if your lookahead gets good enough and definately less than the other 1LLL -ZBLL
3) Where to find 3x3x3 TCLL algs (they may exist but this will definately not be a permantent problem as if this method gets enough interest I will generate the algs (although some others may want to find good ones before then).
4) Lack of cubers experienced at using the method (unavoidable with any new method really, hopefully this will change)
In summary, I believe this method has some potential to be fast in the current form and it still does have room for improvement so has a lot of potential for speed I think. It will be interesting to see how fast someone can get if they use this as their main method (I will learn the whole thing including algs at some point, I just want to actually get decently fast with the method I am currently using (roux) first before really exploring other methods such as SSC and T-CELL although I may learn the TCLL algs for Roux as it does have some application)
If you made it this far, congradulations! I would love to hear what you think of this style of solving and how you would modify it to be more effective. If TL;DR, this is like MGLS but always has an LL skip by using TCLL and L5E by using an altered form of LSish stuff.
 
Last edited:

Robert-Y

Member
Joined
Jan 21, 2009
Messages
3,289
Location
England
WCA
2009YAUR01
YouTube
Visit Channel
"So it means that 1 edge and 1 corner in the D-face are not solved and will be solved at the same time as the last layer in a similar method to LS."

What. So the E layer is complete but the D layer is missing 2 solved pieces?...
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
The F2L is missing two pairs right? Then how do you propose the TCELL cases? It would have much more algs.

the F2L is one missing cross piece and one misoriented corner in the d-layer. Do L5E followed by TCLL style L5C. As neither have an effect on the other, do it as a compound alg without pauses. It comes out to an average of ~24 moves, possibly less in some cases.

"So it means that 1 edge and 1 corner in the D-face are not solved and will be solved at the same time as the last layer in a similar method to LS."

What. So the E layer is complete but the D layer is missing 2 solved pieces?...
Basically yes. It's like an LS slot rotated 90 degrees.
 
Last edited:

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
what would be the equivalent of this approach if you those 2 unsolved pieces were in F2L instead of D?

Insert misoriented corner, solve the remaining edges while inserting, then do TCLL. I wouldn't do this because it has horrible lookahead compared to M-CELL where the lookahead is good enough to basically give a 1-LLL as each orbit does not have an effect the other during the CELL. In addition, there are multiple insertions which is quite inefficient. There are also less rotations in M-CELL as all L5Es are done from one orientation and the TCLLs will probably be in only one orientation for the most part as well (at least for the cases I've looked through).

One other thing I would like to add is that for T-CELL and B-CELL it is better to do the corners before the edges as that means you will not have to learn an extra 3 cases for each step to work around which corners are solved.
 
Last edited:

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
One other possible approach (although this is a bit different from the other style with straight TCLL):

1) F2L-2 (one e-slice piece missing as well)
2) Insert the last F2L piece and orient one LL corner.
3) L3C
4) L5E.

This is basically a lower alg alternative to the T-CELL I outlined in the OP although I still believe it is almost as efficient and the lookahead is still good (L3C can easily be recognised while inserting the LS and the L5E can also easily be recognised while L3C is being done as the L3C can be done using a corner 3-cycle. This method would require <10 algs while still allowing the potential for a no pauses LL.
 

2180161

Member
Joined
Oct 11, 2014
Messages
750
Location
Illinois
WCA
2014HEDR01
YouTube
Visit Channel
What about EO then the FB? It reduces the L5E alg count down so less algs, but something a little harder to master, but...

even less algs. EO Roux FB. 2x2 in BDR, Last F2L pair. COLL, EOL5E. not too sure about move-count tho...
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
What about EO then the FB? It reduces the L5E alg count down so less algs, but something a little harder to master, but...

even less algs. EO Roux FB. 2x2 in BDR, Last F2L pair. COLL, EOL5E. not too sure about move-count tho...

It's a nice idea but it has the same flaws as ZZ/Roux/ECL? - awful lookahead into FB and less fluid solving because you can only think in <R,U,L,D,M2> which is a pain and also increases movecount without giving much for tps.

Also, it doesn't remove too much in the way of algs (or decrease the movecount for them) as L5EOP gives ~10 moves but L5EP is about 6 which, when combined with the Eo is about 11-12 moves which is more than we started with.

Also, half the point of this method is to give a method that is relatively easy with no abstract concepts (in direct contrast with my last method SSC) while still giving a low movecount, good lookahead and potentially a method which can sub-20 (definately- around 2-3 tps), sub-15 (maybe- it's ~3-3.5) or even sub-10 (if you can get sub-10 you should be good enough to average ~40-50 moves, - 23 F2L-2 (which I can usually do at a reasonable speed), TCLL- <14, L5E- <10 total- 47 STM (and I'm not particularly good at block building) which is 4-5 tps which is not too hard considering it has a good lookahead.
 

2180161

Member
Joined
Oct 11, 2014
Messages
750
Location
Illinois
WCA
2014HEDR01
YouTube
Visit Channel
It's a nice idea but it has the same flaws as ZZ/Roux/ECL? - awful lookahead into FB and less fluid solving because you can only think in <R,U,L,D,M2> which is a pain and also increases movecount without giving much for tps.

Also, it doesn't remove too much in the way of algs (or decrease the movecount for them) as L5EOP gives ~10 moves but L5EP is about 6 which, when combined with the Eo is about 11-12 moves which is more than we started with.

Also, half the point of this method is to give a method that is relatively easy with no abstract concepts (in direct contrast with my last method SSC) while still giving a low movecount, good lookahead and potentially a method which can sub-20 (definately- around 2-3 tps), sub-15 (maybe- it's ~3-3.5) or even sub-10 (if you can get sub-10 you should be good enough to average ~40-50 moves, - 23 F2L-2 (which I can usually do at a reasonable speed), TCLL- <14, L5E- <10 total- 47 STM (and I'm not particularly good at block building) which is 4-5 tps which is not too hard considering it has a good lookahead.

ok, but what about inserting the last F2L edge with ZBLS, COLL, and L5EOP?
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
ok, but what about inserting the last F2L edge with ZBLS, COLL, and L5EOP?


1)CFOP style
1) solve 3 cross pieces and place on left
2) do the CE pairs like ZZF2L using the R-face to match up and then insert
3) last two pairs (one with a misoriented corners, insert)
Modifications for B-CELL: insert one of the corners for the F2P (first two pairs) of the case for that is much easier/better, otherwise, you can't really change much. It may also be worth learning VHLS/ZBLS to orient the edges if the opportunity presents itself although this is not exclusive to B-CELL and anyway, the orientation in L5E is still quite efficient and there is no guarentee that you can. In addition, the TCLL cases are quite ergonomic and just as fast, if not faster, than standard COLL and the ZBLS will not really reduce alg count

From the OP^^
 
Joined
May 27, 2015
Messages
64
WCA
2012BRAS01
I've been looking for a good method for me to start using ever since I decided CFOP was bad, and I want to switch to SSF2L+T-CELL, since I think it is a superb method, comparable to Roux. I have some questions though.

Firstly, I'm still confused on how you do the 2x2x2 block. Could you provide an example solve for that? Maybe I'm bad at it because I'm bad at Roux, idk.

Also, I'm confused to how T-CELL is a 1LLL. From what I can understand, you recognize the TCLL case, and the L5E case, execute the TCLL, execute L5E. In that case, do you use TCLL algorithms that don't affect the edges? Am I missing something?
 

shadowslice e

Member
Joined
Jun 16, 2015
Messages
2,923
Location
192.168. 0.1
YouTube
Visit Channel
I've been looking for a good method for me to start using ever since I decided CFOP was bad, and I want to switch to SSF2L+T-CELL, since I think it is a superb method, comparable to Roux. I have some questions though.

Firstly, I'm still confused on how you do the 2x2x2 block. Could you provide an example solve for that? Maybe I'm bad at it because I'm bad at Roux, idk.
For this, you really do whatever goes. You can do a 1x2x2 in DBR then insert the final edge, place the edges then insert the final pair. Really anything you want or can to solve the 2x2x2 block.
Also, I'm confused to how T-CELL is a 1LLL. From what I can understand, you recognize the TCLL case, and the L5E case, execute the TCLL, execute L5E. In that case, do you use TCLL algorithms that don't affect the edges? Am I missing something?

The TCLL case is essentially a L5C comm so do not affect the edges. So you recognise the corner case then the edge case in order to solve the LL and CELL.
 
Top