• 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 35,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

[Help Thread] The "Square-1 Help / Alg Sharing" thread

Joined
Feb 23, 2019
Messages
489
Location
The FitnessGram Pacer Test is a multi stage...
I managed to derive the niklas variant 1,0 / -3,6 / 3,0 / -3,0 which inserts DLF corner into ULB while flipping U/D so that you can build the first layer on U.

Then if necessary a last edge can be inserted from UR to BD with 1,0 / 3,0 / 3,0 / -1,-1 / -2,1 / -3,0
Hey,
Just wanted to let you know that there is an LBL of sorts. I used to use it. It basically involves building a layer, then permute corners then permute edges. However, a much better method is Lin. This is basically build a layer minus an edge, then insert the edge while permuting corners with an alg, and then EPLL. Helmer Ewert uses an advanced version of this method (CSP and solving the whole last layer while inserting the edge) and is currently tied for 6th in the world in average with a 7.56 average at Jonkoping side events 2019.

Hope this helped.
 

kadabrium

Member
Joined
Jun 13, 2019
Messages
65
Hey,
Just wanted to let you know that there is an LBL of sorts. I used to use it. It basically involves building a layer, then permute corners then permute edges. However, a much better method is Lin. This is basically build a layer minus an edge, then insert the edge while permuting corners with an alg, and then EPLL. Helmer Ewert uses an advanced version of this method (CSP and solving the whole last layer while inserting the edge) and is currently tied for 6th in the world in average with a 7.56 average at Jonkoping side events 2019.

Hope this helped.
Thanks for the reply

Lin does look like better way to build one layer. However is OLECP+EPLL better than simple OLE+PLL? I would think recognition for the latter is easier but not sure about execution.

compared to Vand, do lbl methods have a greater movecount? Vand CO only takes 2 moves, EO and CP around 4-5 each. building a layer takes at least one move per piece before the last edge or at least 6 in total, and then some another 6 moves for the edge itself.
 
Joined
Feb 23, 2019
Messages
489
Location
The FitnessGram Pacer Test is a multi stage...
Thanks for the reply

Lin does look like better way to build one layer. However is OLECP+EPLL better than simple OLE+PLL? I would think recognition for the latter is easier but not sure about execution.

compared to Vand, do lbl methods have a greater movecount? Vand CO only takes 2 moves, EO and CP around 4-5 each. building a layer takes at least one move per piece before the last edge or at least 6 in total, and then some another 6 moves for the edge itself.
the answer to your first question is, yes, it is. Particularly if you don’t use CSP, because parity PLL isn’t fun, while parity EPLL isn’t all that bad. Plus, as I said, there is PLL+1 where you solve OLE and LL at the same time with only like 72 algs w/o parity (at that point most would know CSP). You still should know full PLL as a world class solver in case you can insert both edge with an m2.

I can tell you that maximum for 1st layer minus 1 edge is 8 slices, and then PLL+ 1 looks to be around 10-12 slices on average with the algs I have (I have a doc)
 

kadabrium

Member
Joined
Jun 13, 2019
Messages
65
the answer to your first question is, yes, it is. Particularly if you don’t use CSP, because parity PLL isn’t fun, while parity EPLL isn’t all that bad. Plus, as I said, there is PLL+1 where you solve OLE and LL at the same time with only like 72 algs w/o parity (at that point most would know CSP). You still should know full PLL as a world class solver in case you can insert both edge with an m2.

I can tell you that maximum for 1st layer minus 1 edge is 8 slices, and then PLL+ 1 looks to be around 10-12 slices on average with the algs I have (I have a doc)
do you have all the lsll algs?
 

kadabrium

Member
Joined
Jun 13, 2019
Messages
65
Is 1-look l10p possible? there are 5 ol2e cases (1-1, 3-cycle, adj-opp, opp-opp, D-opp) each of them containing no more than 76 algs
 

Sue Doenim

Member
Joined
Nov 9, 2016
Messages
447
Is 1-look l10p possible? there are 5 ol2e cases (1-1, 3-cycle, adj-opp, opp-opp, D-opp) each of them containing no more than 76 algs
Assuming parity is correct, you'd have PLL (21), swapped PLL (22), PLL+1 (76), swapped PLL+1 (76), opp (76), and adj (76), so 271 algs if I did it right. I say that's pretty feasible, if someone is dedicated enough to square one, especially since with M2 you can easily change the case if you get a case you haven't learned yet.
 

kadabrium

Member
Joined
Jun 13, 2019
Messages
65
Assuming parity is correct, you'd have PLL (21), swapped PLL (22), PLL+1 (76), swapped PLL+1 (76), opp (76), and adj (76), so 271 algs if I did it right. I say that's pretty feasible, if someone is dedicated enough to square one, especially since with M2 you can easily change the case if you get a case you haven't learned yet.
how does one go about generating sq1 algs? subsets of particular interest to me are ol2e-cp (for a 2look method that can somewhat handle parity) and l6e (full roux stylistics)
 

Sue Doenim

Member
Joined
Nov 9, 2016
Messages
447
how does one go about generating sq1 algs? subsets of particular interest to me are ol2e-cp (for a 2look method that can somewhat handle parity) and l6e (full roux stylistics)
Jaap Scherphuis made an optimal solver you can find out about here, but it's not very user-friendly. I think maybe you could also use kSolve, but I don't know anything about it.
 

kadabrium

Member
Joined
Jun 13, 2019
Messages
65
Jaap Scherphuis made an optimal solver you can find out about here, but it's not very user-friendly. I think maybe you could also use kSolve, but I don't know anything about it.
generated OL2E-CP of the 3-cycle and Adj sets:
3-cycle(UF-DB-DF)
Front swap D3B2C1A54G8H7E6F-
4,0/2,11/0,3/3,0/3,0/9,0/0,3/1,10/8,0 [8|19]

Back swap A3C2B1D54G8H7E6F-
0,11/0,3/4,1/0,9/2,8/1,10/8,5/3,0/0,4 [8|21]

Left swap D3C2A1B54G8H7E6F-
1,0/5,2/3,0/9,0/1,10/5,5/0,3/0,9/0,4 [8|20]

Right swap C3D2B1A54G8H7E6F-
0,11/4,7/9,0/0,3/11,2/7,1/3,0/9,0/2,0 [8|20]

Diag swap D2C3B1A54G8H7E6F-
4,0/2,0/9,0/0,3/0,3/9,0/8,0/7,1/8,0 [8|18]

Adj(LB)
Front swap A5C7D4B31G8H2E6F-
1,0/3,11/3,0/3,3/3,0/2,0/9,0/0,9/0,7 [8|19]

Back swap B5D7C4A31G8H2E6F-
0,5/9,0/1,0/0,3/8,0/4,0/0,9/3,1/11,0 [8|18]

Left swap B5A7C3D41G8H2E6F-
0,8/1,4/5,2/1,4/2,8/3,0/0,7 [6|17]

Right swap A5B7D3C41G8H2E6F-
1,9/9,0/3,0/2,8/4,1/2,5/0,9/7,7/2,0 [8|22]

Diag swap B5A7D4C31G8H2E6F-
9,0/0,9/4,0/8,0/0,3/9,11/3,0/3,0/0,7 [8|18]

D-opp swap
Adj A1B4D2C37E6F5G8H-
0,8/3,0/0,9/3,0/4,7/2,8/3,10 [6|16]
Diag A4D2C1B37E6F5G8H-
1,9/3,0/9,0/0,9/5,11/10,1/9,0/0,3/8,3 [8|21]

Idk what kind of coincidence it is that all except one are 8 slices long. I just set up the cases minimalistically so it is almost certain there will be shorter algs that do different EPs elsewhere on the cube.

Edit: added the bottom swap set although the diag one may not be more efficient than simple insert + nperm. The opp-opp insert set looks to me as almost certainly not worth it.

we need a better name for this subset too. cdmll?
 
Last edited:
Top