Contents
Back to Main Page

OCLL/PLL: This orients the last layer corners in one
step (OCLL), then permutes the last layer corners and edges
simultaneously in the final step (PLL). This is similar to
Fridrich's OLL/PLL last layer, but much fewer OLL algorithms are
required since the last layer edges are already oriented. OCLL requires
a minimum of 6 algorithms and PLL requires a minimum of 14, giving a
total of 20 algorithms for both steps.
The average move count is 7.93 for OCLL and
11.21 for PLL which gives a total of 19.14 moves
average.^{[10]}
Using partial corner control during insertion of the last
1x1x2 block eliminates the H and Pi OCLL cases, allowing even fewer
moves and faster algs during OCLL or COLL.
Algorithms: speedsolving.com wiki: OCLL
 speedsolving.com wiki: PLL
 AlgDB.Net PLL
 speedsolving.com: Partial Corner Control

COLL/EPLL (ZZVH):
This orients and permutes the last layer corners (without affecting edge
orientation) in one step, then permutes the last layer edges in the
final step. This
may be preferred since it has a lower move count than OCLL/PLL,
and is regarded
by some to have easier case recognition. Learning COLL/EPLL
is also a useful intermediate step
to learning ZZLL or ZBLL. COLL has 40 cases to recognise, solvable by 25 algs min.
EPLL is just 4 cases (and only 3 algs if Uperm is mirrored)  a total of 28 for the
whole LL. COLL is 9.78 moves average, and EPLL is 8.75 average,
yielding a move count
of 18.53 (slightly fewer than
OCLL/PLL).^{[10]}
Algorithms: COLL on the Algorithm Database
 Lars Vandenbergh's COLL Page
 Bob Burton's COLL Page
 speedcubing.com: COLL Page

OCELL/CPLL:
This orients the last layer corners, while permuting the last
layer edges (without affecting edge orientation) in one step,
then permutes the last layer corners in the
final step. The advantage of using this method is that the majority
of algorithms (all of OCELL) can be done 2gen (turning only two sides).
This may give it an advantage over COLL in OH (onehanded) cubing.
Like COLL there are 40 cases to recognise, solvable by 25 algs min.
Again CPLL is just 4 cases, 2/3^{rds} of which are Aperm.
The remaining cases are EPerm, HPerm and a 1 in 12 chance of CPLLskip.
The average move count for OCELL with 2gen algs is 12.10, and for CPLL it's
9.17. In total that gives 22.01 moves on average, including
AUF.
^{[10]}
Algorithms: speedsolving.com: OCELL 2gen algorithms and move count statistics

ZBLL (ZZa):
Often labelled as the holy grail of speed cubing, this method completes
the last layer by orienting the corners and permuting the corners and
edges, all in one step. It involves 494 distinct cases, and requires
learning a minimum of 177 algorithms assuming
mirrors and inverses are applied.
Completing the last layer has an average move count of
12.08, which is a significant advantage over the 2look
options.^{[10]}
Algorithms: ZBLL algs on SpeedSolving.com Wiki (complete set)
 Bernard Helmstetter's ZBLL algs
 Chris Hardwick's ZBLL Page
 Jason Baum's ZBLL Page
 Lars Petrus's ZBLL Page (warning: many java applets!)
F2LL
During the final stages of F2L it's possible to manipulate the last layer cubies to reduce the number of LL cases. This partial solving of LL during F2L is called F2LL. The following are examples of F2LL options:

ZZLL (ZZb or Phasing): Involves permuting two opposite LL edges
during insertion of the final 1x1x2 F2L block. It is a relatively lightweight
technique, using an estimated average of ~2 moves.
It reduces the number of LL cases from 494 to only 167,
enabling completion in one step with knowledge of 80 algorithms.
Phasing also serves as a useful transition between COLL/EPLL and full ZBLL.
Further info: Michal Hordecki's ZZLL algorithms
 Michal Hordecki's Phasing algorithms
 James Stuber's ZZLL algorithms
 speedsolving.com: Phasing Explained

Winter Variation (WV): This orients the LL corners during insertion of the final 1x1x2 F2L block,
reducing the last layer to just PLL.
It involves 27 cases and is said to have a lower average move count than OCLL/PLL.
Algorithms: Sarah Strong's WV Algs
 Winter Variation on Cyotheking
 Sebastien Felix's WV Page
 Winter Variation Article

ZZCT: The first step (TSLE) orients all corners during insertion of the final F2L edge. During this step the final F2L corner is not placed, and may end up in the top layer. TSLE has 108 cases, all solvable with 2gen algorithms. The second step (TTLL) is similar to PLL, except the final F2L corner is also being placed. It has 72 cases, and
is said to have recognition and execution similar to PLL.
Further Info:

MGLS: On the last F2L slot only the edge is placed. The CLS part of the MGLS
method then orients the LL corners, while simultaneously placing the final Dface corner.
This leaves only PLL to complete the cube. There are a total of 105 cases to recognise,
solvable by a minimum of 56 algorithms, in an average of ~9.5 moves. It's possible to
reduce the number of CLS algs required if the Dface corner is also placed along with its edge,
but not necessarily oriented. This results in a subset of MGLS cases called EJF2L, solvable
with a total of 16 fast 2gen algorithms.
Algorithms:

Blah's Method: This deorients the LL corners during insertion of the final 1x1x2 F2L block,
allowing completion of LL in onestep using a reduced subset of ZBLL.^{[7]}
Back to Main Page
Last updated: 7^{th} August 2016