2-Look PLL Explained

What two-look PLL does

After OLL, every last-layer sticker faces upward, but corners and edges may still sit over the wrong side centers. PLL means Permutation of the Last Layer: moving those oriented pieces into their final positions. Two-look PLL first fixes corner positions, then fixes edge positions. A final U-face adjustment aligns the solved layer.

The method reduces 21 one-look PLL cases to a small complete toolkit. A common set has two corner cases and four edge cases, although teaching systems can choose different algorithms or repeat one algorithm to cover a situation. Focus on the two-stage logic rather than arguing over a count: every oriented last layer should be solvable without improvising.

First look: decide how the corners must move

Ignore edges and compare each last-layer corner with the two side centers beneath it. Correctly placed corners have the right three-color identity even if neighboring edges are wrong. Look for headlights: two same-color corner stickers on one side. Adjacent-corner and diagonal-corner cases require different recognition and often different algorithms.

Turn only the U face until the case matches your reference. If one side has headlights, align them as instructed before executing the adjacent swap. If no useful pair exists, use the diagonal case. After the algorithm, pause and verify that all four corner color pairs match their centers. That check makes the edge stage much easier to read and catches a wrong setup before another algorithm compounds it.

Second look: read the edge cycle

With corners solved, inspect side bars and edge destinations. A solved bar on one side usually signals a U permutation, where the other three edges cycle clockwise or counterclockwise. No solved bar can indicate an H permutation, with opposite edge swaps, or a Z permutation, with adjacent edge swaps. Learn the relationship, not only the case name.

For U permutations, determine the cycle direction before turning. Follow one edge from its current side to its matching center, then see how the other two must follow. This is more dependable than guessing Ua versus Ub from an algorithm's visual silhouette. For H and Z, identify whether the swap relationship is opposite or adjacent, align the cube to your chosen algorithm, and execute from a familiar grip.

Choose a coherent beginner algorithm set

Use one trusted reference so setup angles and labels agree. Mixing algorithms from several tutorials can create conflicting recognition cues even when every sequence is valid. Favor algorithms built from familiar R, U, M, and trigger patterns that you can perform accurately. A sequence with slice moves is not automatically advanced; it may be the cleanest option if your M2 execution is comfortable.

Evaluate the whole action: pre-adjustment, algorithm, and final adjustment. A fast body with a slow setup is not fast. Learn where your thumbs begin, which finger performs each U turn, and how the last move leaves the cube for AUF. Repeat slowly until layers stay aligned. Only then shorten pauses and increase speed.

Learn corners and edges in separate blocks

Spend one session recognizing and solving only corner permutation. Create an oriented last layer, identify adjacent or diagonal corner relationships, execute, and check all corners. In the next session, start from solved corners and drill the four edge relationships. This isolation reduces the temptation to memorize six unrelated pictures at once.

After both blocks are accurate, combine them in full two-look PLL cases. Mix easy and difficult states so you cannot predict the next algorithm. Use spaced review: a few correct reps today, tomorrow, and several days later are more useful than fifty imitations in one sitting. Mark a case learned only when recognition and execution survive a random presentation.

Make the two looks flow together

During the corner algorithm's last moves, begin checking which side bars will remain. You may not fully identify the edge case yet, but you can reduce the next search. Finish with both hands in a grip that supports the likely M or U moves. The transition should develop gradually; accuracy remains more important than forcing zero pause.

Treat AUF as part of PLL timing. The puzzle is not solved until the top layer matches the side centers. Train the possible U, U', and U2 finishes instead of stopping mentally at the last algorithm move. In competition, the timer stop follows a completed puzzle, so a forgotten adjustment can turn a good algorithm into a costly mistake or unsolved result.

Fix recognition and execution as different problems

If you choose the wrong edge algorithm, do recognition-only drills: name bar, cycle, and required adjustment without moving. If the choice is correct but the cube locks, slow execution and examine finger assignments. If recall disappears under the timer, practice one mixed case after a normal solve rather than doing another isolated block. The failure stage determines the drill.

Avoid rotating the entire cube several times to hunt for a bar. Learn two-sided recognition first, then work toward reading from the front and right faces. For corner permutation, headlights may sit on a hidden side; infer their location from visible sticker relationships rather than turning the cube. Every removed rotation preserves grip and shortens the visual search.

When and how to learn full PLL

Move to full PLL when two-look PLL is automatic, F2L is reasonably stable, and you can support regular review. Full PLL uses 21 cases to solve permutation in one algorithm. Learn distinct, ergonomic cases first, then add visually related pairs. U, T, J, and A permutations are common early choices, but your easiest recognition groups may differ.

Keep the two-look route as a fallback for a new case. If you cannot identify the full PLL confidently in a solve, use the reliable two-stage solution and review afterward. Replace the fallback only when the new case is quicker end to end. An algorithm list is not finished because every line was viewed; it is finished when random cases are recognized and executed without destabilizing solves.

A focused practice session

Warm up with five slow PLLs. Do ten recognition-only corner cases, then ten edge cases. Execute each missed case three clean times, not twenty rushed times. Finish with an average of 12 normal solves and record every PLL pause, wrong case, and AUF error. That final block tests whether isolated practice transferred.

Review the log after several sessions. If one U-perm direction remains confused, pair those two cases until the visual cue is explicit. If M moves lock, reduce speed and adjust grip. If PLL is already consistent, stop overtraining it and move to the next phase. The purpose of two-look PLL is dependable coverage; once it provides that, use the saved attention elsewhere.