角块在 FR 槽位上方且白色朝右,棱块在 U 层已与角块配对。
CFOP · 第 2 步
F2L 算法
CFOP 方法的全部 41 种标准 F2L 情况。高效地将角-棱配对插入前-右槽位。
What F2L really is
In the beginner method you solve the first layer in two separate phases: place four corners, then thread in four middle-layer edges. F2L collapses that into one idea — you pair each first-layer corner with its matching middle edge and drop them into their slot as a single unit, four times. That one change is the biggest single time saving in the whole solve. Dan Harris calls F2L the stage where “most people gain a lot more time,” and it is the reason a CFOP solver finishes in 20 seconds where the beginner method takes 60.
Hold the cube, then read the case
Keep the cross on the bottom and build every pair in the top layer, inserting downward into the four side slots — these algorithms target the front-right (FR) slot. To recognise a case, read it in two steps: first find the corner (is its white/cross-coloured sticker on the U face, the R face, the F face, or is the corner already down in a slot?), then find its edge. That two-step read is exactly how the case families below are grouped, so it doubles as a recognition routine.
The anatomy of every F2L algorithm
Almost every F2L algorithm does the same job in the same order: a set-up move or two to position the pieces, the pair-up that joins the corner and edge, the insert that drops the joined pair into its slot, and sometimes a restore that puts back any pair the pair-up disturbed. Once you can split an algorithm into those phases, you can rebuild a case you've forgotten and stop breaking slots you've already solved. Watch the two below and you'll see the phases happen.
R lifts the slot pieces into the top layer and connects them with the waiting pair, U aligns the joined pair over its slot, and R' drops it home. There is no separate set-up or restore — a connected pair sitting in the right place is the gift case of F2L.
R U R' U' pairs the corner and edge up in the top layer; the trailing R U' R' connects and inserts the pair below the corner. Notice how the work splits cleanly into pair-up then insert — that is the template behind dozens of the cases below.
One trick, four angles
The most common myth about F2L is that filling all four slots means learning four times as many algorithms. It does not. Learn each case as a hand motion — how the pieces move relative to each other — rather than a fixed string of letters, and the same trick works from any side: you simply substitute the face that is in front of you. The front-right pair-up R' U R becomes F' U F, L' U L, or B' U B for the other three slots, with no new memorisation.
Look-ahead: the real speed lever
Getting faster is not mainly about turning your hands faster — Harris's warning is that “the hands are quicker than the eyes.” Turn at full speed and your perception can't keep up, so you stall after every algorithm to hunt for the next pair, and those dead stops cost more than turning a little slower would. The fix feels backwards: slow your turning down just enough that your eyes can track the pieces, let the current algorithm run on autopilot from muscle memory, and spend the freed attention finding and orienting the next pair. Chain those together and the whole solve runs with no pauses.
The genuinely tricky cases
Most cases are short. The long ones almost all involve a piece trapped in a slot the wrong way — a corner wedged in with its cross colour pointing sideways, or a pair already in the slot but flipped. These are long for a reason: the algorithm has to eject the badly-placed piece into the top layer before it can rebuild and re-insert the pair. You'll find them under the “Corner in slot,” “Both in slot,” and similar filters below.
Finger tricks for F2L
A finger trick is a short sequence ripped off in one fluid motion instead of re-gripping between turns, and F2L is built almost entirely from a handful of them. Your index finger and thumb alone can drive the whole trigger family — these are the same pairs of moves that pair up and insert, so finger-tricking them is fast F2L. The famous “sexy move” R U R' U' is just two of these stitched together.
All 41 cases
Here is the complete set. Now that you can read an algorithm's phases and adapt it to any slot, treat these as phrases you understand rather than strings to parrot. Filter by case family, copy any algorithm, and step through it in 3D.
情况 1 的镜像。配对已在 U 层形成,可从左侧插入。
配对已在 U 层对齐,准备好从左侧直接插入。
配对已在 U 层对齐,准备好从右侧直接插入。
角块白色朝上,棱块在侧面。标准分离算法。
情况 5 的镜像。思路相同,左侧。
角块白色在 F 面,棱块在 U 层。
情况 7 的镜像。
角块白色在 R 面,棱块在 U。使用左手设置。
角块白色在 R 面,棱块在 U 层后方位置。
角块白色在 F 面,棱块在 U 层后方位置。
较棘手的情况,需要三重 sledgehammer 设置。
角块白色贴纸朝上,棱块在 U 层。标准插入。
情况 13 的镜像。
角块朝上,棱块在 U 层左侧。
角块朝上,棱块在 U 层右侧。
角块朝上,棱块在 U 层后方。设置较长。
角块朝上,棱块在中层槽位(朝向错误)。
角块朝上,棱块在中层朝向错误。
情况 19 的镜像。
角块在 U 层,棱块已在中层(错误槽位或翻转)。
棱块在槽位且白色朝前,角块在上方。六步解。
棱块在槽位需要重新插入,角块在上方。
角块朝上,棱块在槽位。设置 + 插入。
较棘手的情况,使用左手握法。
角块已在 FR 槽位(朝向错误),棱块在 U 层。
情况 26 的镜像。
角块在槽位指向错误方向,棱块在 U。
情况 28 的镜像。
角块在槽位错误,棱块在 U。较长的恢复设置。
角块在槽位错误,棱块在 U 后方。三重设置情况。
两块都在各自槽位但朝向错误。长算法。
角块在槽位已朝向,棱块翻转。常见情况。
用 sexy move 重复图案翻转配对。
角块在槽位正确,棱块需要在相对槽位翻转。
配对在槽位但角块-棱块位置互换。
简单变体,角块正确而棱块翻转。
只有棱块在槽位(翻转)。角块在上方的 U 层。
只有角块在槽位(扭转)。棱块在 U 层。
配对在槽位但翻转(角-棱连接颠倒)。
配对在槽位但角块和棱块位置互换。
OLL 算法
F2L 之后,用 57 种 OLL 情况调整最后一层朝向
参考CFOP 入门
了解 OLL 如何融入完整的 CFOP 方法