We Played 300 Levels of Screw Sort: Here Are the 7 Hardest
300 Levels In, Seven Scenarios That Keep Coming Back
The screw-and-bolt sort format looks deceptively approachable in its first twenty levels. You have a board populated with colored bolts; each bolt can be removed from its post and placed on another post if the colors match or if the target post has space. The goal is to consolidate all bolts of each color onto dedicated posts. Simple enough — until the boards grow to twelve or fourteen posts with six or seven colors distributed across them in configurations designed to produce forced bad-state branching.
We played through 300 levels across SortFun's screw and bolt sort section and tracked the board patterns that stopped us, slowed us down, or required more than twenty moves to resolve. Seven configurations came up repeatedly, across different games and difficulty tiers. They are not random hard — they are structurally hard, meaning the same underlying constraint pattern produces the same type of difficulty every time it appears. Understanding what makes each one hard is more useful than memorizing solutions, because the same structure will appear in different visual arrangements across different levels and different games.
This post documents all seven. For each one, we describe the board configuration, explain the specific mechanism that makes it hard, and give the approach that breaks it open. If you want the foundational strategy concepts that underlie these approaches, our 7 techniques for screw sort covers the basics, and the constraint-propagation logic in our 3-step solving method applies directly to the screw format. If you want to understand why these puzzles feel so hard at a psychological level, our piece on the Zeigarnik effect covers the cognitive side.
Hard Configuration 1: The Single Receiver Lock
Board description: One color — typically the one with the most bolts on the board — has exactly one post with that color in the top position, meaning only one post can currently receive more of that color. All other posts holding that color have other colors on top, blocking the matching bolts underneath.
Why it is hard: the natural instinct is to start moving things around to expose the buried bolts. But every move that temporarily displaces another color to expose a buried bolt of the locked color uses a post or creates a new blockage elsewhere. Players who move reactively — "clear the top of post 4 so I can access the red bolts underneath" — typically create two new blockages for every one they resolve. The board gradually fills with partial-color posts that cannot receive anything, and the moveable pieces run out before the consolidation can complete.
The approach: before touching anything, count how many posts you will need to stage displaced colors during the unblocking sequence. If the locked color requires exposing bolts on three posts, and each post has two colors on top of the buried bolt, you need six staging moves before you get to consolidate a single segment of the locked color. Do you have six posts with available space? If not, the consolidation sequence will collapse partway through. Identify the staging capacity first. If it is insufficient, find a different color to consolidate first — one that frees up posts as it completes — and return to the locked color when the board has more breathing room.
Hard Configuration 2: The Mutual Block
Board description: Two colors are blocking each other. Color A's bolts are buried under Color B bolts on posts that Color A needs to consolidate. Simultaneously, Color B's bolts are buried under Color A bolts on posts that Color B needs. Moving A's surface bolts to free Color B's receiving posts requires a post that is currently occupied by B's surface bolts. The block is symmetric and complete.
Why it is hard: there is no locally obvious move. Every move that helps Color A harms Color B and vice versa. Players who scan for moves that help a single color will find none, which typically triggers either a random move (bad) or a decision to undo everything (expensive). The mutual block is the screw sort equivalent of a deadlocked traffic intersection — two streams of vehicles each waiting for the other to clear.
The approach: identify a third color that is not involved in the mutual block and has a clear consolidation path. Completing that third color's consolidation will free at least one post, which creates an asymmetry in the mutual block and breaks the deadlock. The general principle is: when two colors are blocking each other, the solution is almost never to resolve one of them directly. It is to find a third color that can be resolved independently, and use the freed post as the staging ground that breaks the symmetry. This is a specific instance of the backtracking step in our 3-step method: instead of solving the mutual block directly, identify the board state that would make it solvable (one free post) and work backwards to create it.
Hard Configuration 3: The High-Bolt-Count Pyramid
Board description: Boards with eight or more posts and seven or more colors, where one color has significantly more bolts than the others — say, eight bolts of blue when every other color has four. The "pyramid" color requires two complete posts to consolidate, but because it has more bolts in total, it also has more bolts buried under other colors and more colors blocking its receivers.
Why it is hard: the pyramid color dominates the board's constraint structure. Every move made for any other color has to account for whether it leaves the pyramid color's receiving posts accessible. Players who treat the pyramid color as just another color to consolidate eventually corner themselves: they consolidate three or four other colors cleanly, then discover the pyramid color's consolidation is now impossible because its remaining bolts are locked under posts that were just filled with the other completed colors.
The approach: identify the pyramid color in the first scan of the board. Treat it as the primary constraint throughout the entire solution, not as one color among many. Every sequence you consider — for any color — should be evaluated first for whether it preserves at least two accessible receiving posts for the pyramid color. This is the look-ahead step of the 3-step method applied with maximum priority to a single color. The pyramid color is the chokepoint. Everything else moves around it.
Hard Configuration 4: No-Reset Dead End (The Minimum-Undo Trap)
Board description: A board where the first twelve to fifteen moves look clearly correct — constraint propagation reveals several forced completions, the look-ahead analysis seems favorable — and then, around move fifteen, the board is suddenly deadlocked. No moves are available that do not create an immediate deadlock on the following move. And the undo budget (if the game implements one) is down to two or three moves, which is not enough to back out to the decision point that caused the problem.
Why it is hard: this is the trap that specifically punishes players who trust constraint propagation and look-ahead but skip backtracking. The forced completions were correct. The look-ahead analysis was probably correct for the first six to eight moves. The deadlock was created by a move in the nine-to-twelve range that seemed fine locally but created a global constraint that only became visible fifteen moves later. Without enough undo budget to backtrack to that move, the board requires a full reset.
The approach: on boards that appear to have a clear early sequence, deliberately pause after the forced completions (Step 1) and run a full board-wide constraint analysis (Step 2) before starting the exploratory moves. The deadlock that appears at move fifteen is always caused by a specific decision in the middle phase. The question to ask, mid-sequence, is: "Does this move permanently close off any color's consolidation path?" If the answer is yes — even hypothetically, two or three moves downstream — it is the wrong move. Finding the right move at that junction requires the backtracking target-state analysis from Step 3.
Hard Configuration 5: The Cascading Displacement Chain
Board description: A board where every post has bolts of at least three different colors, and no post has the same color in both the top and second position. This means every single move displaces a color that was not on top — creating a new surface color on the source post — which in turn may be a color that was not supposed to be on that post's surface, creating a new constraint problem with each move made.
Why it is hard: the board is in a state of maximum mixing. Every move creates at least one new problem while solving one. The displacement chain — move A creates problem B, solving B creates problem C — can run for ten or fifteen moves before the board either begins to consolidate or locks completely. Players who are not tracking the full displacement chain will lose their place in the sequence and start making reactive moves, which accelerates the lockdown.
The approach: write it down, literally, if the game allows a pause. Identify the full displacement chain for your planned next three moves: "If I move the bolt from post 2 to post 7, post 2's new top will be yellow. Yellow needs to go to post 4. Post 4's current top is green. Green needs to go to post 9. Post 9 has space." Three moves that each create a new surface but each resolve into an existing path. Tracking the chain explicitly prevents you from making the move that creates a surface color with no valid destination. The universal sorting tips guide covers displacement tracking as a general technique; in screw sort it is more demanding than in water sort because the bolt-by-bolt discreteness means every move is a single-unit displacement rather than a fluid pour.
Hard Configuration 6: The Forced Bad-State Sequence
Board description: A level where the optimal solution requires deliberately making the board look worse for five to eight moves in a row. The sequence that leads to a clear board passes through a midpoint where every post has mixed colors, no completions are visible, and the board objectively looks harder than when you started. Many players abandon the solution at this midpoint, convinced they made a mistake.
Why it is hard: it violates the intuitive heuristic that progress should look like progress. Most puzzle games — and most real-world problem-solving — generate visible signs of improvement as you move toward the solution. The forced bad-state sequence does not. The board gets worse for a sustained period, then suddenly resolves into a clean completion in three or four moves. Players who trust appearances over analysis will bail out of this sequence every time.
The approach: before starting a sequence that you identified via backtracking analysis — that is, a sequence you derived by working backwards from a target board state — commit to it fully. Do not evaluate each individual move on its apparent merit; evaluate it only on whether it is the next step in the sequence you derived. The sequence is correct if the backtracking analysis was correct. Each move that makes the board look worse is a move that is creating the conditions for the resolution. If you find yourself stuck mid-sequence and uncertain, the question is not "is this move good?" — it is "is this the move the sequence calls for?" If yes, make it. The bad-state sequence is the hardest thing to trust in screw sort, and the most important skill to develop for levels above difficulty 100.
Hard Configuration 7: The Last-Two-Colors Trap
Board description: You have successfully consolidated five of seven colors, and the board now has four posts — two holding the last two colors mixed together, and two empty. The remaining bolts of each color are distributed across the two mixed posts in a pattern where the only way to access color A's bolts requires moving color B's bolts to a post that already holds color B, and vice versa. The two empty posts are not enough to stage both displacement chains simultaneously.
Why it is hard: it feels like a logic error. You did everything right for five colors, and now you are trapped with two. The board looks nearly solved, but the final two colors are deadlocked in a configuration that seems to have no exit. This is the most demoralizing configuration in screw sort because it occurs late in a level, after significant investment, and at a point where resetting the entire level feels like wasted work.
The approach: the last-two-colors trap is almost always caused by a mistake made earlier in the level — specifically, by failing to leave the two-empty-post staging capacity in the right configuration when transitioning from the fifth to the sixth color. The correct fix is to undo back past the consolidation of color five, because the trap was set at that transition point. The two empty posts that feel like they should be sufficient are not sufficient for the displacement chain the final two colors require. You need three effective staging moves, and you only have two empty posts — which means one of your completed posts must serve as temporary staging, which means one of your earlier completions has to be partially undone. It is not a pleasant realization, but it is the correct one.
The broader lesson this configuration teaches is that the end of a screw sort level is not the time to relax the backtracking analysis. The trap that appears with two colors remaining was created five or six colors earlier. Players who run the look-ahead analysis only on the hard parts of a level and switch to reactive mode once it "looks easy" are setting themselves up for this exact scenario. Our full technique guide for screw sort covers the end-game staging problem specifically. For players who want to understand how these constraint patterns relate to other puzzle formats, the same structural issues appear in harder parking sort levels and in the late levels of games like Bubble Sorting Deluxe — the constraint logic is universal even when the physical metaphor changes.
After 300 levels, the most consistent finding is this: screw sort is not hard because the moves are complicated. It is hard because the consequences of moves are delayed. The mistake that causes the deadlock at move twenty was made at move eight, and it looked correct at the time. The discipline of the backtracking analysis — asking not "is this move good?" but "what board state does this move lead to in five moves?" — is what separates players who consistently clear hard levels from players who hit the same walls repeatedly. That discipline is trainable, and the seven configurations above are the specific scenarios where training it pays the most.
