The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →A puzzle is difficult when its rules, constraints, or solution path are hard for a particular solver to discover or reason through. A longer solution can add work, but move count alone does not tell you how hard a puzzle will feel. The number of plausible choices, hidden dependencies, planning demands, and knowledge a player brings all matter.
Why puzzle difficulty depends on the player
The same puzzle can be easy for an experienced solver and frustrating for a newcomer. Players differ in the patterns and rules they recognize, while puzzles differ in how clearly they communicate legal moves and useful deductions. A player may be stuck not because the solution requires unusually deep search, but because they have not yet understood what actions the puzzle permits.
Kenneth Kotovsky and Herbert A. Simon’s 1990 study of Chinese Ring problems illustrates the point: almost none of the participants solved the puzzle within two hours without being shown how to move it; even after a demonstration, only half solved it. The authors concluded that “Discovering how to make moves, rather than other features of the problem search space, was the source of its inordinate difficulty.” That result describes one puzzle and participant group, not a general rate for puzzle solving.
What can make one puzzle harder than another?
More plausible choices and dead ends
A puzzle can offer many possible next moves, only a few of which lead anywhere useful. When a solver must test alternatives, backtrack, or avoid deceptive dead ends, the puzzle may demand more effort even if its shortest solution is not especially long.
#1 Best Overall
In computer search, the branching factor is the number of successor states available from a state. It can vary across a puzzle’s search space, and the algorithm and heuristics used affect how much of that space is explored. This is a way to describe computational effort, not a direct measure of how difficult a human will find the puzzle. Korf, Reid, and Edelkamp’s analysis of heuristic search on sliding-tile puzzles and Rubik’s Cube, for example, describes heuristics as reducing effective search depth in their model rather than changing the asymptotic branching factor.
Rules or dependencies that are hard to see
Some puzzles make their legal moves and constraints obvious; others require players to infer how the pieces, clues, or rules interact. Hidden dependencies can make even a small-looking puzzle difficult because the solver must first discover what matters before deciding what to do.
Rank #2
Sudoku shows why clue count is not enough to predict difficulty. Mária Ercsey-Ravasz and Zoltán Toroczkai’s 2012 study proposed an escape-rate measure that correlated with human difficulty ratings for the puzzles they examined. Their analysis found that clue positions and constraint structure matter; a global measure such as the number of clues alone does not characterize hardness. Their scale’s reported mean values—η=0.816 for easy, 1.439 for medium, 1.782 for hard, and 1.809 for “absurd”—belong to that study’s sample and model, not a universal Sudoku rating system. Read the Sudoku study.
Planning subgoals and making counterintuitive moves
Some puzzles require a player to work toward an intermediate goal before they can make visible progress toward the final one. A move may appear to take the player away from the objective, or several subgoals may need to be coordinated mentally. These demands can make a short solution harder to plan than a longer but more direct one.
Recommended Free Tools
Rank #3
A 2004 Tower of London study manipulated goal-hierarchy ambiguity, the need to generate subgoals, and the presence of suboptimal alternatives in three-move problems. Those factors affected preplanning time, showing why shortest-solution length does not capture every planning demand. The study’s authors cautioned that judging difficulty solely by minimum moves is insufficient. See the Tower of London study record.
Knowledge the puzzle expects
A puzzle may rely on a rule, pattern, or inference technique that some players already know. What feels like a clever deduction to a novice may be routine to an expert. In a 2024 paper applying its methods to The Witness, Shen and Sturtevant define puzzle entropy in terms of the information needed to describe a solution, given a model of a player’s knowledge and inference rules. This is a formal framework, not a standard scale used across puzzle games. Read the puzzle-entropy paper.
Does a longer solution mean a harder puzzle?
Not necessarily. A longer shortest solution can mean more steps, but it does not say whether the next move is obvious, whether the rules are clear, or how much planning and backtracking are likely. Conversely, a puzzle with a very short solution can be difficult if the key move is hard to recognize or seems counterintuitive. The Tower of London findings on three-move problems demonstrate that planning structure can affect difficulty even when solution length is held short.
For computational puzzles, solution depth is one factor in search effort, alongside branching and the search method. A computer’s node count or runtime and a human’s time or perceived difficulty answer different questions. They should not be treated as interchangeable scores.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteBest Value
- 4 Activity Books Kit-Endless Fun: Our activity books for adult set includes 4 books with different color covers, each packed with 39 pages of puzzles. Over 150 pages total provide dozens of hours of engaging entertainment and a perfect escape from digital screens.
- 8-IN-1 Brain Games for Adults: Includes 8 popular puzzle types games: Word Search, Crossword, Sudoku, Maze, Dot to Dots, Spot the Difference, Find Hidden Objects, and Logic Games.Fully meet the needs of different interests, keep the brain active and challenging.
- Great Use Experience: The page size is moderate (7.5“ x 10.4”), with clear and easy-to-read text and graphics. The lay-flat binding and large, clear print make solving puzzles comfortable and enjoyable.
- For Stress Relief and Focus: Immerse yourself in a world of fun challenges. It helps you step away from electronic screens or work fatigue, improve focus and mindfulness, and achieve a state of calm relaxation.
- The Thoughtful Gift for Women and Teen: This thoughtfully designed activity book offers a variety of puzzle games suitable for any occasion and audience. Whether given to friends, family, or colleagues as a birthday or holiday gift (such as for Christmas, Thanksgiving, Halloween, or just to say "thank you") or as a thoughtful leisure gift for travel, retirement, or recovery periods, it makes a considerate and heartfelt present.
How to compare puzzle difficulty fairly
When comparing two puzzles, name the kind of difficulty you mean rather than relying on one rating or a move count:
- Solution depth: How many moves are in a shortest solution, if that is well-defined?
- Choice structure: How many plausible next moves are there, and how likely is dead-end exploration or backtracking?
- Constraint visibility: Are the rules and dependencies apparent, or must the solver infer them?
- Planning: Must the player create subgoals, coordinate several steps, or make a move that initially seems to move away from the goal?
- Knowledge assumptions: What patterns, inference rules, or genre knowledge does the puzzle expect?
- Who or what is being measured: Is the comparison about a human novice, an expert, or a particular search algorithm—and is the result time, effort, nodes, or runtime?
These are useful comparison axes, not a validated universal scoring system. Puzzle ratings are most meaningful when they identify the genre, intended solver, and measurement method.
What research can—and cannot—tell us
Different studies isolate different sources of difficulty: move discovery in Chinese Ring problems, planning structure in Tower of London tasks, constraint patterns in a set of Sudoku puzzles, or algorithmic search effort in formal models. Their findings explain mechanisms, but none establishes one authoritative scale that works for every puzzle or player.
For example, Adi Botea Eaton and Vadim Bulitko’s 2022 crossword-solving method used constraint propagation to reduce candidate choices in a Romanian competition crossword model. Some instances took an average of up to 1.2 minutes of CPU time, while a comparison timed out after four hours. Those are results from the paper’s model and experiments, not a general benchmark for crossword difficulty or human solving. Read the crossword-solving paper.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
