How to Compose an Original Tsume Shogi Problem
**Tsume shogi** (詰将棋) refers to composed shogi checkmate puzzles. Designing an original tsume shogi is a creative process that experienced Japanese composers have written about in guides and interviews. This article compiles their advice on planning a puzzle, methods of construction (backward vs forward design), and thorough testing to ensure a sound puzzle. We will explain technical terms in English so that even advanced shogi enthusiasts who don't read Japanese can follow along. Accuracy is crucial in tsume-shogi composition, so every tip here is backed by expert sources, ensuring you learn reliable methods without any unsupported claims.
Planning the Puzzle
Before placing pieces on the board, it's important to plan what kind of puzzle you want to create. Top tsume-shogi composers emphasize starting with a clear theme or key idea for your problem. This "theme" might be a favorite tactical motif (手筋) — for example, a clever sacrifice (捨駒) or an unexpected piece drop — that you find interesting and want to showcase. Once you have a motif in mind, think about how to present it in the most striking way: How many moves will the puzzle be? Will the king be chased to the corner or wander in the center? Which move in the sequence should feature the thematic tactic? These are questions composers ask when planning a puzzle. For instance, if you intend to compose a 7-move tsume (詰将棋７手詰), there will be four checking moves by the attacker, of which up to three could be sacrifices — you would decide at which point(s) in those moves to insert your highlight tactic. You should also consider the geometry and aesthetics of the position: do you want the puzzle to take place near the edge of the board (上辺) or involve a full-board chase (入玉型)? Sometimes simply rotating the board 90° or shifting the location of the scenario can make the idea clearer. Composers often experiment with different placements and orientations to find the most elegant setup for their idea.
Another aspect of planning is deciding on the puzzle's style and difficulty. Tsume-shogi authors each have their own style depending on what they "find fun" in a puzzle. Some prioritize a puzzle-like intrigue with many clever tricks and decoy moves, while others live for setting up brilliant foreshadowing (伏線) that only pays off at the end. There are those who love extreme difficulty and complex variations, and others who derive joy from achieving a beautiful initial position or satisfying certain composition conditions. There is no single "right" style — the world of tsume-shogi is richer when a variety of themes and styles coexist. So, choose a concept that genuinely excites you, whether it's a crafty unblock (移動合) or a visually pleasing arrangement, and let that guide your design.
Once you have a theme, a practical tip is to start small and gather ideas. Many Japanese experts advise composing lots of very short problems (3-move or 5-move tsume) as exercises. Try to create a tiny problem where one major piece is used or one clever sacrifice is the focus, and use as few pieces as possible to accomplish the mate. The goal at this stage isn't to produce a publishable masterpiece, but to practice efficiently realizing a tactical idea on the board without extraneous pieces or moves. These miniature compositions serve a dual purpose: they improve your "composition muscle," and they also become a library of useful motifs or building blocks you might later incorporate into larger works. In fact, an experienced composer notes that short "practice compositions" can become the seeds for adding a particular theme into a longer problem or form the basis of a good endgame (収束) sequence.
It's also wise to keep a notebook or idea log for your tsume-shogi musings. Write down any intriguing move sequences or tactical concepts that cross your mind — even if they seem impossible to implement at first. Japanese composers sometimes refer to these as "mousou tejun" (妄想手順), meaning "fantasy sequences." This is essentially a dreamed-up solution you'd love to see in a puzzle. Don't worry yet about whether it's legal or how to make it work; just record the idea. The technical work of realization can come later — what matters in the planning stage is expanding your creative vision. By nurturing a collection of such ideas, you'll have a source of inspiration to draw from when you sit down to compose.
One more planning tip from veteran creators: be prepared to produce many attempts, because only a few will turn out to be truly great. Tamiya Katsuya, a respected tsume-shogi composer, proposed the "3/100 theory," which says that in every 100 problems you compose, perhaps 3 might be really good. This isn't meant to discourage you, but rather to highlight that even for experts, creating a masterpiece is hard work and often a numbers game. Don't be afraid to make "practice" compositions that might not be perfect — each one teaches you something. In the words of one author, facing failures is part of the process: you'll be "fighting against alternate solutions, no-solution issues, and overly long variations" as you chisel the intended solution into the board. The key is to learn from each attempt and treasure even your first simple creation. It might not win a prize, but it's still your own work and a step toward the next idea.
Backward vs. Forward Design
When it comes to actually constructing the puzzle, tsume-shogi composers generally follow one of two approaches: the backward design method (逆算式 gyakusan-shiki) or the forward design method (正算式 seisan-shiki). The choice affects how you proceed with placing pieces and moves:
Backward Design (「逆算式」)
Backward design means you start from the end — the final checkmating position — and work in reverse to build the initial puzzle. In other words, you decide on the ideal mate position (tsume-agari 詰上がり) first and then add moves before it to construct a longer solution. This method is popular because it gives the composer a clear goal to aim for. As 8-dan composer Itō Ka explains, when using the backward approach "the final mating portion is fixed, so the composition has a certain stability". Knowing exactly how the problem must end can make it easier to ensure everything leading up to it is consistent and necessary. Many experts recommend backward design to beginners because it's systematic and reduces guesswork.
How do you actually create a tsume-shogi by backward design? Experienced composers outline a step-by-step process:
- Start from a simple mate position. Imagine or set up the board with the checkmating configuration you want (for example, the king trapped and checkmated by a certain piece). This could be as small as a one-move mate (詰みまで1手) position or whatever final tableau you envision. It's often easiest to begin with a very short mate. For instance, you might place a king and a few pieces such that ▲Kx is mate in 1 move. This will be the "target" that your full problem must achieve at the end.
- Work backward to add moves. Once you have the final position, think about what position could precede it — in other words, remove the last move. To do this, ask: "What was the move just before mate, and what did the board look like then?" In a one-move mate, for example, one move earlier the checkmating piece wouldn't yet be on that final square. So you create a position where the mate can be delivered in one move. This gives you a 3-move tsume (since now the mate will take one move from that position). Now continue to extend: from that new position, again ask what came *before*. Each iteration usually adds two ply (one full move by the attacker and one response by the defender), effectively lengthening the solution by two moves. At each step, you are "thinking of the position before the current one," which is the reverse of normal shogi thinking. Repeat this process to grow the problem to the desired length, adding pieces or adjusting the king's position as needed.
- Iterate and adjust when problems occur. If at any point your backward extension "doesn't work out," be ready to revise the plan. For example, perhaps the move you added allows the king an unexpected escape route or introduces an unintended solution. In that case, you might step back and try a different preceding position or add an extra piece to fix the issue. Backward composition is often an iterative dance: you add a move, then verify if the solution is still unique and forced; if not, you adjust and try again. Don't hesitate to undo a step and explore a different idea for that stage. The goal is to ensure the sequence remains tight and correct as you lengthen it.
By following these steps, you gradually "reverse-engineer" the puzzle. A simple example: Suppose you start with a basic 3-move mate as your final sequence. To make a 5-move puzzle, you need to add two moves before that. One common technique is to lure the enemy king or a key defender out of position by sacrificing a piece early. In fact, sacrificing a piece to force the king to move is "the first basic procedure to consider" when reverse-composing. For instance, if the king in your final position is on 2☗3 (using shogi notation), you might place the king initially on 1☗2 and plan a sacrificial check that compels the king to step forward to 2☗3, where your final mating moves then unfold. Moving the king by a sacrifical lure is a classic way to backward extend, and it also tends to make the puzzle more interesting because the attacker appears to waste a piece (a moment of "disadvantage") to achieve a greater aim. Japanese composers even have a term for this impression — furi-kan (不利感), the sense that the attacker's move seems disadvantageous. Incorporating a momentary self-harming move often adds depth to a tsume-shogi. In short, when adding moves behind your final mate, think about what sacrifice or forcing move could logically precede the mate.
As you extend the sequence backward, continuously check for any issues. A typical challenge is that new escape routes for the king appear when you add earlier moves. For example, one composer describes adding a rook drop as the new first move in a problem, only to realize the king now had extra escape squares on 1☗5 and 3☗5 that weren't guarded. The solution was to place an enemy pawn at 1☗5 and a gold at 3☗5 to block those escapes. In another case, adding a move created an unintended alternate mate: the attacker could drop the rook at a different square (2☗2 or 2☗1 instead of the intended 2☗3) and still eventually mate. This kind of alternate solution (called a "余詰め" — more on this later) is unacceptable, so the composer sought to eliminate it. One idea was to place a pawn at 2☗1 from the start, which would block both of those improper rook-drop options. While that would have solved the immediate problem, the composer rejected it for aesthetic reasons — simply adding a pawn there felt against the puzzle's concept. Instead, he reconsidered the prior position and devised a different setup (switching to an alternate "B diagram") that avoided the flaw altogether. This illustrates an important point: you have to balance practical fixes with the overall elegance of the puzzle. Sometimes a straightforward fix (like plugging a square with a pawn) might make the composition less elegant or violate some self-imposed theme or condition, so a more creative adjustment (or even reworking that section of the puzzle) can be preferable.
Throughout the backward design process, remember that it's okay if a certain added idea doesn't pan out. The experts advise flexibility: if you "get stuck, go back one step and consider a different position" instead of stubbornly forcing a flawed idea. In practice, backward composition often involves a lot of trial and error — adding a piece here, removing another there — until the moves flow correctly and all side variations are under control. In the end, however, many composers find this method rewarding. Because you knew the final mating picture from the start, you can focus on polishing the route to get there without worrying that the solution might fall apart completely. "Since the final shape is decided, you can confidently work it into a good composition," Itō Ka notes about the gyakusan approach. It provides a sense of security that the puzzle has a solid backbone, and you're mainly decorating or extending that backbone with interesting play.
Forward Design (「正算式」)
Forward design takes the opposite approach: you start with an initial configuration and try to invent a mating sequence *forward* from that. In this method, nothing is fixed at the end — you freely place pieces on the board, almost as if setting up a position from a real game, and then you search for a mate in that position. The term "seisan-shiki" (正算式) isn't a standard dictionary word; it was coined in tsume-shogi circles as the antonym of gyakusan (backward). Forward composition is a more open-ended, experimental process. Its biggest advantage is freedom: because you haven't predetermined the conclusion, you can explore novel patterns and let your creative instincts run wild. Itō Ka points out that with forward design, the "ending portion is free, giving you a wide breadth" to work with. You might stumble upon a unique idea that wouldn't emerge if you were constrained by a fixed final move. However, this freedom comes at a cost: there's less inherent "stability," and you may have to contend with many unforeseen solution paths or even the possibility that the position doesn't mate at all until you tweak it.
In practice, forward composing often starts with a burst of intuition or imagination. One way is to scatter some pieces on the board in an interesting way and see if you can make them into a mating problem. For example, you might decide "I want a puzzle that looks like a realistic endgame" (実戦型) and place the enemy king with a typical endgame configuration of defenders, then add your attacking pieces by gut feeling. Itō Ka recounted doing exactly this as a teenager: he "casually placed pieces one after another" aiming for a setup with a knight and lance on the board edges (like a natural game position) and a few attacking pieces positioned by instinct. He set the king's defenders in what "felt like a real endgame situation," and placed his attacking dragon, bishop, and pawn in a way that "felt right" to him. At this stage, you may or may not know if the position is actually a mate — you are essentially composing *and* solving at the same time, working out the moves as you go. In Itō's case, after he stepped back to analyze the setup he'd created, he realized the position was *almost* a mate but "agonizingly not quite solved" — it failed by a narrow margin. Specifically, he discovered that if he only had one more knight in hand, the puzzle would be solveable. So he went ahead and granted himself that extra knight (changing the set of pieces in hand), and incredibly, with that one change the position became a completely sound mate — no alternate solutions, no shorter solution, nothing missing. He had unintentionally produced a correct 25-move tsume-shogi! He calls this the "miracle of forward composition," marveling that by a stroke of luck the puzzle had no flaws. This anecdote shows both the excitement and the uncertainty of forward design: you might conjure up a brilliant idea by chance, but you'll also need a bit of luck (or a lot of testing) to get a flawless result.
Another anecdote from Itō highlights the almost whimsical nature of forward composing. He once literally created a puzzle in a dream and solved it in his sleep — upon waking, he rushed to his magnetic board to set the position he saw in the dream. In the dream position, the king was at 1☗3, an attacking bishop was at 4☗2, a pawn at 2☗5, and so on. Initially this dream puzzle didn't work (it was "extremely simple" and not complete). But after waking, he added a pawn at 1☗5 in the position (which hadn't appeared in the dream), and that single addition made the whole sequence come together as a valid, sound mate. He refers to this as another "miracle" of seisan-shiki — sometimes a puzzle will just fall into place forward in a way that feels magical. Of course, not every forward composition will be a miraculous instant success. More often, you'll need to adjust and tweak repeatedly: add a piece here to plug a hole, remove a piece there if it's unnecessary, and so on, until the solution emerges clearly.
So what's a practical approach to forward design? One Japanese source suggests that forward composition may suit those with strong solving ability, because you must find a mating sequence in whatever position you create. Here are some tips for using the forward method effectively:
- Start with an interesting setup: Place the king and other pieces in a configuration that appeals to you or that suggests a particular conflict. You might base it on an intriguing position from an actual game's endgame, or just an arrangement you find challenging. The key is that something about the position should hint at a mating idea, even if you haven't fully worked out the moves yet.
- Identify a potential mating sequence (solve your position): Once the pieces are down, treat it like a tsume problem that someone else composed — try to solve it. If you find a mate, great! If you're close but not there, identify what's missing. Is the king escaping somewhere? Is a key sacrifice or piece drop missing? Use your solving skills to pinpoint the flaw.
- Adjust the pieces to fix flaws: Now take the insights from solving and modify the position. If the king had an escape, add a blocking piece or move a piece to cover that square. If you needed an extra piece to deliver mate (as Itō needed a knight), give yourself that piece in hand. If some defender was too strong, remove or relocate it. This step may loop multiple times: each change you make, test solve again to see if the mate is now possible and if any new alternate solutions have appeared.
- Be open to new ideas as you refine: Forward composition can be unpredictable — you might discover a new theme or better idea while adjusting the position. For example, in one forward example a composer initially had a direct 5-move solution, but by looking at it from a "different perspective," he rearranged the pieces to create a more impressive 5-move solution with a sacrificial twist. The forward process allows these creative shifts, so embrace them if they improve the puzzle.
Forward design can be more free-form and might feel chaotic compared to backward design. You don't have a predetermined road map; instead, you're creating the road as you travel. It requires a lot of verification because it's easy to overlook a hidden solution path when you haven't "locked down" the ending from the start. Japanese experts note that with seisan-shiki you often encounter mysterious instances of success or failure — puzzles that suddenly work perfectly, and others that stubbornly resist being sound. This is part of the fun: "Forward style is interesting because you can encounter these strange miracles," writes Itō. If you enjoy a more exploratory, intuition-driven creation process, forward design is a great way to compose. Just be aware that you will need to rigorously test the final product, since the lack of an initially fixed solution means there could be more unintended possibilities lurking. We'll discuss testing next.
Testing and Refinement
After constructing a tsume-shogi problem (by either method), you enter a critical phase: testing and refining to ensure the puzzle is sound. A sound tsume-shogi has **exactly one** solution route: the attacker can force mate in the given number of moves and there are no alternate ways to achieve mate. In composition terms, a correct puzzle must have no "y余詰め" (yōzume), which means no unintended solution other than the author's intended sequence. It should also ideally avoid any "早詰" (hayazume), an unintended mate in fewer moves than the intended solution. (In fact, a hayazume is just a particular type of alternate solution — a shorter one — and is especially undesirable.) A puzzle with multiple solutions or a quicker solution is considered flawed, and composers label such a problem a failure. The goal is to achieve a "complete composition" (完全作), meaning a puzzle with no alternate or shorter solutions — only the single intended mating line. Ensuring this requires thorough checking of every possible variation.
The first step is to **verify that your intended solution (作意手順) actually works** as you think. Play through the moves from the start: does the king get checkmated on the last move? Are all pieces behaving as expected? Sometimes in composing, especially via forward design, an oversight can lead to a position that doesn't actually mate the king at all — this is called a "不詰" (fuzume, no-mate) situation. Obviously, a tsume-shogi must have a solution, so fix any no-mate issues by adjusting the position or moves. For example, if the king can run to an escape square that you didn't account for, you may need to place a piece to cover that square or restructure the moves to seal that escape. In one backward-designed example, a composer found that after adding a certain move, the king wouldn't take a bait because it could flee to an open square (making the sequence unsolvable). He solved it by adding a blocking piece on that square, turning the position back into a forced mate. The lesson is to double-check that for every move by the attacker, the king (or defending pieces) has no choice but the responses you anticipate.
Next, you must systematically **hunt for alternate solutions (余詰)**. Remember, in tsume-shogi the solver is obligated to find the shortest mate — any deviation that results in a mate in more moves, or using fewer/different pieces, is considered an alternate solution even if it's longer. This is a bit different from chess problems: in tsume-shogi, a longer mate or one that leaves unused pieces is still a composition flaw because the solver isn't required to follow the composer's intended line once a mate is found. Therefore, you must check **every possible check and move** that the attacker could try to see if it accidentally leads to mate. Common alternate solution scenarios include:
- Using a different piece or square for a drop: Perhaps you intended the first move to be a drop at 2☗3, but could the solver drop that same piece (or another available piece) on a different square to still force mate? In our earlier example, dropping the rook at 2☗2 or 2☗1 also led to mate in a longer sequence. That was an unintended solution, so it had to be eliminated. To fix such issues, composers often add a piece that covers those alternative drop squares or make the alternate sequence fail. In the example, adding a pawn at 2☗1 would have prevented both the 2☗1 and 2☗2 drops by occupying one and blocking the line to the other. Ultimately, the composer chose a different fix (for aesthetic reasons), but the principle stands: identify alternate drop locations and block them with cleverly placed pieces or pawns.
- Alternate order of moves: Check if the attacker can rearrange the move order to succeed. Maybe your solution is A then B to mate, but could they do B then A? If the sequence still works out of order, that's a problem. You might need to introduce a subtle point (like making one move only work after the other has been played) or add a piece that makes the wrong order fail. Sometimes the concept of "move order" issues overlaps with needing to force certain responses; a good composition will have a tight logic that forbids moves out of the intended order.
- Extra piece not needed: Consider whether all pieces you provided are truly necessary. If the attacker can mate without using a particular piece, that means that piece was extraneous and its omission doesn't increase the move count — essentially an unintended solution that "leaves a piece unused." In Japanese this is described as the solver "ending up with a piece still in hand" (持駒を余らせる) in the solution, and it counts as a yōzume (alternate solution) as well. For example, if you gave a gold and a silver in hand but it turns out the gold was never needed to mate, the solver had an unnecessary resource. The puzzle should be reworked so that every piece plays a role. Often, removing the unused piece from the initial setup is the simple fix — unless that piece was put there to prevent something else. In that case, you must find a different way to prevent the alternate without leaving an idle piece.
- Shorter mate (早詰): The most severe alternate solution is one that mates faster than your intended solution. For instance, you intended a 7-move mate, but there's a hidden 5-move mate. Because tsume-shogi rules demand the shortest mate, the 5-move line becomes the solution and your composed 7-move line is invalidated. Early mates typically arise from a defender being in a slightly wrong spot or an attacker's piece having an unexpected direct route. Fixing a hayazume might involve adding a defender to guard the quick route or altering a piece's position so the quick mate no longer works. In Itō's forward-story, he was delighted that after adding the knight, the composition had "no hayazume" — implying he carefully checked that no faster mate existed. Always verify move counts: if a mate can be done in fewer moves, you must either lengthen the intended solution or remove the shortcut.
When you find any alternate solution, address it methodically. Japanese composers often talk about the "fight against yōzume" as a battle that continues throughout the composition process. Each time you plug one hole, a different leak might spring elsewhere, especially in complex compositions. For example, one author added a Lance on the board to block an alternate, but that piece's presence inadvertently allowed a different 5-move mate to appear. The fix to that was to add yet another piece (an enemy gold at 4☗1) to stop the new alternate. This illustrates that fixing alternates can sometimes be a chain reaction. Be patient and persistent; iteratively refine the position until no move by the attacker except the intended ones will result in a mate.
In the refinement stage, also pay attention to **"dead pieces" or unnecessary pieces** on the board. A well-composed tsume-shogi has a sense of unity; ideally every piece on the board or in hand contributes either to the mate or to preventing alternates. If you have a piece that is only sitting there with no involvement, testers will notice and it will "leave a bad impression". For instance, if you placed a knight just to block a certain escape but that knight is never captured or moved in the main solution, it's considered a blemish. One expert after finalizing a composition admitted that a Silver he placed at 6☗6 was "too powerful" and had to be moved away to avoid an alternate, but in the final position that silver ended up doing nothing except existing as a blocker. He noted this is not ideal, saying the piece being left over is aesthetically poor. The advice is to either find a way to utilize that piece in the actual solution or remove it by redesigning that part of the puzzle. Sometimes you can incorporate a blocking piece into the solution by, for example, making it a piece that gets sacrificed or captured as part of the main line, rather than a dormant wall. Other times, you might replace an always-there blocker with a piece that is dropped during the solution (i.e. a move by the attacker), which is more elegant. As one source puts it, "rather than a blocker that is placed from the start, a blocker that appears mid-problem is better". For example, instead of initially placing a pawn to clog a line, maybe the attacker can drop that pawn during the solution — achieving the same block but as an active move. This kind of refinement elevates the quality of the puzzle.
Finally, after you've manually checked everything you can think of, it's highly recommended to **have the puzzle tested by others or by a computer program**. Many modern composers use shogi-solving software (tsume-solving engines) to run an exhaustive search on their composition for any yōzume or hayazume. These engines can confirm if multiple solutions exist. For example, one composer, after thoroughly checking himself, ran his nearly finished problem through a solver and was shocked to discover a hidden 25-move alternate solution the engine found. This was something "even the composer was surprised by". It turned out the Lance he had placed at 3☗3 in the initial position was "not good" because it allowed that long alternate; moving that Lance to 3☗2 eliminated the alternate. This kind of deep alternate might be impossible to find by human effort alone, so engine assistance is invaluable for final verification. Once the engine (or skilled test-solvers) give you the green light — reporting "no alternate solutions" — you can confidently consider the puzzle complete.
To summarize the key points of refinement:
- **Ensure uniqueness:** Check that only your intended solution works — no other mating line (shorter or longer) should succeed.
- **Guard against alternate moves:** For every attacking move, consider if a different move or drop could also mate. Add or adjust pieces to close those possibilities.
- **Eliminate shorter solutions:** Confirm the mate cannot be done in fewer moves than intended (no hayazume). If it can, restructure to increase the minimum moves.
- **Maximize piece efficiency:** Remove or repurpose any piece that doesn't contribute to the solution or a necessary defensive obstacle. Avoid "junk" pieces lingering unused.
- **Block escapes and unwanted defenses:** Make sure the king has no escape squares at every step, and that any defensive interpositions (合駒) other than those intended fail. If an enemy piece can nullify your idea by blocking differently, adjust the position to force a unique defense. For instance, if a defender could drop a piece to stop your mate in more than one way, you might need to arrange the position such that only one specific block works and all others lead to immediate failure (thereby guiding the solver along the intended path).
- **Test thoroughly:** Use other solvers — human or computer — to double-check for any solutions you missed. It's easy to become "blind" to alternate solutions in your own composition, so a fresh perspective or an engine's brute-force search is extremely helpful.
Through careful testing and refinement, you gradually sculpt your creation into a polished tsume-shogi problem. The process can be time-consuming — top composers might spend weeks or months refining a single problem — but it is highly satisfying when you end up with a puzzle that is sound, original, and elegant. As one author quipped, some creators will even "keep revising a 7-move puzzle for over 10 years" to get it just right (an extreme case, to be sure!). While you might not take a decade per puzzle, the underlying message is to value quality over speed: iterate until you have full confidence in the result.
In composition, there are also subjective judgments about what makes the final product "good." For example, how rich should the false trails (紛れ) and side variations be? Some puzzles are very clean, with almost a single line and minimal complication, while others intentionally include deceptive moves and subtle variations to make solving more interesting. As a composer, you can "thicken or thin out the flavor" of these variations as you see fit. Adding a few plausible-looking but wrong moves for the solver to consider (without actually leading to mate) can increase the puzzle's challenge. On the other hand, too much complexity might frustrate solvers. Strive for a balance that matches the level of complexity you want your puzzle to have. The refinement stage is where you tweak this balance by possibly adding or removing pieces to adjust the number of misleading possibilities. Ultimately, a "good tsume-shogi" often exhibits a clear core idea (the theme you started with) executed through a logically sound sequence, with no unintended solutions and no superfluous elements. If you achieve that, you've done well.
Conclusion: Composing tsume shogi is both an art and a science. Japanese tsume-shogi masters recommend starting with a strong idea, choosing a method (backward for structure or forward for creativity) that suits your style, and then refining meticulously to ensure a unique, satisfying solution. It's normal to face challenges like dead pieces or unexpected alternate solutions along the way; even experts battle those, employing various tricks to overcome them. By applying the principles outlined above — working methodically, checking all possibilities, and learning from the wisdom of past composers — you can craft tsume-shogi problems that are not only sound and solvable, but also capture the imagination of those who attempt to solve them. Happy composing!
**Sources:** The guidance above is based on original Japanese-language writings by veteran tsume-shogi composers and analysts, including Itō Ka 8-dan's series on tsume-shogi creation, the Kaze Midori note "How to Make Tsume Shogi", commentary by composer Komai Mei on reverse-construction techniques, and a tsume-shogi creation lecture from a shogi class blog illustrating backward design step-by-step, among other sources. All Japanese terms have been explained in context. This compilation presents verified methods and tips endorsed by those experts, translated and summarized for English readers.