Middlegame · 形勢判断 keisei handan
Who is better, and why.
Before choosing a plan, strong players ask who stands better. Japanese teaching gives the question a fixed shape, 形勢判断 (keisei handan): four factors, checked one by one. None decides alone, and the art is weighing them against each other.
Why count at all
The right plan depends on the answer. Ahead, you trade pieces and keep the position simple. Behind, you look for complications and play for the king. A player who never asks tends to attack when they should defend and to defend when they could have won.
1. Material (駒の損得 koma no sontoku)
Count each side’s pieces, on the board and in hand together, since in shogi a captured piece is only one drop away from fighting again. The usual order of value is:
rook ≥ bishop > gold ≥ silver > knight ≥ lance > pawn
Point values, and why no table is exact, are on piece values. Rules of thumb: a knight or a lance is worth about three or four pawns, and two pieces for one is usually a gain when the two include a gold or a silver.
2. Piece activity (駒の効率 koma no kōritsu)
A piece is only worth its value if it is doing something. The same count of pieces can hide a large difference in how much they reach. Look for:
- a rook on an open file, or a bishop on a long diagonal, against one boxed in by its own pawns;
- pieces that take part in the fight against pieces that never left their starting squares;
- promoted pieces, a dragon or a horse above all, deep in the other camp;
- pieces in hand, which can go wherever they are needed next move.
3. King safety (玉の堅さ gyoku no katasa)
Count the golds and silvers near each king, then look at the shape. With the same number of guards, an Anaguma is harder to break than a Mino, and a Mino harder than a boat castle. Ask also how easily the king can be checked: a king that can be checked from several directions is weaker than its guard count says.
Each castle and how it holds up is on the castles pages.
4. Whose move it is (手番 teban)
In the opening a tempo rarely matters. Near the end it can decide the game outright, because an attack that arrives one move earlier usually wins even from behind on material.
Counting who arrives first, move by move, is its own skill: speed counting.
How the weight shifts
- Opening: avoid losing material and finish your castle. Tempo is a minor factor.
- Middlegame: material and piece activity decide most plans; king safety decides whether an attack can be risked.
- Endgame: king safety and tempo take over. Material still counts, but catching the enemy king comes first.
Make it a habit
After each exchange, and whenever the opponent’s move surprises you, run through the four in order. Write the answer as a sentence, “I am a silver up but my king is thinner”, and let it choose your plan. The app’s review shows how the engine’s assessment moved through the game, which is a quick way to check your own.
Check your judgement.
Play a game against the bot, then open the review: the graph shows where the balance turned, and you can compare it with what you thought at the time.