Chess Piece Values and How to Think About Trades

Most chess books for beginners give you the piece values early on. A pawn is worth 1 point, a knight and a bishop 3, a rook 5 and a queen 9. The king has no value, because it can never be captured.

These numbers are not rules. They are a quick way to judge whether a trade wins or loses material. Used well, they prevent a large share of beginner losses. Used badly, they can make you cling to material in a position you could have won. In this post I cover where the numbers come from, how to count a trade, the positions where the numbers mislead, and why checkmate puzzles throw pieces away so freely.

Where the numbers come from

According to the Wikipedia article Chess piece relative value, attempts to put numbers on the pieces go back to the Modenese school in 18th-century Italy. In the 19th century Howard Staunton gave the queen 10 points, so 9 has never been the only answer.

A more recent example is Larry Kaufman, who estimated piece values from a large collection of games. In his article reprinted on Chess.com, he found that a lone bishop and a lone knight are worth about the same, and that a pair of bishops is worth roughly half a pawn more than two single bishops.

I take two things from this. The 1, 3, 3, 5, 9 scale is a fairly reliable average built on long experience. And it is only an average, which shifts with the position.

Count a trade by adding up

The values help most when you count an exchange. The method is simple: the points you gain minus the points you lose.

Say White’s knight takes a black pawn, and Black takes the knight back with another pawn. White gained 1 and lost 3, so White is 2 points down. I think this is the most common loss in beginner games. You see a capture and take it before checking whether the piece can be taken back.

Trading a bishop for a knight is 3 for 3, an even trade. Giving a rook for a knight or a bishop is 5 for 3, a loss of 2. This trade has a name, “the exchange”, and the side that gets the rook is said to be “up the exchange”. The Wikipedia article notes that most theorists put the real difference closer to one and a half points than two. You can compare other trades the same way, such as two minor pieces for a rook and a pawn.

Count before you capture

To add things up, you first need to know how far the captures will go. This is the order I suggest.

1. Count the defenders. How many of your opponent’s pieces protect the piece you want to take? If none, it is free.
2. Count your attackers. How many of your pieces hit that square? If you have more attackers than they have defenders, your piece is the one left standing at the end.
3. Capture with the cheapest piece first. When both sides pile onto one square, you normally take with a pawn before a knight. Take first with the queen and a pawn recapture costs you a lot.
4. Add up the points at the end of the sequence. If either side can stop partway, add up that version too.

It will feel slow at first. After a few dozen games most of it becomes counting by eye.

When the numbers mislead

The values are averages, so some positions break them. These are worth knowing early.

A pawn close to promotion. A pawn on the seventh rank with nothing to stop it is not worth 1 point. If it becomes a queen next move, it carries almost the weight of a queen.

Pieces that cannot move. A bishop blocked by its own pawns, or a knight left on the edge of the board, does less than its 3 points suggest. A piece’s value depends heavily on how many squares it controls right now.

The bishop pair. As in Kaufman’s estimate, two bishops together are a little stronger than the sum. When you give one up, it is worth being a little more careful even if the points are equal.

The king in the endgame. The king has no point value, but once the pieces come off it becomes a strong fighting piece that walks up the board. The Wikipedia article puts its fighting value at about 4 points.

Material that cannot mate. More points do not always mean a win. If your opponent has only a king left and you have only a king and one knight, you cannot checkmate. When trading in the endgame, ask whether the pieces you keep can actually finish the game.

Mate puzzles give pieces away

If you start checkmate puzzles right after learning the values, they can be confusing. The solution often hands over the queen for nothing.

The reason is simple: checkmate ends the game. When a mate is there, the points no longer matter. The smothered mate is the classic case. You sacrifice the queen, the opponent’s rook has to take it and fills the last free square, and a knight gives mate. I described it briefly in Chess Checkmate Puzzles for Beginners.

So the habit of counting points and the habit of looking for mate only work together. For most of a game you judge by material. When the enemy king has few squares, you set the points aside and look for mate. I believe learning to make that switch is one of the most important steps after the beginner stage.

Compared with shogi

Shogi also has the idea of gaining and losing material, but the situation is different. In shogi a captured piece goes into your hand and can be dropped back onto the board, so losing a piece also means giving one to your opponent. The swings are larger than in chess.

As far as I know, the only official point count in shogi is for “jishogi”, the impasse that happens when both kings have entered the enemy camp and neither can be mated. The Japan Shogi Association’s game rules count the rook and bishop as 5 points and every other piece as 1. That count decides the result; it is not a measure of piece strength. Shogi players also have a saying, “in the endgame, speed matters more than material”. It is the same idea as giving away pieces in a chess mate puzzle.

Daily mate practice

Chess365 is a checkmate puzzle app with 1,000 puzzles each of mate in 2, 3, 4 and 5. You always play White, you can make any legal move, and Black answers with the most stubborn defence. Every puzzle is generated and checked by the app’s own solver: the key move is unique, every white piece has a job, and in puzzles of two moves or more the key move is never simply capturing the queen. Many solutions look like a loss of material if you only count points, so it is good practice for the switch described above. Stalemate is a draw, so it counts as a failure.

How to use the app is in Getting Started with Chess365. If you want to learn how the pieces move and the rules of chess first, StudyX Mindsports has beginner lessons.