Chess Piece Values Table
| Piece | Classical | Kaufman (statistical) | Engine estimate |
|---|---|---|---|
| Pawn | 1 | 1 | 1 |
| Knight | 3 | 3.5 | 3.2 |
| Bishop | 3 | 3.5 | 3.33 |
| Rook | 5 | 5.25 | 5.1 |
| Queen | 9 | 10 | 8.8 |
| Pawn by file | Value | Passed pawn by rank | Value |
|---|---|---|---|
| Centre pawn | 1.00 | 4th rank, connected | 1.15 |
| Bishop pawn | 0.95 | 4th rank, passed | 1.30 |
| Knight pawn | 0.85 | 4th rank, passed & protected | 1.55 |
| Rook (edge) pawn | 0.70 | Bishop pair bonus | +0.50 |
Chess piece values are the game's exchange arithmetic: a roster of rough numbers that tells you whether trading your rook for his queen is a bargain (yes) or trading bishop for knight is a wash (usually). The classical scale - pawn 1, knight 3, bishop 3, rook 5, queen 9 - has survived two centuries because it is easy to add at the board, and it quietly encodes the mobility hierarchy: minor pieces, the long-range rook, and a queen worth roughly two rooks or three minor pieces.
The modern refinements matter more as you improve. Larry Kaufman's statistical revaluation lifts the knight to 3.5 and the queen to 10, and adds the famous bishop-pair bonus of about half a pawn - two bishops together reach squares one alone cannot. Engine analysis agrees the classical numbers undervalue the pieces that scale with open positions, and shows pawns are not interchangeable: a center pawn is worth a full point while an rook's-file pawn trades closer to 0.7, which is why flank pawn grabs lose endgames.
How to use
- Use the numbers for exchange decisions, not strategy: the chart answers should I trade questions - giving a knight (3) for a rook (5) wins material, trading bishop for knight is even, and sacking a rook needs about two pawns plus position to break even - but it says nothing about tempo, king safety or development, which decide more games than a half-point of material.
- Add the bishop pair when counting: Kaufman's bonus of roughly half a pawn applies in open positions where both bishops can act - a bishop and knight versus two knights may look level on classical counts, but the paired bishops tip the material balance and often justify a pawn investment to keep them.
- Revalue pawns by file and rank: a center pawn is the full 1, a bishop-pawn about 0.95, a knight-pawn 0.85 and an edge pawn 0.7 - and a passed pawn grows as it advances, reaching 1.55 or more when protected on the sixth rank, which is why endgame technique is mostly about escorting one pawn while freezing the rest.
Frequently asked questions
How many points is a queen worth in chess?
Nine points in the classical system - the convention that survives in clubs and books - though Larry Kaufman's statistical work and several engines push the queen toward 10 because her mobility scales so well in open positions. For practical arithmetic the queen is about three minor pieces, two rooks, or rook plus minor piece plus pawn; that last comparison is the one to memorize, because queen-for-rook-bishop-and-pawn trades decide whether a middlegame attack was worth it.
Is a knight or bishop better?
The chart says they are equal (3 points each), and in a vacuum they are - but context revalues both. Bishops thrive in open positions with pawns on both wings and gain about half a pawn as a pair; knights prefer closed positions and outposts, and a knight on the sixth rank plays like a rook. Kaufman's revision gives the knight 3.5 versus the bishop's 3.5 - the real rule is positional: trade your bad bishop, keep your good one, and never give up the bishop pair for free in an opening position.
Why is the king not on the value chart?
Because the king is the game, not a piece you trade: losing it ends everything, so no finite value applies. Analysts do assign the king a fighting value for endgames - commonly three to four points when it is safe to participate, meaning a king that walks to the center of a pawn endgame acts like a strong piece. But on the chart the king's column stays empty on purpose: every exchange question in chess is asked with the assumption that both kings survive, which is exactly what the point system is for.
What is an exchange sacrifice?
Giving up a rook (5) for a minor piece (3) - deliberately conceding two points of material for position, initiative or a decisive square. The chart's arithmetic frames the cost: you need roughly two pawns of compensation, or a bind worth more, to break even. Famous examples turn on the bishop pair or a paralyzed enemy piece being worth more than the missing two points, which is the deepest lesson of piece values: the numbers price material, but strong players price squares, and the best sacrifices buy squares with material.