# Is a bishop less than half the value of a bishop pair?

In valuing a rook, I observe that a rook on the open board commands 15 squares; his own and seven others in each direction. Divide by three and you get a value of 5.

In valuing a knight, I observe that a knight on the open board commands nine squares; his own and eight others in each direction. Divide by three and you get a value of 3.

If I were to value a bishop same way, I would observe that a bishop on the open board commands 13 squares; divide by three, and I would get 4 (plus).

But most valuations have the bishop at 3, rather than 4 despite his range.

Interestingly, some experts value a bishop pair on the open board at 7-8, even though the sum of 3+3 is only 6. Some would say that the pair is worth more than 7, even valuing single bishops at 3.5. A value of eight for a pair would be in line with my theoretical value of 4, above.

Put another way, a bishop and extra pawn is almost always at a disadvantage against a rook, but two bishops and two extra pawns are not necessarily at a disadvantage against two rooks.

Could a bishop be worth only three because of its single-colored move, even if two bishops are worth eight?

Given the concern that computers beating humans has made the game less interesting, have any experts, human or computer proposed or play tested a variant whereby a bishop can change its color by making a single move in a lateral direction to a square of the opposite color? This, presumably, would increase the value of the bishop, perhaps to four.

More to the point, have the Grandmasters made any more conventional arguments in favor of an eight-valued bishop pair?

• An 8-pt Bishop pair implies they are worth almost a Queen. It's pretty rare that someone trades a Q for B+B+P. The debate is usually about N+N vs N+B vs B+B. Perhaps Knights are over-valued? Commented Sep 8, 2013 at 15:02
• @TonyEnnis: I've been "told" (by World Champion JR Capablanca in "Chess Fundamentals" that a bishop and knight can give a queen a hard time. That said, it would not surprise me that two bishops (worth more than a bishop and knight) plus a pawn can hold their own against a queen. Commented Sep 8, 2013 at 18:34
• Can you find an example of a Master making that choice? Commented Sep 8, 2013 at 18:45
• @TonyEnnis: In this 1945 radio match, chessgames.com/perl/chessgame?gid=1125455 the Russians caught Reshevsky with a "prepared" variation in which (after netting out a pair of rooks), they had a rook and the bishop pair against a queen and three extra pawns (one doubled) for Black. They claimed they had a decisive advantage and proved it by winning the game. Assuming that the queens worth 9.5 points and the three pawns (one doubled) were worth 2.5 points, that's 12. A "decisive advantage" would imply 13 for White. Subtract 5 for the rook, and the bishop pair is worth 8. Commented Sep 9, 2013 at 12:38
• Interesting, but it's one game in the last 68 years. I'd expect this sort of thing to happen more frequently if they were actually equal. Commented Sep 10, 2013 at 0:41

A bishop pair is rarely worth more than a pawn. That would mean you're willing to sacrifice a pawn to obtain the bishop pair.

The value of pieces depends on what those pieces can accomplish, not just how many squares they control.

knight/bishop = 3 pawns because at worst, a minor piece can usually sacrifice itself for 2 pawns. So it's worth at least 2 pawns, but minor pieces are more useful than 2 pawns, and can potentially win another pawn, so theoreticians estimate its value at 3 pawns. 4 pawns is pushing it, it's less likely a minor piece can force exchanging itself for a value of 4 pawns.

The same can be said for the rook, a rook can usually exchange itself for a minor piece and pawn at any time, so it's worth at least 4 points. But the minor piece can't force the rook to make this exchange as easily as the rook can. 5 points is a fair estimation. Similar things can be said for the bishop pair and queen. It's the potential for these pieces to win more material than they can be exchanged for that increases its practical value.

Who are these "some experts"? I was not able to find any IMs or GMs that value the pair at 1-2 pawns. Keep in mind that these values are not absolute. It depends entirely on position and how your strategy incorporates the bishop pair. In a closed game, two bishops are not going to be better than 2 knights most of the time.

This topic may be of interest to you:

What is the bishop pair worth?

The value of pieces is not so one-dimensional as you make it out to be. Not only does it depend on position, but there is more to the picture than just how many squares the piece can control. A knight can jump over pieces. A bishop cannot. Thus the advantage shifts to the knight in a closed position. The knight can influence both colors; the bishop cannot. A simple example is in an endgame where the bishop cannot control the queening square, whereas a knight could. Point comparisons are meaningless here.

It would also be simplistic to say that controlling 1 square = x value. Every square is not the same, and it depends on position.