Training Diary 14 Jul 2026

The Amateur's Mind, Part 2: Knights vs Bishops

Continuing through IM Jeremy Silman's The Amateur's Mind, second half of the knights versus bishops chapter. Part one was about value shifting with position, this half goes deeper into knights specifically, closed positions, the rim, support points, and one imbalance I'd never have spotted on my own.

Rule 6: Knights love closed positions

Silman's example here has White with pawns on the light squares forming a pyramid shape, which leaves White with a bad bishop that can't do much of anything. Black, meanwhile, has a strong knight rooted just outside White's pawn chain, threatening to pick off the base pawn.

Fairly clear in this position that the knight is stronger, though the moment those pawns start trading, the bishop's value climbs straight back up. Closed positions favour knights, open ones favour bishops, and the position can flip that balance mid-game.

Rule 7: A knight on the rim is grim

One of those chess sayings most players have heard at some point. A knight sitting on the edge of the board simply doesn't control enough squares, and it usually takes several moves to get it back to where it actually belongs, the centre.

Edge squares look fine on the board but do very little. Worth checking whether a knight move is heading toward the centre or away from it before playing it.

Rule 8: Knights need supported, advanced squares

Knights aren't long range pieces, so they rely on finding an advanced square that's properly supported, what Silman calls a support point. He breaks down how a knight's value changes depending on which rank it's sitting on, from your own perspective:

Rank 1, ineffective, purely defensive. Rank 2, still defensive, weaker than a bishop. Rank 3, a strong defender, ready to attack. Rank 4, strong if secure, roughly on par with a bishop. Rank 5, a powerful attacker, better than a bishop. Rank 6, overpowering, dangerous enough that some players will give up a rook just to remove it. Rank 7 and 8, the value actually drops again, since it starts losing control of squares out there.

How powerful a knight becomes depends heavily on which rank it sits on from your side of the board, combined with how secure it actually is there. Rank alone doesn't tell the whole story, security does the rest.

Rule 9: Knights beat bishops when all the pawns are on one side

Silman's example is a closed position where White has a light squared bishop that can't affect anything, since all of Black's pawns sit on dark squares. Black's knight, on the other hand, works with the pawns and threatens to team up with the king to start picking them off one by one.

Same theme as everything else in this chapter, just from a different angle, how dramatically a knight's and bishop's value can shift depending purely on the position in front of them.

Rule 10: Deprive knights of support points, create your own

To devalue an opponent's knight, take away its advanced support points. If you have knights of your own, actively work to create support points for them instead. Silman demonstrates this with a position from his own game, Silman vs Gross, American Open 1992, move 20, White to play. I can't do justice to Silman's writing on this one, and I'd genuinely recommend picking up the book if this section interests you.

The key imbalance in the position is that Black has two knights, while White has a knight and a dark squared bishop. White's plan is to prevent any advanced support points for Black's knights, starting with b3 to restrict one of them immediately. Silman mentions showing this position to some of his students, most of whom spotted the knight versus bishop imbalance, but didn't rate it highly enough to actually try to exploit it.

Something this simple, a knight and bishop against two knights, can be enough to build a winning plan around. Not something I'd have picked out myself, which is probably the whole point of reading the chapter rather than just knowing the piece values.

Where this leaves me

Part one established that piece value isn't fixed, part two is really the extended breakdown of that idea specifically for knights, closed positions, the rim, ranks, and support points. Reading it is one thing, actually noticing any of this mid-game under time pressure is another. That's the real test, whether any of it survives contact with an actual clock.