The title is somewhat of an intentional anachronism. Knights had swords, not rifles, so they would never employ a bayonet. But I’ll explain later.
In my second game of the Championship Qualifier, I got skewered by the Bayonet. I tried my usual improvisation through the late opening against a prepared opponent and got an inferior position. In the critical moment, my opponent found a combination that wasn’t quite winning in all variations, but scary enough to rattle me and I chose badly, giving up my queen and pawn for rook and minor. The rest of the game was losing for me. I tried to hang on, but couldn't see any counterchances, and I resigned before a second queen came after me.
Kramnik apparently won a couple King’s Indian Bayonet games against Kasparov, causing the latter to avoid the opening and then the King’s Indian’s popularity waned at all levels. Teimour Radjabov remains a steadfast champion, essaying it five times against the likes of Kramnik (draw), Gelfand (win), Aronian (loss), and Carlsen (draw) in the 2008 Corus Wijk aan Zee Tournament.
This week, I started some chess lessons with my opponent. He’s not much higher rated than I am, but he knows a lot more theory and he's beaten me 4-0. In our first lesson, we ran through the ideas of the King’s Indian Defense, Mar del Plata Variation, Bayonet Attack. It was quite valuable because he corrected a lot of misconceptions that I had about the Bayonet. The King’s Indian is usually a no-holds-barred opposite side attack similar to the Yugoslav Dragon but with locked pawns. Here are two of my major misconceptions: 1) I thought the Bayonet was an accelerated attack on c5 and d6 and 2) Black’s queen bishop is too valuable to give up for a marauding knight at e6. Black’s queen bishop is a key piece that is not only his good bishop, but also often sacrifices for a pawn on h3, delivering the last blow of the battering ram and destroying White’s fortress.
This is what I learned. The idea of b4 is not necessarily to quickly advance c5 and attack d6 so much as to provide space for Qb3, Rb1, and Bb2/a3. White bides his time, keeps his knight at f3, and patiently waits for Black to play his thematic f5. White responds with Ng5 and eventually Ne6 with Bxe6 likely forced and then pressurizes d5 and the crumbling center. If Black uses time to prepare f5 with h6, then White switches back to the flank idea with Nd2, c5, Nc4. In the lines with Ng5-Ne6, Bxe6, Black maintains his chances mostly by maneuvering his knights around the center. For example, right after the Bayonet move 9.b4, Black plays 9…Nh5 eyeing f4. White typically plays 10.Re1 to create a retreat square for the Be2 in case of 10…Nf4. The game often continues: 10…f5 11.Ng5 Nf6 12.f3 Nh5 13.Qb3 Bf6 14.Ne6 Bxe6. Black continues to maneuver his knights all around to help encircle the pawn on e6 and get whatever good posts they can get. The win of a pawn helps compensate for the loss of the key queen’s bishop.
From the start of the game, Black’s king knight typically goes Ng8-Nf6-Nh5-Ng7(after Bf6)-Nxe6 while the queen knight typically goes Nb8-Nc6-Ne7-Nc6-Nd4-Nxe6. I was thinking about the movement of the knights in the game, especially the circular movement of Black’s king knight and I thought of the Knights Errant doing De La Maza’s Circles.
12 hours ago