Everyone loves a trick shot, but only the best have what it takes to attempt one when the stakes are high. Roger Federer once played a ‘tweener’ shot to bring up a match point in a Grand Slam final against Novak Djokovic. Zinedine Zidane, before his infamous head-butt, scored a ‘panenka’ penalty for France in the 2006 Football World Cup Final. During a tense run chase, Jos Buttler resorted to his trusted ‘scoop’ in the 2019 Cricket World Cup Final. The common denominator between all the above instances is, despite the obvious flamboyance, they represented the most logical course of action according to the match situation. Of all the fancy tricks the rules allow, the chess equivalent perhaps is under-promoting a pawn. When it’s done out of necessity or perhaps even as a last resort, it’s a pleasing and enjoyable spectacle that draws wows and aahs. If we discount the situations where players do it for sadistic or whimsical purposes, there are very few circumstances where it makes sense to under-promote. Let’s have a look at some of them. When the knight is right The unique movement of the knight means it can actually do stuff the queen cannot. This attribute makes knight under-promotions perfectly logical in certain circumstances. [caption id="attachment_144112" align="aligncenter" width="300"] White to play and win - (Emanuel Lasker) [/caption] The situation appears bleak for white. His back rank is weak, his passed pawn is not secure and it seems he can’t even create any threats. But he has a nice combination starting with 1.Rc8+ Rxc8 (1...Kxb7 2.Rxd8) 2.Qxa7+! (aha!) Kxa7 3.bxc8=N+ (the point!) Kb7 4.Nxe7. White ends up with two additional pawns and a winning endgame. When a rook is better than a queen The queen is the most potent offensive force on the board. The conclusion of the famous Saavedra study is a fantastic example of a situation where one is better off without some of those powers. [caption id="attachment_144114" align="aligncenter" width="300"] From the Saavedra study - White has just played 1.Kb3-c2[/caption] No longer able to stop the promotion of the white pawn, black finds a surprise defence with 1...Rd4!. Now 5.c8=Q Rc4+ 6.Qxc4 is a stalemate. White turns the tables with 2.c8=R Ra4 3.Kb3, winning by attacking the rook and threatening 4.Rc1#. Stalemate avoidance, as we saw above, is one of the major motivations to underpromote a pawn to a rook or a bishop. Still they are exceedingly rare in tournament practice. Perhaps the best known example is the following. [caption id="attachment_144117" align="aligncenter" width="300"] Alapin - Rubinstein (1908) - Black to play[/caption] Black has several ways to win, but Rubinstein found the prettiest (and the most efficient) method. 1...Nf2+ 2.Bxf2 gxf2 3.Kxh2 f1=R! 4.Kh3 Rh1#. A rare under-promotion Let’s have a look at a slightly more difficult example of stalemate avoidance. In the following mate-in-four problem, black still has two pawn moves left before stalemate becomes a factor. White must allow him a third move, and deliver the checkmate with the fourth. [caption id="attachment_144120" align="aligncenter" width="300"] White to play and mate in four moves - Ovon Kroshofer (1904) [/caption] White has a neat checkmate in five moves starting with the obvious promotion. 1.c8=Q b3 2.Ra1+ Kb2 3.Qh8+ Kc2 4.Rf3! followed by either 5.Qh2# or 5.Qc3#. But finding a mate in four is far from easy. 1.c8=B! The whole point becomes clear later. 1... b3 2.Bg4 b2 3.Bd1 (avoiding the stalemate) Kxb1 4.Bb3#. What an entrance by the bishop! The whole concept of under-promotion epitomizes the beauty of chess. It can lead to tricks, laughs and despair. Imagine the following position occurring in one of your blitz games with only ten seconds left on the clock. What would you do? [caption id="attachment_144123" align="aligncenter" width="300"] White to play[/caption] If you found the ‘solution’ 1.Kg3 f1=Q 2.Qh2# immediately, congratulations. You just earned the ‘privilege’ of reading this column again.