This was from the game Christiansen-Spassky Linares 1981. Linares was a very strong tournament, and in 1981, the current World Champion Karpov was playing. But the sensational result of young US GM Larry Christiansen, to come equal first with Karpov, ahead of elite players such as Spassky, Larsen and Portisch was the main story.
In the above sharp position with opposite side castling white has an advantage, but the unbalanced nature of the position means that white has to be careful about maintaining his advantage. So what might be a train of thought here?
1. Bxf6 wins the d-pawn as black's knight on f6 is the first defender. 1.Bxf6 Qxf6 2.Nxd5 wins a pawn and the knight improves it's position and gains time by hitting the queen. The question is whether there is anything better than this, as white's bishop is pinning black's knight, so trading it might not seem the best try.
2. e4. Seeing the d5 pawn is pinned, why not attack it again? Black's knight is not a defender as it is also pinned. After 1.e4 there doesn't seem to be a good way to defend d5, and white retains the strong bishop on g5. If black plays 1..Qd6 unpinning, then white will have to capture on f6, when we have a similar type of position to above except white's pawn will be on e4 rather than e3. Which would be better? Is it really that big a difference?
3. As the d5 pawn is pinned, and the f6 knight, the e4 square is undefended. This allows us to consider moves that we might not see at first. How about 1.Ne4? This was the amazing move that Christiansen played, threatening to take on f6, while forking black's rook on c5 at the same time. The knight can be taken 3 ways, but they all lead to a loss of material to black.
The other position I showed yesterday was from the game Schiffers-Chigorin Hastings 1895.
Black has just blocked the h4-d8 diagonal with the move ..f6. The question is how should white respond?
1. Move the attacked bishop. The only sensible retreat seems to be to 1.Bf4, where the bishop stays active. Only black will then play 1..Bc5 and be able to castle. White really wants to keep the black king in the centre for as long as possible.
2. Qh5+ would be good if black couldn't play g6 to block. Then how about a sacrifice, 1.Bxf6 gxf6 2.Qh5+. This looks good. I would just need to continue the analysis of both 2..Bf7 and 2..Ke7 and come up with convincing lines for white. If there was any doubt with either of these moves, then back to the drawing board.
3. Assuming I can't make Bxf6 work, let's go back to 1.Bf4 Bc5 but look a bit deeper. Now 2.Qh5+ is a fork winning a piece, so black can't play 1..Bc5. This might be best.
So now my head is beginning to spin with possibilities, so it's probably time to take a break and set my mind on calculating Bxf6 deeply when I return.
4. Thinking about it, with black's king stuck in the centre, how about just carrying on with development? 1.Qe2 allows black to capture on the bishop, 1..fxg5 but white can now double on the e-file 2.Re1 when black's king is coming under a tremendous attack through the centre of the board. In fact, the only defences of black's light squared bishop are 2..Ke7 which just allows me to triple on the file and win the e6 bishop, so I guess it must be 2..Kf7.
White now has the stunning tactical shot 3.Rd8!! winning black's queen as the black queen cannot take the rook as mate follows by Qxe6.
This final line was the one chosen by Schiffers, Qe2, but his calculation did let him down, because Bxf6 does win, as shown by the analysis in the game.
So chess analysis is just a continuing process of asking questions and extending your chess imagination to include possibilities that would never have appeared to you before. Part of this process is going through material that has been analysed therefore building a database of typical and non typical manouvres and tactical themes. But a big part is just looking at games and asking over and over again, "why this", "why not that", and proving to yourself why one move works better than another. Sometimes you'll get things wrong, but even this is a good learning experience, as if someone questions your judgement, and you correct your own findings, it means you are being self critical which is an important part of being able to grow.