Day #1 of Analyzing Random Games

In this position it does not take a long time to work out that Black has the advantage. The a5-pawn is a potential danger, the doubled pawns are obviously weaknesses and Black leads in development.

Nevertheless there are two things that I would normally consider doing here - comparing pieces and finding ideal squares. In this situation, in order to save space, I will just find the ideal squares for Black: the a8-rook is good right now, the knight might go to f4 but for now it is well placed, the bishop could hardly be better but might at come point go to c6, the f8-rook should probably be on d8 and the queen on b6.

White has no good squares easily available for his pieces. The queen cannot improve and nor can the knight (on f5 it would be quite lonely), but the bishop might be better on g5, although this takes time. Meanwhile the queen's rook is simply miserable. Only king's rook and the king are easy to improve by castling.

About exchanges. White should seriously consider exchanging on d5, for with pawns on both side of the board the bishop vs knight situation would be beneficial for him. The exchange of queens might also be in White's favor because it is more problematic for white to find a good home for the queen. So as Black we know what we want to prevent. If we look at it from a static point of view the correct move is 20...Qb6. The reason is simple - this is how we want to place our pieces.  This also prevents Nxd5 for at least a few moves, homing on the g1-a7 diagonal. The only problem is the hanging piece. However, there is a rule called the 90% rule, which states that in 90% of all situations the move which is correct for positional reasons works out tactically. Here it makes curious to see if the move we would like to play can be played. It turns out that it works just fine for Black.

20... Ob6! 21. 0-0

After the double exchange on d5 there is a simple check on b1, winning the exchange and the game. Now black easily exploited the weaknesses to win: