Good question, but rest assured, the most important part has already been achieved! Namely, you've taken full notice of a pattern in how you often lose your games. To remedy the problem of avoiding combinatorial/tactical blunders can be a somewhat systematic (thus definitely achievable) process, so it's definitely the easier part.
Always check for forcing moves:
These typically are, ordered according to priority:
- Checks: clearly, the most forcing moves are checks, without exception and almost in a computer like manner you should exhaust the list of checks your opponent can give. Example (1).
- Captures: Take note of the possible captures in the position, it shouldn't matter how crazy it might appear at first glance (like giving up a queen for a rook, but a mating threat might be following, example (2))
- Positional gains: Relative to the first two, this one's admittedly rather vaguely defined. However, as a shortcut, let's simply treat them as possible combinations in the position, which I will further elaborate below.
Types of combinations to consider:
- Checkmating combinations: This is the easiest combination to understand, as it is purely concrete and when it works, immediately decisive.
- The combination to win material: Can be any forced sequence of moves, e.g. starting from a bait (a sac'ed pawn), which ends in winning or trapping more important material.
- The third type, is the toughest, the strategic combination: again it might start from a forced sequence, but for a specific type of a goal, such as creating the king's position vulnerable, or in contrast to create a defensive setup, it might be to an exchange sac opening up the position for our pieces or permanently damaging our opponent's light/dark squares, it might also be a sequence that creates a weak pawn or one that ends up with you having an endgame with a bishop pair, and the list goes on.
How to efficiently check for all these:
- Keep calculating variations both on your own and your opponent's time, irrespective of how winning or equal you might find the position.
- Keep taking mental notes about positional aspects of the position throughout the game, namely anticipating critical moments where: the pawn formation might change, or potential of changing it thereof (e.g. noticing a minority attack could be achieved, see here for more details), the king safety can be compromised (either by a pawn advance, a sacrifice, or a simple trade of a good defensive piece guarding the king), and last but not least, the combo of pieces (bishop pair, Q+N vs Q+B, etc).
- Keep asking yourself questions about your opponent's decisions, no matter how strange they might appear sometimes: e.g. "why did they play
a6? what are they trying to prevent?" or "why are they not resolving the center?" ... Again this helps keep you sharp during the game and not relax too early (and miss simple ideas).
- Take note of whether one of your defensive pieces are possibly getting overloaded.
- Take note of tactical themes: hanging/attacked pieces, pinned pieces, discoveries, etc.
- Step up your tactics training! Sounds obvious but this really should be part of your daily practice no matter the level of play. This will not only add to you bag of tactical ideas and motifs, but it will also prevent your calculation skills from getting rusty.
- And finally, don't make assumptions about variations that you'd like to see occur, instead consider your opponent's alternative options specially when making a committal decision in the game.
Finally, for the promised examples of (1) and (2), here's one where both are combined:
[title "Meier vs Kramnik 2013"]
[fen "1r3bk1/2R2pp1/p3p3/4q3/Qn2N2p/5BP1/4PPKP/8 w - - 0 1"]
[startflipped "1"]
1.Qd7 {Here's the position of interest, white's last move Qd7 both defends the rook on c7 and creates a threat on f7} Qf5 {Kramnik defends f7 with Qf5 while also preparing h3+} 2.g4 {Meier notices that black is playing a potentially dangerous game allowing the battery against f7 and having only their queen as defender. White then try to decoy the queen from defense of f7} Qf4 3.g5 {Setting up a neat combination with Nf6+, which is a tactic of type (1) that Meier is planning for, a move that would blind the queen's access to f7} Nd5 {casually played by Kramnik, almost in a manner suggesting he's missed Nf6+, but it is in fact a bait} 4. Nf6+ {let's first see the candidate lines I reckon white had anticipated, namely gxf and Nxf} Nxf6 (4...gxf6 {other losing line} 5.Qxf7+ Kh8 6.Qh7#) (4...Qxf6 {!!! Here's our case (2), white had simply dismissed the crazy possibiliy of queen taking on f6 instead of pawn or knight. The queen sacrifice comes with a checkmating combination, surprisingly of the backrank type despite white's fiachetto pawn formation!} 5.gxf6 {clearly then seeing the previous line, white cannot take the queen, so game continued as follows} (5.Qxd5 Qg6 6.Qe5 Rb5 {and with g5 falling white resigned soon thereafter.}) Nf4+ 6.Kf1 Rb1+ {and mate is unavoidable.}) 5.Qxf7+ Kh8 6.gxf6 Qxf6 7.Qh5+ Qh6 8.Qxh6+ gxh6 9.Ra7 {likely a draw, despite black's shattered pawn structure.}
To finish, I leave you with a problem (though relatively simple so it mimics what you might have experienced in your game), where to recognise the combination in a real game we would:
- take note that white has back rank issues
- rook on
f4 is attacked by the queen
- and our rooks are doubled along the open e-file
These already bring to mind: forcing moves such as captures and checkmating combinations.
[title "black to play and win"]
[fen "4r1k1/p3rppp/2pqb3/2p4Q/N2p1R2/1P1P4/P1P3PP/5R1K b - - 4 24"]
