My Thinking System

When you play a game of chess, how do you decide on your move?  I mean that literally: how do you do it?  What goes on in your brain?  For a long time, I couldn't answer this.  When I was 1800, my process was as follows:

1.     Stare at the board.
2.     Have a move pop into my head.
3.     Calculate to see if it works.  If so, play it.
4.     If not, go back to step 1.

Essentially, I played chess randomly.  If my intuition suggested good moves, I did well; when it didn't, I played terrible.  There was no middleground.  Worse, the random nature made my results random.  I could play the game of my life and follow it up by dropping my Queen on move 9.  Needless to say, I plateaued at 1800, and I attribute the main culprit being this complete lack of structured thinking.

I eventually discovered GM Smirnov, and his advice basically boils down to cultivating an efficient thinking algorithm.  This has done wonders for me: I surged past 1800, approaching 2200 in both long-form correspondence and short-form blitz.  More importantly, it opened my eyes and helped me appreciate the beauty and logic of chess.  Adding structure has made chess much more enjoyable.

This post describes my current thinking system.  I've written about it before on various forums, but I can dive into detail here, so let's do it.

The Summary

When looking for a move, I ask myself the following questions:

1. What is my opponent's idea?
2. Do I have any tactics, any breakthroughs?
3. If not, how can I improve my position?
4. Is my move safe? 

This approach lets me apply all of my chess knowledge on every move.  Each question prompts me to think in a certain way, and altogether it encapsulates everything I need.

Step 1: What is My Opponent's Idea?

As soon as your opponent makes a move, your first thought should be, "What is my opponent's idea?"  There's nothing groundbreaking here; this advice goes back centuries. If your opponent is threatening something, you want to see it!  Indeed, being able to see your opponent's threats and then react is one of the first skills budding chess players learn as they start improving.

This is normally framed as, "What is my opponent's threat?", but I prefer using the word "idea".  Sometimes our opponent has a clear plan, even if it is not a threat.  For example:

In this position, White's last two moves were Qe2 and Bd2.  White clearly wants to castle Queenside; that's the idea.  By recognizing this, I can prepare for it.  My previous moves had prepared the e5-break, but ...c5 looks better now, because it responds to White's idea.  I might have played ...c5 regardless, but noticing White's idea makes it much easier.

This also goes into chess psychology.  By looking at my opponent's moves, I can predict what sort of player he or she is.  In the above example, castling Queenside is aggressive and prepares a dynamic middlegame.  These sorts of players often dislike endgames and more strategic positions where they cannot attack.  Therefore, I might want to trade Queens and head for a drier position, which might rub against my opponent's natural inclinations.

That might sound farfetched, but I have numerous examples of using this sort of chess psychology as a weapon.  More fundamentally, though, if you ask what your opponent's idea is, you will see his threats and you will be much less likely to blunder.  Beginners that internalize this step often make mammoth rating gains just from blundering less.  It's really powerful.

Step 2: Are there any Tactics or any Breakthroughs?

After noticing my opponent's idea, I immediately look for tactics.  This encompasses two related ideas.  First, we have what GM Smirnov calls counterblows.  In essence, it means when our opponent attacks us, our first thought should not be to defend or to retreat but instead to see if we can hit back even harder.

For example, we want to notice that White is attacking our Knight with his last move, but we do NOT want to automatically retreat.  Instead, can we create a bigger threat?  Here, ...Qa5+ comes, gives check and picks up the undefended pawn next move.  If we automatically retreat, we miss these sorts of tactics.

Second, there is a well-known phenomena where players have tactics ratings several hundred rating points higher than their normal rating.  Most of us have solved tactical puzzles involving amazing sacrifices, but few of us use those same sacrifices in our games.  A big reason for that: we think differently when we play games versus when we solve puzzles.  If we treated them the same, we'd have similar results.

By asking, "Are there any tactics?", I'm treating the position like a puzzle.  I look at the "normal" puzzle moves, the big checks and common sacrifices, and I see if they work.  If they do, great, I win!  If not, oh well, at least I looked.  If you don't look, you'll never find them.

Obviously, this is not exactly the same as a puzzle: I have a ticking clock and no assurance a tactic exists.  If you do this too much, you run out of time, so you have to pick and choose which positions get the "full puzzle treatment" and which get only a quick glance.  Nevertheless, every move we want to check if we have something forcing, because if we do, we win.

I include the word "breakthrough" because of a personal weakness.  For example, look at the following position:

White has no tactics to win material, and so I would default to something like Rad1 to finish development.  I routinely missed that I had the option to sacrifice on h6 to break open the cover and attack the King.  These moves never crossed my mind.  By asking whether I have any breakthroughs, I look to these sorts of moves, and indeed, Nxh6! is a winning shot.

Nowadays, sacrificing on h6 looks intuitively obvious and is my first thought, but I still miss breakthrough-style tactics regularly when I don't focus on them.  That's why I still remind myself, "Do I have any tactics?  Or do I have any breakthroughs?"  You can tailor this to your own situation, but for me, it is a very important reminder.

Edit: just in case it is not clear, this is where calculation comes in.  Looking at forcing moves, checking for tactics, examining breakthroughs, all of this is backed by calculation.  I felt the need to include this here, though, because otherwise I wrote an entire piece on thinking in chess without mentioning calculating!

Step 3: How do I Improve My Position?

If I can't see any tactics, I move to the positional realm and ask, "How do I improve my position?"  This is a loaded question, because it basically encompasses everything: piece quality, open files, piece quality, weak pawns, king safety, potential outposts, evaluation of plans, you name it, it's covered by this step.

Admittedly, this is a bit of a cop-out.  It basically encapsulates every concept written in every chess book, and obviously it is impossible to cover all of that on every move.  Fortunately, for most positions, you don't need to think about most things.  Let's look at a few examples.

 In this position I have no tactics and am down a pawn.  It's opposite-side castling and Black is preparing h6-g5 to get his pawn storm going.  If I play slowly, Black will either just attack me or start trading down into a won endgame.  I need to do something to generate counterplay, and Black's advanced pawns near his King are the obvious target.  Opening lines with b3 and doubling Rooks on the Queenside files immediately jump out.  Indeed, no other plans make sense.  

I will want to keep evaluating whether a sacrifice on c4 works, but otherwise, I know how to improve my position. Passed pawns don't matter, outposts don't matter, nothing else matters.  Simple. Next, here's a completely different type of position:

In this position I again have no tactics but have equal material.  I have more space and Black has nothing to compensate, so I'm just pleasantly better.  Finishing development by bringing the Rooks to the center seems good and obvious.  Black has less space, so avoiding exchanges makes sense; I may want to consider g4, forever stopping ...Bxf3.  On that same theme, Black's pieces look clumsy, especially his Knights, which can barely move.  Playing a4-a5 boots the Knight back to c8, gains space and activates our a1-Rook.  It may also allow Ba3 to attack d6, though the Bishop also looks perfectly fine on b2.

In this position, the concept of space does all the heavy lifting: I focus on it and quickly come up with a plan: push the a-pawn and strive to avoid exchanges.

If you are more of a beginner, this step may seem overwhelming.  It probably is.  My intuition guides me here, and I developed that over literally thousands of games. It's also my favourite part, what I like most about chess, so I don't need sub-questions or anything else to guide me further.  Again, your individual situation may differ.

Step 4: Is My Move Safe?

This is also known as the blunder-check.  Once you decide on a move, ask one question before making it: is it safe?  Step 1 prevents missing your opponent's threats, and this Step makes sure you don't stumble into one.  I think GM Smirnov explains it best: ask what attacking moves your opponent might play in response. Here's an example:

Black has just audaciously castled Queenside.  My first thought is opening it up with d4, using those lines against him.  Let's do it ... but wait.  Is it safe?  What attacking moves can Black do?  Ahh, he has 1...g5!, hitting our Bishop, and after 2.Bg3 g4! hitting our Knight.  And once our Knight moves, we lose our defender of d4.  In other words 1.d4? g5! will win a pawn for Black.

It only takes a few seconds, but it will save you hours of regret when you forget and make a one-move howler.

Reality Check: Do I Actually Ask These Questions?

Do I actually do what I said I do?  Yes and no.  Yes, I go through this general process, but no, I am not consciously asking these every time.  Most of it is unconscious habit at this point.  When my opponent makes a move, I automatically consider his idea, and I (generally) think about my attacking opportunities before anything else.  If I have no tactics, I intuitively check for improving moves, and I ... mostly do a brief blundercheck before moving.

Actually, this is probably the step I skip the most, and it's the one that costs me the most points.  I get so excited about my own ideas that I don't take the few moments necessary to look for my opponent's immediate replies.  This isn't so bad in longer games but absolutely murders me in blitz. 

Anyway, the point: I have done this so often that it is now a habit.  Some are stronger habits than others, but it's a habit.  It happens automatically.  Indeed, this is why I praise GM Smirnov's courses so much: he prominently explains the power of automating these thinking habits, and his exercises help lock it in.  Doing this helped break me from my 1800 plateau and propelled me over 2000, for which I'm forever grateful.

How do you make this a habit?  You just have to do it.  You can start with one question or with the whole series.  I spent roughly one hour a day for three months going through Smirnov's "Grandmaster's Positional Understanding."  That cemented my thought-process.  You don't need an hour, and you certainly don't need three months: different places say different things, but habits form somewhere between 21-64 days.  Do it consistently, do it frequently (ideally everyday), don't slip into old patterns (the hardest part initially) and you will get there.

I will add, though, that I sometimes do ask these questions consciously.  Some positions are so complex, so wild, or otherwise so demanding that I don't know what to do.  My chess intuition turns off.  It feels like I'm flailing about.  That's when slowing down and asking these questions helps me to orientate and make sense of the position. Also, breakthroughs are not automatic yet, so sometimes I need to literally remind myself to look for those during a game.

Anyway, that's how I reason through a position: figure out my opponent's idea, strive to counter it, check for breakthroughs, look to improve my position and then blundercheck.  This results in a pretty universal thought process for almost any position.