How to Find the Best Opening Moves
How do you find the best moves when studying the opening?
This is a question I am asked quite often, since everyone has a natural desire to have a good opening, and this phase only increases in importance as you become stronger.
The short answer is, to find the best move in a position, you have to analyse the position, using the resources available to you - engines, databases, cloud servers, opening books and so forth. There is no substitute for hard work in chess!
However, there are some shortcuts that can accelerate this process for us. (Read to the end to learn the fastest way to improve your openings).
1) Which move is played by the best players?
By best players, I don't just mean the top human players (Carlsen, Caruana, Ding and so forth), but the best engines and the best correspondence players.
Let's return to the position at the start, which is reached by the following moves in the Zaitsev Ruy Lopez:
In this position, the recent trend in correspondence and human games is 18.g4. However, 18.Nh2 is the most common move, and is still played sometimes in correspondence.
Why else might we want to start our investigation of this variation by analysing 18.g4?
2) Which move performs the best?
According to my basic reference database, 18.g4 scores an impressive 75% for White. And this is not just over a few games, but a solid sample size (89 games).
While this is useful for getting an idea of a move's practical merit, the score/performance rating a move achieves won't always tell us the full story. Also, the statistics may vary a lot, depending on whether you're using over-the-board games, engine games, correspondence games, or all of the above, to form the statistics!
For example, a move may score very well, but in the last game with this line, it was refuted spectacularly. After that, everyone knew the move was bad, and no one played it anymore.
A good example of this is the following brilliancy by Anand, which must be one of the 1000 most-quoted games in the chess circles:
What is the story here? Why did Aronian play a variation that everyone now knows to be not that great? We can answer this by looking at the history of the variation.
Initially, White was meeting 11...Rc8 with 12.Rd1, but Black was able to prove equality with the nice 12...c5! pawn sacrifice. White then switched to 12.b4 (which is the modern main line), and there were discussions about whether Black could fully equalise with 12...a5 (which we now know to be the case).
Then, Aronian's team must have come across the following high-level game, which, despite ending in a draw, had a successful opening outcome for White:
However, Anand was able to reconstruct his preparation in the game, and show why 12.Ng5 is not objectively dangerous for Black. I invite you to challenge yourself, and see if you can play like a World Champion:
Well done if you successfully played all of Anand's moves!
3) Which move looks the most critical?
When analysing a position, either in a game or in analysis, it's best to analyse from the perspective of 'What could be wrong with this move?'. It is a more critical way to examine things, than trying to 'sell' all the good points of a particular move. The best 'selling point', to me at least, is that a move has no problems 
So, if we return to our first position (after Black's 17...f5), the move 18.g4 is certainly very ambitious, aiming to resolve the central tension in White's favour. That's another reason why I would analyse it first.
4) Try to be as unbiased as possible.
In most cases, when we analyse an opening variation, it is with the intention of playing it for one side.
This can lead to us, consciously or unconsciously, either not looking for the opponent's best moves as critically as our own (a common mistake at the starting levels of play), or overestimating our chances in assessing positions (e.g. 'the engine says it's equal, but I think White is just better' without a clear reason).
Admittedly, I also made this mistake when I last examined this position.
You see, I had first come across this 18.g4 idea in a Yearbook Survey by Erwin L'Ami, where he showed some wonderful high-level correspondence wins for White.
Here is one example:
After seeing such a game, it would be easy to assume, like I did, that Black's 17...f5 is simply a mistake, and that Black should play the modern main line 17...c4 (which is analysed very well in Kuzmin's 'The Zaitsev System')
However, this snap judgement would fail to take into account possible improvements for Black. For one thing, what if Black plays 18...f4 instead, closing the kingside? After all, is that not where White typically attacks in this line?
Indeed, by sacrificing a pawn with 19.Nb3 Ne5!, Black seems to be able to hold this position. The following correspondence game is a key example for our understanding:
Of course, such a crazy position should be analysed thoroughly before playing it at the professional level, but the point is to show how to turn the data and statistics in our database into logical conclusions for practical purposes.
If I wanted to make 18.g4 work for White, this game is where I would look for a possible improvement:
The engines give White an advantage in this variation at lower depths, but once you get to around depth 40+, it starts to conclude that the position is close to equal. That's a sign that this could be a good practical choice, in that Black has to play somewhat precisely to prove equality.
This brings us to the fifth key point for finding the best moves:
5) See what the engine recommends at a high depth.
I deliberately left this tip to near the end, because, in my experience, many players rely just on the engine to tell them the best move, and in many cases, at an insufficiently high depth to draw the right conclusions, or even find the best move.
That may sound a bit harsh, but I have also been guilty of this at times. There are many methods that players can use to make the most of their engines, and help them draw the best conclusions.
One of the simpler techniques is to not stop analysing too soon. We should continue our analysis until there are no more tactics in the possible, and no more forcing moves (checks, captures, or threats). In such strategic or 'settled' positions, we can more easily make a precise verdict, now that the position has clarified.
The underlying ethos behind such techniques is to verify things, and not take an assessment or move for granted, but try to understand it, and go deeper. Also, different engines will give us different ideas and assessments, and it is a separate skill to know which engine has a better understanding of the position.
As an example of how this can work in practice, let's see an interesting position from a recent spectacular game:
Here is your chance to apply the sixth and last key point for finding the best moves in the opening:
6) Use your human understanding, intuition and curiosity to draw your own conclusions.
How do you assess the position? Does Black have enough compensation for the queen?
I noticed that the older engines often struggled to correctly assess positions with material imbalances, such as several pieces for the queen. At first, Stockfish 10 gives White a big advantage, but at depth 45, it gives the common assessment '0.00'.
Unless it really is a forced draw, '0.00' is often a sign that we should look deeper to see if the position is really equal, or, if it is, who has the onus of 'proving' this equality.
I also fed this position to Leela Zero - the neural network engine based on AlphaZero. It concludes, at depth 13, that it is Black who is slightly better - though both engines agree that 11.Bd3 is a better move, despite allowing further damage to White's structure with 11...exf2.
(If you are wondering whether to use Stockfish or AlphaZero in a position, then I would echo NM Han Schut's advice, that Stockfish is generally more reliable than AlphaZero in forcing positions with a clear solution, while AlphaZero is better in appreciating long-term factors, especially long-term compensation and initiative).
Of course, if you were preparing to play this as Black, then you would analyse the different options, and make sure that you understand such a crazy position before entering it. However, based on the information above, we can draw a logical conclusion that this position is much easier to play as Black, whether it is objectively equal or objectively better for Black.
After all, from a human perspective, the queen often struggles against several pieces in situations like our position above, where Black has very good piece coordination (especially on the dark squares), no weaknesses, and White has weaknesses and difficulties harmonising his forces.
Although it is just one example, the game played out in line with this verdict:
If you applied the previous tips, you will have seen that White can avoid this queen sacrifice idea with 8.Bd3!, with advantage. But we have already gone quite deep in this article.
Do you remember the key methods for finding the best moves in the opening?
Let's review them:
1) Which move is played by the best players?
2) Which move performs the best?
3) Which move looks the most critical?
4) What is the objective assessment of the position? What are the improvements, for both sides?
5) What does the engine recommend, at a high depth?
6) What is my human understanding/feeling for the position? What is my conclusion?
If you apply all these tips well, have very good (and up to date) databases and engines, work hard, and have a good understanding of chess (or are adept at using the engines correctly), then you can have extremely strong opening knowledge - up there with the best in the world, even.
However, I promised early in my post that I would share the fastest way to improve your openings. If you read all the way to here, then I congratulate you on your diligence, which is, after all, an important quality for becoming a great chess player!
The truth is, a lot of players find opening study quite difficult, frustrating, or even boring. Granted, I was passionate about opening study from a young age, but I am more the exception.
Those of you who have read several of my posts/articles will know that I am a Grandmaster, the 2018-19 Australian Chess Champion, and an experienced chess trainer and writer. Many esteemed sources (including Chess.com in an interview) have called me a 'chess theoretician'.
If you liked this post, and are inspired to improve your openings, then I may be able to help you further with this process. Using my coaching and playing experience as well as my own research, I have adapted these techniques to find what variations and moves are the most effective at different levels of play.
To apply to become a private student of mine, fill out and submit this
To stay up to date with more chess and life improvement content, you can
What other challenges are you facing in your chess games?
I will see you in the next post!
