My take on "how many moves ahead could [Insert Top 10 GM] see?"

Notice:

1. I'm a chess newbie, I just want to share my little "discovery".

2. My idea is definitely not a hot discovery. It's just my interpretation of an old-age idea that probably already crossed the minds of millions of chess players.

 

Okay, let's just jump straight to the question, how many moves ahead could [Insert Top 10 GM] see?

TLDR answer: Some IMs and GMs on quora said there's no definite limit to how many moves ahead they could calculate. If they want to they could calculate more than 70 moves ahead.

 

My answer: Indeed there's no definite limit to how much they calculate, but I think at their level those things are not that important. At that level they're not calculating moves anymore, they're calculating approach. It's a little bit hard to explain in chess terms since I'm a beginner, but let me explain using analogy from mathematics.

 

When given the question "what is 10 + 2?", normal people usually answer "isn't it obvious that the answer is 4?" And proceed to answer using simple algebra. But mathematicians approached it in different way. To answer that question they create a theory. When mere mortals juggling with their cute elementary algebra operator to answer that question, mathematicians are juggling with set theory, category theory, topos theory, etc2 [insert another obscure exotic named theory here].

 

Maybe the answer is more nuanced than this, maybe it's actually simpler than I think. Maybe they really only make a move that they feels right, and it's not about what I've mentioned above, it's just simply about talent and preparation.

Reti (one of the founding fathers of the Hypermodern school) was asked how many moves he usually looked ahead, and he replied "As a rule, not a single one."

When you read these words, you are not looking much at individual letters but the patterns of letters and short, one-syllable words like "you" and "the" are seen as one visual unit. Otherwise reading would be exhaustingly slow.

Similarly, master chess players see patterns and they see positions where they know certain moves will result in strong positions in their favor.

Consequently, most of the time they only claim to see a couple moves ahead and that is why.

10+2=4???

The first numeral is in binary, the second in decimal.

It depends on the position--how many moves ahead one can see?

In some endgames even a Class A player can see maybe 20 moves ahead.

In the middle game--it is hard to see 4 or 5 or 6 moves ahead.

As is typical, beginners think in terms of their own ability:

How good am I at calculation?
How good am I at attacking?
How good am I at making long term plans?

This makes sense, because beginners are limited by their own ability:

After 2 moves I lose track of my calculation
I have a lot of pieces near the enemy king but I can't find a checkmate
I made a plan, but next move I see it's no good and make a new one (repeat every move)

 

Grandmasters (and other skilled players) are limited by the position:

How many moves does the position require me to calculate?
Does the position allow for an attack?
Does the position allow for long term plans?

 

This makes sense becauese:

Since so many things can happen, deep calculation is not practical unless there are forcing moves
Attacks are based on things like space and piece activity.
Long term plans require there to be long term differences between the two positions (like pawn structure).

All this is to say:

Sometimes a grandmaster doesn't bother calculating at all, because it's not useful
Sometimes a garndmaster wants to attack, but he knows it's not possible in the position
Sometimes a garndmaster doesn't make any long term plans at all, because it's not possible

it more matters about the memory then the skill of how far you can calculate. for example, i have photographic memory, and could right now calculate 100+ moves, but it gets comletely useless after a point. also, it is more dependent on the ability to find the best moves in the position by looking at it. in a otb tournament i have spend 20 minutes calculate variation after variation before realizing that that move was really, really stupid

You didn't say in the original post that the first number in the calculation "10+2" is supposed to be binary, so a "normal" person would say "isn't it obvious that the answer is 12?". The funny thing is that a mathematician would say the exact same thing "isn't it obvious that the answer is 12?".

The "normal" person certainly wouldn't assume you make a mixture of binary and decimal and neither would the mathematician. Why would they make such an assumption?

Seeing variations isnt the problem

Most people are fine with the this then he does that then i do this and he does that thing

The real challenge is evaluation of a position

In quiet positions the gms know precisely where their advantage lies or is likely to lie ten moves down the line

They work to make the most of it

The point isnt ten moves down the line the point is seeing now in a quiet position out of opening theory what to do

Gms are excellent with that they just evaluate precisely and make the best move now and then repeat every move

Evaluating positions is where most noobs are weak and if dont evaluate a position how will you know what to play

If you arent playing the right move and the right line what does it matter if you could calcuate eighty moves deep in the incorrect line

If you watch masters stream on twitch (or YouTube) you see that they don't actually calculate that often long ahead. They can usually see intuitionally the good moves in quiet positions (eg. I need my bishop here, I need to control this diagonal, there is a weakness on that square, I need to protect this square), which comes from a lot of practice, experience and analyse of similar positions for years. They only stop to calculate more thoroughly when the position gets more sharp and there are tactics available (if I do this, he will do this, I will do that, no that doesn't work, how about if I do this instead and so on...). The more you play the better and faster you are at spotting the situations when you are making a positional, developing move and when there is a time to go for a combination.

It's pretty cool to watch Magnus Carlsen streams to see that he calculates much less than Hikaru Nakamura for example. He just sees the possible moves basically instantly and knows where it leads to without having to calculate very much. He stops to calculate only couple times a game (at least in blitz) and yet completely crushes any opposition. That is what being a generational talent with tons of top level games and practice behind you is like.

So that being said one needs to know how to calculate, but top level players calculate much less than a beginner would assume. They have seen so many of the positions (and patterns) already and can almost instantly recall the correct moves, ideas and combinations from their memory.

 

Magnus Carlsen has a phenomenal chess memory.  And that may be why he doesn't need to calculate as much as Nakamura. 

Pretty sure it was, "I only see one move ahead: The right one."

Seirawan once said that he reached a position which he calculated more than 50 moves ahead. He said he did it as an exercise. He knew he was winning in much fewer moves, but wanted to test his vision. It was a relatively simple pawn endgame that involved some precise calculation. 

 

Basically, GMs can calculate as far as they need to in simple positions. In more complex positions, they may truly see only one move ahead. And in mate in one situations...