How many moves a GM can think?

Nakamura can calculate anywhere from 17-30, not sure.

When I'm calculating and evaluating positions along the way during my calculations....   sometimes I forget whose move it was in the line and I have to start over again.  

That's a time wasting annoyance that GMs don't have to deal with. 

I can think 30 moves ahead... in a king and pawn endgame.

They don't calculate many moves. They probably calculate X moves ahead like a normal player, but they calculate them almost perfectly, imo.

GMs dont calculate moves, they tap into the force and the force moves the pieces for them.

OMG folks this thread is amazing. I don’t play chess (OK, I know how but it would be Riker v. Kolrami in Strategema so let’s pretend I don’t know). I needed to make a point for a marketing presentation about “thinking moves ahead” so I am online, quick Google search, chess thinking moves ahead, and up pops your thread. Nine-year-old thread!

Thank you for the obsessive detail as it did answer my question and was very entertaining. For my meeting the line will be “at least 10 moves ahead”. In reality it is somewhere between 5 and 30 moves ahead quite likely and more likely between 10 and 15 on average. And there are also likely savant esque extreme variations.

The real question seems to be not how many moves ahead can you see forward, but how many you need. And that need varies based on talent, intuition and ability to read your opponent. But you eidetic, logical, GMs have danced around the real issue for the last 9 years.

Theory of a multiverse where in every instance every possible option that can happen, does happen. Not as hard to calculate in a fixed game of chess, but someone has to have, or will have some algorithm that can calculate all possible moves from any point within the game, recalculating with each move the best strategy to win. Needed or not, intuition and raw talent notwithstanding, if you had a mind that could calculate each possible move, from any point forward in a game, and automatically recalculate your position and strategy, to the Nth move ahead, always, you should never lose. No human has that kind of mind.

Question is why hasn’t it been done with a computer yet? I mean chess has its complexities but it’s not like mapping the human genome. There are a finite number of possible moves, from the 24 possible opening moves mapping each possible move would likely be tedious, but doable. Maybe even not that tedious with the right script. And there have been chess programs forever so why hasn’t that logic set been set in stone years ago? And if it has, how can you explain a human GM ever beating the machine?

There are 10^120 possible games of chess, more than the number of atoms in the universe.  With today's computing power it would take the age of the universe to "solve" chess.  Neural networks like AlphaZero are the new paradigm shift to a mathematical brute force solution.

Asking how many moves a GM sees ahead is also kind of vague.  On a clear board with a rook and a couple bishops there might be a 30 move sequence to walk the opponents king across the board into a corner for mate or escorting a pawn up the board for example.  You could say the player is calculating 30 moves ahead, but really they are just using a pattern of moves over and over to edge the king into the desired position.  You really have to define the question more specifically- for the same position for example, how many moves ahead would an 800, a 1600, a 2200 player and a gm calculate?  That would be a more relevant question. Also would you include repetitive moves in known mating patterns or sequences count as individual moves?  

Richard Reti was asked how many moves he looked ahead, and he replied "As a rule, not a single one."

Reti was an artist at the chess board:

yes because they choose reasonable moves to look at. Nobody caculates na3 like a noob