Quote:
Originally Posted by LektorAJ
I agree, they calculate plenty deep - I just think they other poster is saying they don't need to calculate as wide as a computer because they are (or were?) better than a computer at reducing the list of candidate moves at every node to just the lines that need to be analysed - I'm pretty sure he's not saying they only analyse 3 moves ahead total on every line - just that 3 moves is often enough to discard a particular line.
I also think that the rest of are not saying that Anand can calculate every possible position 20 moves in advance, just the main lines. So I don't really understand why everyone's posts have the form and tone of disagreement here.
Regarding a GM being better than a computer at reducing candidate moves, this is both true and not true at the same time. If you start at a point before any thinking(or in the computer's case, calculating) has occurred, this is true. A GM will intuitively eliminate moves that are irrelevant, and won't even register other possible moves outside those few. However, a computer will arrive at a similar list almost instantly, for essentially the same reason. Typically, only a few moves have any value, and arriving at those doesn't take much.
This speaks more about the power of the human subconscious than anything else. That's why chess engines have only fairly recently surpassed humans as well. The human intuition and subconscious is a beautiful thing. An enormous amount of work and advancement in technology was required for a computer to rival a skilled human's ability to see why a move is better due to the effect it will have at a much later point. That's what's really fascinating, I think.