I tried playing GO in college, I wasn’t good at it and after a while I gave up. But I remember a few very basic things. The point of the game (if I remember correctly) is to surround your opponent’s pieces, each person putting one piece at a time on a line intersection, so you can remove them from the board. It is possible to build safe positions that can’t be penetrated by the other side. To my very rough and inexperienced eyes, it appears white has the dominant position in most areas of this board. I did a quick rough count, and it appears white has around 6 more pieces on the board, which if so would mean they have already removed some black pieces. Again to emphasize, I am not an expert, and the game strategies are, to me, very complex.
There are both local tactical, and global, considerations to estimate the value of a position. There is surely a lot of pattern recognition going on— for instance at a very basic level to figure out which groups of stones are alive, dead, or still unsettled. This is something that can be practiced by playing games on a small board; e.g. look at the following picture from Wikipedia:
[ok it doesn’t work; scroll down here: Life and death - Wikipedia]
to Status of Groups of Four. Surely someone with a little experience does not have to puzzle out each configuration like that anew for a couple of minutes each time.
Accurately assessing territory, especially in the early phases of a game, is a fairly advanced skill and there isn’t really a simple algorithm. On this board, it seems that white has a large territory on the upper left and black has a large territory on the lower left, except for the small white group in the corner. But whether these are really territories depends on whether they can be invaded. Assessing whether an invasion is possible can be very difficult. If it were not, the game wouldn’t have much interest. Smaller groups like some of the ones on the right side of the board need to be assessed to determine whether they are alive, dead, or unsettled. This is also an extremely difficult skill to learn. I have a book about life and death problems which I bought in 1992 and have read through a number of times since then. I still don’t have the ability to read through complicated life and death problems accurately.
BTW that is a very good documentary. I watched the Lee Sedol - AlphaGo games live when they happened and they were riveting. A lot has changed in the go world since 2016. Lee Sedol has sadly retired from playing professionally (either because he is demoralized by AIs or because of dispute with a go organization; accounts differ). AIs have completely dominated the game and no one expects a human player to beat a good AI anymore. Professionals study AI games and they have inspired new strategies and styles unlike anything played in the previous 4000 years
In some sense, the skill of being able to look at a board and tell who’s winning is the same skill as the skill of being able to play the game well. Every time you make a move, you’re choosing between a number of next positions for the board, and you want to choose the one where you’re winning, rather than one where you’re losing.
This applies equally to go, chess, checkers, or to many other similar games.
