How long do you spend when replaying a pro game?

There's so many pro games available that you'll never get through all of them no matter how fast you go-- this is a fact. There's no way to run out of them. There's so many thousands of pro games on the Internet that I get the urge to go through them quickly, one after another, because there's just so many.

But on the other hand, pros spend up to 8 hours on a game, sometimes more for older games. Since I'm a kyu, I should expect to spend even more time than that going over one game if I'm going to even begin to get some good out of it. Fujisawa Shuko said that he spends about 2 hours going over one game.

Then again, I also like skimming games, let's say in about 5 minutes. I don't think it's reasonable to expect to spend long amounts of time going over each move and each possible line of play for every game you go over. Plus, especially when I'm new to a player, I like to skim several of their games, so I can get a sense of what's going on, and that way I can find a game that interests me more more easily. Also, when you study more games more quickly, sometimes some things make more sense-- a line of play that you thought was particularly clever becomes obviously an old joseki when you see it a second or third time, for instance. Sometimes an odd move becomes perfectly clear when you see the follow-up 20 moves later, and if you go too slowly you might have forgotten about that original move by the time you get there. I also think it's easier to predict moves when you're more familiar with a person's style-- "Shuei liked responding to this attack by playing on the fifth line," or "Go Seigen likes to tenuki here," maybe. And it's harder to get used to moves like that unless you study more games more quickly rather than fewer more slowly.

Then again, when you skim a game in 5 or 10 minutes, unless your reading is lightning-fast and your instincts razor-sharp, you're probably not going to get all that much out of it. It makes more sense to look at just about every single move, only going faster for standard joseki, evaluating which groups are weak, and making rough counts from time to time. Usually I can do this for about 20 or 30 minutes before my concentration is shot, and if I'm not in koyose by then, I just skim the rest of the game. I've tried putting aside games and coming back to them, but usually unless I let them sit for a month or more, or unless I get a dan to review them for me, I've found that I usually can't read much more out of one game when I go back to it.

How do you study pro games?

first, studying pro games is not the most efficient way to improve, especially at kyu level. playing and solving tsumego is traditionally most praised

once you are aware of that, i will follow with my method of studying pro games

when you say you can skim game in 5 minutes, i guess you are replaying them online - i prefer studying offline, with printed kifu and at real goban. it helps me to stay concentrated and slows me down, at computer i tend to just start clicking through the game, not really watching the moves.

i enjoy slow reviews. in best case i look at the board, ask myself where would i play, then when i am really decided, i look at the diagram and if the moves are different (usually), i try to figure out why is pro's move better than mine. even in fast playthrough you should spend more time looking at the board than looking at the kifu

after 120-150 moves the game enters endgame, moves become less exciting and harder to find at the diagram, and i also sometimes lose my concentration, so i often skip endgames, focusing on the "important" part of the game

i think it is good to replay each game twice or three times during few days, until you remember basic played lines and the flow of the game.

i don't try to know every pro player, i mostly follow my favourite players, whose style i like, in my case it means Kato Masao and Go Seigen. i wonder, does it sometimes happen to you that you replay a game and you dislike some player at the first sight? of course not speaking about him personally but about his style. i remember that when i first replayed a game with Lee Changho or Sakata Eio, i thought something like "Come on, is this guy serious, can it be really played like that?"

this is my way of studying pro games. i don't know if it is good or bad, but i like it.
(note: i don't replay much games, i could say few games in few months, so on those occasions i don't regret spending more time with one game. for everyday study it would be way harder to follow my system)

_________________
Spilling gasoline feels good.

I might be wrong, but probably not.

This does not logically follow. It would probably take you days to look at the game and see everythign the professionals see, if you could to it at all, but this is irrelevant to 'get some some good out of it'. The moves that will help your game are not those that the professionals will themselves look at, in general, and there's no reason it should take you particularly long to find them.



I think there is value in skimming pro games. It just makes you think 'maybe I could try that', and hints to your brain that it should remember particular shapes.

OVerall, I think you should study them however you feel like it. If your goal is to improve, do whatever you enjoy the most. If your goal is to improve no matter what, play games and do a lot of tsumego. If your goal is to understand them in detail, spend hours trying to understand everything (laman's suggestions are all good), but as laman says this is not going to be the most efficient way to study until you're stronger.

maybe good way of studing pros is buying books with their games and comments. This is good start in beginning.

_________________
sorry for my english - im Marsian

What I've currently been doing is:

1) Read a commented game in Catching Scent of Victory.
2) Replay the game slowly on CGoban , trying to guess/remember each move & fooling around with variations. This takes about 2 hours.
3) Re-reading the game in the book.
4) Replay the game fast on Goscorer, to see how much I remember. This takes about 30 mins.

I find that this:

a) Is a lot of fun.
b) Helps me make sense of the commentary.
c) Makes me think about what I would do and why.
d) Feels a bit as if I were playing a nice looking game myself with a pro sitting behind my shoulder pointing out better lines of play.
e) Has helped me formulate plans in my own games.

_________________
Patience, grasshopper.

Shusai Meijin said that you could become a pro by replaying 1,000 pro games. (OC, he was assuming talent. ) I think that he was also thinking about spending a whole day's study on a single game. (This was a man who once thought for 8 hours on a single move in the endgame.) Fujisawa spent 2 hours replaying pro games. Takemiya told me he spends 15 minutes.

IMO, one of the best ways to go over pro games is to use something like the GoGoD software to guess the next play. That takes me around one hour to an hour and a half.

I would recommend studying some games from the late 19th century. Not that go hasn't progressed, or that individual players were better, but, because they took their time, their play in the last 100 moves or so was nearly perfect. Perfection is hard to beat.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Just adding to the anecdotes, a little.

The pupils in the Kitani school were expected to play over a pro game before breakfast each day. Kato Masao said he used to choose games with an early resignation.

Go Seigen famously got a crooked finger on both hands from holding open magazines while studying pro games.

There used to be, and maybe still is, a system at the Nihon Ki-in where pros could request copies of handwritten scores of games by other pros to be put in their pigeon holes. They were able to read through the game (or enough of it) just from that, no board needed.

whats a pigeonhole?

Pigeon Holes are a common 'mailbox' arrangement for an office mail system.. Try google, or, this picture found using google.

Of course, I already knew what I was looking for, so, perhaps that helped my googlefu.

Oh, interesting! Thanks for all the comments!

I wonder if Kato Masao liked games with early resignation because he liked games where one player got the overwhelming lead, and he liked to try to figure out why, or if he was just lazy. I know I would never be able to follow a 300+ move game with lots of ko fights before breakfast.

Laman: I do try to guess the general region of the next move, but I don't often go to the level of specific points. I just tried replaying a Shuei game, making specific predictions before some of the moves, and counting how many I got right. I found that I got a fair number of them right, about the same number as I got completely wrong-- but about half of my predictions were almost right. Like, I'd say, "That stone is threatening to connect underneath, so white has to descend," and White plays a keima attaching underneath the enemy stone rather than descending straight down. Or I'd say, "That stone has to move out right now," and he'd move out with a keima instead of a one-point jump. Then again, there's the predictions I got completely wrong-- "X group is the weakest group on the board, he should settle it now" and he plays a probe somewhere completely different. Do you pay more attention to the specifics of the little details you get wrong, or when you predicted something completely different?

Bill: As it happens, right now I'm studying a lot of Shuei games. (I consistently find myself impressed with his skill in the oyose-- all of his groups tend to expand a lot, and all of his opponents' tend to shrink.) Do you have any tips for getting the most out of studying perfect endgame?

At the moment, I tool I'm using is the "guess" feature on Just Play Go. For each top professional game it shows, I keep clicking until I get the right move. AFAICS, the score system is 1 point for right first time, then 0.5, then 0.3, 0.1 and then other 0.0x values for further guesses. At 100 moves, I check my score to see how I did. Each time along the way, like laman, I try to guess what I think is important and why. Sometimes, the pro plays exactly my move _after_ a tenuki to some other kikashi somewhere else, reminding me the value of sente and the whole board vigiliance pros have. Sometimes, a pro plays the same sort of move, but locally in a different position, and I can try to recognise the improved shape / efficiency of their move. Sometimes, a pro plays something completely different altogether, and I get to chastise my tunnel vision.

Regardless, I find it fun and instructive

For the most part, I score between 56% and 59%, with the exception of that recent Cho Chikun game I linked to where I got 41% - that game I simply couldn't feel for some reason.

- I think so too. Doing exhaustive search for all different possibilities and latent threats is, well, exhaustive. I'm interested in shapes professionals use. What direction they are playing? Sometimes when you see something unorthodox it's a good spot to stop and ponder why the pro played like that.

Studying pro games has only limited value because of concept blindness. That means the things you see on board don't hold same meaning to kyus than they do to dans. The stronger you are the more possibilities are obvious. Weak player might recognize move as an attack, but stronger player sees that the attack is about splitting and utilizing thickness to build territory on some other part of board. The problem is how to make that leap from weak player to strong one.