Memorizing Games of Professionals

[Apoah]

Yes, that is exactly how I look at it. Like learning a language or music. Go is very musical to me.

Most importantly, I really enjoy the process and it is more of something to pass the time than active studying. There is such beauty in pro games that can only be unlocked with many, many replays. At my level there are many confusing moves and it is very gratifying to to finally come to some kind of answer as to why a move was made. Even if it is wrong, that is the process of go, right? Constantly working on an assumed logic set and actively adjusting that logic as your understanding of the game grows?

On reflection, perhaps what I do isn't study as much as it is fun. Like someone who can sight read music sitting down to play a Beethoven sonata. They aren't "practicing piano" per say. I do not think I would suggest this to someone as a way to study.

Then again, if I ever hit 5d, perhaps other people will see it differently.

[Lady_Light]

Doesn't it take all the fun out of playing a game if your memorize all the moves? What kind of challenge would that be. Plus, wouldn't that take a lot of time to try and figure out what kind of moves your opponent would be making without actually having the opponent there? I am just confused.

[Dusk Eagle]

It would be absolutely impossible to memorize Go in the same way Tic-Tac-Toe can be memorized. It's even impossible for computers: there are more possible games of Go than atoms in the entire universe. No, Go is way to hard for any human to truly achieve perfection at it. This topic is about memorizing individual games between two professional players (players who, for the most part, have dedicated their lives to the game). The theory being that you can learn to emulate good moves if you see them often enough.

While I definitely believe studying/memorizing pro games has benefits, I personally think that other methods of study may be more efficient, such as doing tsumego. I also think you might plateau prematurely if your only method of study is pro games.

[LovroKlc]

It would be absolutely impossible to memorize Go in the same way Tic-Tac-Toe can be memorized. It's even impossible for computers: there are more possible games of Go than atoms in the entire universe. No, Go is way to hard for any human to truly achieve perfection at it. This topic is about memorizing individual games between two professional players (players who, for the most part, have dedicated their lives to the game). The theory being that you can learn to emulate good moves if you see them often enough.

While I definitely believe studying/memorizing pro games has benefits, I personally think that other methods of study may be more efficient, such as doing tsumego. I also think you might plateau prematurely if your only method of study is pro games.

I am not sure how useful is memorising... I mean, when I review a game in detail, I remember a good part of it, and that should be enough probably...

[Bantari]

It would be absolutely impossible to memorize Go in the same way Tic-Tac-Toe can be memorized. It's even impossible for computers: there are more possible games of Go than atoms in the entire universe. No, Go is way to hard for any human to truly achieve perfection at it. This topic is about memorizing individual games between two professional players (players who, for the most part, have dedicated their lives to the game). The theory being that you can learn to emulate good moves if you see them often enough.

While I definitely believe studying/memorizing pro games has benefits, I personally think that other methods of study may be more efficient, such as doing tsumego. I also think you might plateau prematurely if your only method of study is pro games.

I am too tired to do research right now, but I think the issue is not really how many 'possible' games there are, but how many 'feasible' games. I mean - even on move #1 there are 361 'possible' moves, but only what - 9 feasible moves? Or a a few more? In most position, I seldom consider more than 5 moves, and the same probably goes for the top players - although the 5 moves they consider might be different from the ones I see. Granted, there is a problem of narrowing it down, but still - most points in any position can be really dismissed out of hand even by beginners.

If this 'feasibility' factor holds even in some approximation throughout the game, the actual 'feasible' game tree of Go might be much narrower than people usually claim.

[serapis]

In my first two years of playing go I improved largely by memorizing professional games. I began around 8kyu and by the end of the first year I had reached 1dan. A friend of mine, when reaching a barrier of around 1kyu/1dan or so, took an even more extreme path. He stopped playing any go, only studying/memorizing games for about a year. When he was finished he was about 6dan.

Despite my apparent success with memorizing go games as well as my friend's, I can honestly say that I almost never recommend it. It's a quaint idea while you're thinking about it, but spending day after day, week after week replaying previously memorized games and trying to memorize new ones, is unbelievably difficult and anyone crazy enough to go through with it may start hating the game.

I think that, whatever you do to study, the most important thing is keep your mind open. Through the course of my teaching I've encountered students who weren't really trying to learn new things, rather they were seeking validation that what they knew was correct. Whereas, for me, when studying, it's always been essential to keep in mind that everything i know can be wrong.

[hyperpape]

@Bantari But if your goal were to play by just memorizing appropriate plays rather than using game knowledge, you'd have to know how to respond to any move your opponent makes.

In any case, if you gave the same board position to players ranging from 30 kyu to 9p, they'd propose more than 5 moves in any given position, even if each player considers 5 moves or less. So even the set of games that might actually be played by humans is bigger than what you're describing.

[karaklis]

In my first two years of playing go I improved largely by memorizing professional games. I began around 8kyu and by the end of the first year I had reached 1dan. A friend of mine, when reaching a barrier of around 1kyu/1dan or so, took an even more extreme path. He stopped playing any go, only studying/memorizing games for about a year. When he was finished he was about 6dan.

It may be quaint, but was obviously effectful (8k=>1d, 1k=>6d in a rather short time). I've been stuck at 7k for one year now, and by just playing/reviewing I cannot improve any more, so that I've almost stopped playing. Also a tough tsumego program (about 1400 tsumego from July to October 2009) didn't help much. It took me exactly from 8k to 7k. That's why I've been looking for alternatives. Did you or your friend have any preferences which games to memorize?

[serapis]

I cannot improve any more, so that I've almost stopped playing.

This is a bit worrisome. If you're at the point where you're frustrated enough to stop playing, I think it would be a bad idea to take up something even more frustrating.

That said, i primarily focused on games which emphasize the importance of understanding the basics. In other words; this game was a nice example of how to properly reduce influence or that game is a great example of the value of having 2 weak groups to chase etc.. Games that were complicated dog fighting from start to finish i mostly avoided because, at around 8kyu, i couldn't follow it anyway.

[Stable]

Where did you find the info on which game is a good example of x?

[karaklis]

This is a bit worrisome. If you're at the point where you're frustrated enough to stop playing, I think it would be a bad idea to take up something even more frustrating.

I think this is mainly a matter of motivation. I've always been the type that prefers studying to playing. I don't need the adrenaline rush every day, it's sufficient to have it every now and then at tournaments. Playing for me fulfills two things: to measure the skills and the social component: hanging around and exchanging knowledge with go playing people. That's why in my spare time when doing something go related, I do tsumego or read go books. The frustration comes in when you see that these activities yield no measurable results. That's why I've been looking for other ways to improve. Your posting woke me up a bit from my lethargy and revived my spirits. I know myself well enough to know that I need a clear path to the goal, because I had success this way with another activity, and that gives me motivation. If I just knew how many matches you memorized and which ones within the two years, it would be of great help.

[jts]

If this 'feasibility' factor holds even in some approximation throughout the game, the actual 'feasible' game tree of Go might be much narrower than people usually claim.

This begs the question. By "feasible" you seem to mean "most likely to lead to victory," but the original question was "can you memorize the game tree of Go so that you know which moves are most likely to lead to victory (as you can in Tic-Tac-Toe)?" No one would doubt that knowing which moves are, as you put it, "feasible" makes you stronger at go.

[tapir]

This is a bit worrisome. If you're at the point where you're frustrated enough to stop playing, I think it would be a bad idea to take up something even more frustrating.

I think this is mainly a matter of motivation. I've always been the type that prefers studying to playing. I don't need the adrenaline rush every day, it's sufficient to have it every now and then at tournaments. Playing for me fulfills two things: to measure the skills and the social component: hanging around and exchanging knowledge with go playing people. That's why in my spare time when doing something go related, I do tsumego or read go books. The frustration comes in when you see that these activities yield no measurable results. That's why I've been looking for other ways to improve. Your posting woke me up a bit from my lethargy and revived my spirits. I know myself well enough to know that I need a clear path to the goal, because I had success this way with another activity, and that gives me motivation. If I just knew how many matches you memorized and which ones within the two years, it would be of great help.

It would help immensely as well if you could enjoy playing. (Both your strength and your well-being.) I mean you would not stop talking a foreign language you learn, while still studying grammar books daily, would you? The other way round looks more healthy to me, (especially when frustrated already), almost nobody can improve without practice.

[daal]

Karaklis - I'd like to recommend Catching Scent of Victory by O Rissei. I got it for myself for Christmas, and have enjoyed it immensely. In the book, O Rissei reviews 20 of his own games pointing out the stragegies he employs (largely meaning his thoughts on the fuseki), and his descriptions of critical moments, sometimes asking the reader what he would do in that situation. I found that this type of presentation made me want to replay the games.

I am sure that there are other comparable books, but in any case, I found it a good deal more fun to try to memorize a commented game than one that I more or less randomly pulled out of the database. It is quite interesting to recall O Rissei's thoughts while replaying the games, and memorizing did not at all seem like a frustrating task.

Also it's worth noting that most of the games happen to be on GoGoD, and one can replay them using the "simplified" Go Scorer program included on the GoGoD CD. BTW, is a full version of the program available?

[TMark]

We're now talking a bit of pre-history. The original GoScorer was a standalone program John wrote for DOS, for use with the GoScribe files (ishi format) that we started with, before we switched to sgf files. I don't know if John has a backup version; I certainly don't. Again, John probably wrote that intro back in the 90s, when he put the GoScorer module into GoGoD95 and then forgot all about it. I know people who like using it for testing their study, but we haven't had too much feedback about it and so we tend to concentrate on things that give us lots of feedback, the extent of the database and Go Seigen commented games.

Best wishes.