Enter general discussion about replaying those kind of games.

I'm thinking about doing it on my real board. I never use it much besides twice a week at club, so why not? It'll be interesting to try reading stuff out to see what is best played first, and to put myself in the shoes of those far better than me. Any of you do this? Do you propose it better replaying on sgf file rather than real board so as to review all the variations? Other suggestions?

_________________
TucsonGo.org - Play Go in Tucson
We meet twice a week, visit us!

One reason I could give you is aesthetics. It's a beautiful game played out on a real board. I have heard that replaying them on a board gives additional benefits because you take longer and automatically review the game more diligently. I have found it just takes longer for me, and I still don't understand much of it. But I enjoy it a lot at times. It's time better spent then watching television.

I think it's a lot easier to remember a sequence (including a whole game) when you play it out on the board. When you simply click through an .sgf file, you don't have to pay too much attention to where the stones are actually being placed. You don't actually have to engage your mind at all in order to click the forward button in your go client. But when you use a physical board, you are forced to pay attention to each and every move. I find that this makes the moves stick much better in your head.

In fact, I always make a point of playing out pro games that I want to study on my board, even if I have an .sgf or video commentary of the game with me. I also always use my board when I am studying joseki, as I know I will quickly forget what I have studied if I do not.

_________________
We don't know who we are; we don't know where we are.
Each of us woke up one moment and here we were in the darkness.
We're nameless things with no memory; no knowledge of what went before,
No understanding of what is now, no knowledge of what will be.

The recent Kong Jie-Lee Sedol game interested me, so I decided to try replaying it, seeing how easy it would be to memorize.

I unfortunately do not have a board that I can easily replay it on, so I do use the SGF. However, what I do is actually click the next moves on the board in my SGF viewer (instead of using the "Next" button). If I have no clue where the next move is, then I step back and study the sequence to see if I can understand (in a loose way) the goals of each player.

Two interesting points have come up in the 20 minutes or so I took last night to try this:

1) The majority of the moves make at least some sense to me. In this game at least, the pros stick with a lot of natural feeling moves. The real beauty is when you step back and try to fit those natural moves into the overall board position (which can be much more complicated).

2) There are certain moves that stand out particularly BECAUSE I do not understand them. In this game, Lee Sedol's K15 stone is one of those moves. I had to step back and think for a bit before I realized that it made sense in relation to Kong Jie's K17 and J16 stones. It's a move played on a shape point. It was confusing because initially I had been trying to relate the move to Lee's M17 stone and could not see the benefit of an Elephant Jump like that (indeed, a short time later Kong Jie did separate the two stones by hitting the Elephant's Eye).

I guess the point I'm trying to make is: when I first thought about this kind of study (when I was just getting to SDK ranks) I did not see the benefit, particularly because all I had were SGF files and I simply clicked through the SGF using the "Next" button. I tried something different when studying this game, and it makes a lot more sense to me now why it could be beneficial for studying purposes to do something like this.

Playing it out on a board definitely forces you to think about the moves in a more tangible way, but it isn't impossible to do so using an SGF Viewer so long as you make the effort to learn the game in a more interactive way.

I usually use a real board for playing through pro games. Partly because, as Hushfield mentioned, there is a nice aesthetic to it. And it's the only time I get to use my floor Goban!

I have my own particular method of studying games, which I've mentioned in threads in the past. But the bottom line is that I think that taking the extra time involved in laying out the stones by hand adds to the process of imprinting upon our pattern memory. Maybe it's just that it's slower, so one's brain has a longer awareness of each move.

Having said that, I will admit one can review a lot more games with a computer!

The problem that I have as a mid-sdk is that there are many moves that make sense. If you want to know the difference between your crude moves and the professional move, it would be very helpful to have comments that are understandable for you. When these are missing it's a bit like poking in the dark.

This is a lot harder than I expected when it comes to the end game moves (seems almost randomly around the board), but on the other hand it is a lot easier than I expect for the opening and mid-game. I'm also finding it helpful just looking at and sometimes playing out what would happen if the move that was played hadn't been played. Sure I'll probably never be able to fully understand these kinds of moves, but there are plenty of insights.

_________________
TucsonGo.org - Play Go in Tucson
We meet twice a week, visit us!

It's a well-known thing that how fast you can play through a game on a real board using only one or two diagrams is a rough measure of your go strength. Here's an exercise you can do when playing out a game. After you have played a move, look at the board and try to figure out at least what area of the board the next move will be in, then look for it there in the diagram. If you find the move in that area it validates your sense of direction. If you don't find it there then you have something to think about, such as why was the actual move in the game bigger or more urgent, or was this a miai situation, etc.

I've never been convinced by this. Doing lots of this sort of work speeds you up no end. I've known EGF 4 dans take 20 minutes going through a pro game with 3 diagrams (one per set of 100 moves). It normally takes me about 10 if I'm just playing it out and not analysing, and I understand T Mark Hall, with his work for GoGoD, generally takes 4-5 minutes per game.

yeah, some people just do the opening and mid-game moves and skip the endgame.

as I've posted many times before, replaying pro games is one of my favorite go activities, and feel its helped improve my level of play

there was a thread on GoDiscussions about this very thing. I don't think it was proven one way or the other.

TMark posted an image of a game record, and many of us timed ourselves in transcribing to sgf. If I recall, TMark's time was ~20 mins. And in spite my being just barely a SDK, I was able to almost match his time (but mostly because I also had some experience doing the same thing already).

Perhaps it's relative to the individual. Just imagine how much stronger you need to be to do it in 1 min

_________________
FusekiLibrary, an opening library.
SGF converter tools: Wbaduk NGF to SGF | 440 go problems | Fuseki made easy | Tesuji made easy | Elementary training & Dan level testing | Dan Tutor Shortcut To Dan

I'll certainly need more nimble fingers

I think the idea is that the stronger you are the more you know where the next move will be so that you don't have to search large areas of the board. If you have a 200+ move game on one diagram and you are not a fairly skilled player you will waste a lot of time scanning the whole board, probably missing the move a few times and having to do it over again.

I've recently started doing almost exactly what Marcus describes.

I replay Shuei's games. First, I look through it using next, to see what events happen. Then I go through it again a bit later playing the moves from memory, in the SGF, so I can see when I make a mistake. My memory is very poor though, so I always forget, at which point I evaluate the board (groups' health, big areas ...) and read - and find the right move.

My direction, positional judgement and instinct are bluntly, bad and I tend to invade/reduce inelegantly. Shuei is amazing at all of those things, so by re-finding the right moves for him, I hope to improve myself. I also really want to capture the feeling of his reductions in my own games.

Already, I feel my fuseki has become stronger and I am much more confident with 3-4s and reduction plays (I play sketches of his techniques in my games, get great results).

_________________
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.

You can memorize moves best, when they make sense to you. So if you are stuck at a special place this is a mark that you've problems with this move and should study it.

With that attitude you'll go very far.

_________________
FusekiLibrary, an opening library.
SGF converter tools: Wbaduk NGF to SGF | 440 go problems | Fuseki made easy | Tesuji made easy | Elementary training & Dan level testing | Dan Tutor Shortcut To Dan

I could comment on this. I've done some research into methods of teaching and learning and (without being able to say more specifically) you use different areas of the brain to process information visually, and physically. I mean, that a pure visual form like clicking in a .SGF does work well in conjunction with a physical method such as placing stones on a real board because you're learning in different ways, and they reinforce each other. The brain processes an abstract top down picture of a goban very differently than the physical moves of placing stones on a goban, with the touch and sound associated with that as well. In short, if you want to learn well from replaying pro games, it's best to play them out on a real board as your basic practice. But any further, it does help to replay them in other ways as well.

Replaying games on a real board is great, but I have found that SmartGo Pro iPhone's "Guess the move" feature is a lot of fun. While replaying any game you can select "guess" mode and you have to click where you think the move is. If you are close but not correct, SmartGo shows a green cloud of points around where you clicked, indicating the region that the move is in. As you continue making incorrect guesses, the cloud gets smaller, until only one possible point is left.

If your guess is in the wrong area of the board altogether, a red cloud shows you where the move "isn't". If you click a point and it turns blue, it means that the move was played, but not yet.

It's very satisfying when you get the move right first time in a pro game. Also, if you run through the game a few times, you become much better at recalling it correctly than if you were just browsing through it with a standard SGF viewer: it forces you to think about the sense of what is happening on the board.

That sounds prtty cool. I might have to upgrade the version I have.