Life In 19x19

Here's a commentary of Iyama Yuta's winning game against Gu Li in the Bosai Cup.

Younggil praises Iyama Yuta's play, but shows that there were chances missed by both players, due to time pressure.

It's a very interesting review.

An Younggil's commentary is both insightful and enjoyable. Commentary such as this makes such a game much more accessible. An Younggil seems particularly good at recognizing situations in a game where a weaker player might easily make an incorrect evaluation, such as around move 33 or the variation of move 16. He also tosses out a few good tips, such as at move 46 and 61, and peppers up the review by showing moves that exemplify a player's style. Great stuff! Thanks for sharing.

I was impressed my An's comment that black overlooked move 74 when he replied to white's move 50. As if An just takes for granted that a pro should see everything 24 moves deep without error.

If I understand the set-up correctly, David of gogameguru in Australia has a pro more or less at his right-hand - a unique position of access.

There are many questions I'd like to put in that position, but there is one in this game that is fairly typical of pro commentaries. It's possibly rude to ask, but David may be willing to pass on this query, and if so the answer would be very enlightening.

The commentary says, of White 14, "This is one of Iyama's favourite moves. A or B is more common instead."

Two problems (for me) with this.

1. Out of 176 games in the GoGoD database with the position as in the lower half of the board, B is the commonest (51 cases) but A is rare (3 cases and 11th most common). The second commonest in database terms is the move shown by the variation number 1.

2. Of these 176 cases, Iyama has played this move just once - hardly a favourite (he has also only faced this Black position once, though).

I can imagine an explanation of (2). The phrase in the Oriental mind starts as "Iyama has played this move in the past." That becomes in what might might be called meta-English "Iyama likes this move", which ends up in polished but flawed English as "This is one of Iyama's favourite moves". Another possible explanation doesn't seem to fit: that the reference is not to this position but to the high distant White approach in general. If that's the case, there are just 7 examples by Iyama compared to 63 high close approaches.

However, I have no easy explanation of (1). The explanation of both might be that An has access to an even bigger database than GoGoD - but a startlingly different one?

Dare we ask?

A comment that jumped out at me was at move 102, An says that Iyama's style is sharp in a way that's atypical of Japanese professionals. I hadn't heard that before, and it makes me think of John's post or two on "souba" viewtopic.php?p=63031#p63031.

Is that observation about Iyama a familiar one?

John,

Younggil is my teacher and my friend. The concept of our site started out as an idea to help Younggil get more students. Younggil has plenty of students now and the site has evolved into something else.

There are now three of us working on the site and it's really an experiment in promoting Go, which we're just at the beginning of. As people who visit the site regularly will see, there's still no real information for beginners. We're still at the 'preaching to the converted stage'. This is despite the fact that I've been working on drafts for weeks now (writing clearly for beginners is surprisingly hard - maybe I just need to stop trying to perfect the articles and start publishing them). Anyway, I hope the experiment works and that we can introduce Go to more people.

To come back to your question, I've asked Younggil about database related things before when he has made similar statements. I can try my best to tell you what I think his general response would be. If you're not satisfied with it, just let me know and I'll see what he says. Though I've noticed that pros tend to be puzzled by the relevance of such questions sometimes.

At one stage, a couple of years ago, a friend and I did show Younggil a database search using Kombilo. He was intrigued by it, though seemed to regard it more as an entertaining diversion than a serious study tool. He said he didn't think many pros used such tools to study Go, which surprised me because I thought I had heard someone (you or T Mark? - maybe it was in The Go Companion?) say that they did.

Anyway, I think Younggil's view is that pros already have this kind of information in their head. I find this plausible and I imagine you might too. He also puts a high value on trusting your intuition and playing your own style. I imagine you've met enough pros to have heard similar statements before, though I'd be interested to know what other pros have said about GoGoD,

In one sense he does have a different database, because he has the experience of the thousands of games played as a yeongusaeng and later as a pro, studying and training with other pros. Many of these games wouldn't have been recorded and won't be in GoGoD. Likewise, Younggil won't have replayed all the games in GoGoD. This is the conclusion I came to. There must be so many games played by pros that are not recorded.

With regards to your question about the two space high approach. I have questioned similar things before and so have others. For example, see Uberdude's comment here: http://gogameguru.com/commentary-gu-li- ... omment-873

In this case I didn't actually question the statement about Iyama's move, because that is also my impression of Iyama Yuta's style from playing through his games. I think you're right that that's maybe the key to interpreting "This is one of Iyama's favourite moves" - That is Younggil's impression of Iyama's style. I'm pretty sure Younggil wouldn't have consulted a database to write the article. Did you look at the recency of the different moves?

With regards to editing, I'm not completely sure that this is what you were asking, but we try to edit Younggil's articles as little as possible. We make very minor edits to fix obvious grammatical and spelling errors, but we try to keep his individual 'voice' intact. Jing or I will usually send the corrections back to him to help him in his study of English. On a side note, his English has actually improved incredibly quickly since he arrived in Australia - he's a very smart guy and he has many good people helping him.

Anyway, "This is one of Iyama's favourite moves" is exactly what Younggil wrote.

I hope this goes some way towards answering your question. If it doesn't please let me know. Either way I'd be interested in hearing your (and other people's) thoughts about this.

David, Actually I would like to hear An's reaction to the specific point about Iyama, because I don't think he has access to games by Iyama that we are lacking, and I don't think he's ever studied with him. Nevertheless, I do expect a language-based explanation similar to the one I outlined (and I'm not trying politely to give him a get-out; I base this on actual experience).

He said he didn't think many pros used such tools to study Go, which surprised me because I thought I had heard someone (you or T Mark? - maybe it was in The Go Companion?) say that they did.

I don't think we ever said anything like that. In fact, in Go Companion I said, "Now it should be pointed out that this database approach to go seems to be anathema to pros we have talked to in Japan, China and Korea. Indeed, many of them seem to be reluctant even to look at games digitally. They prefer to look at a written record, register the opening moves and then do the rest in the heads."

I've heard of only a few pros who use databases. Specifically, I can only recall one, and that is Sakai Hideyuki. An article about him made it plain he was exceptional. That may have been because of his scientific training as a doctor.

As with the book study vs. repetitive workouts discussion elsewhere on L19, I think we have to accept that pros often operate in quite different (and mysterious) ways from us amateurs.

gogameguru wrote:In one sense he does have a different database, because he has the experience of the thousands of games played as a yeongusaeng and later as a pro, studying and training with other pros. Many of these games wouldn't have been recorded and won't be in GoGoD. Likewise, Younggil won't have replayed all the games in GoGoD. This is the conclusion I came to. There must be so many games played by pros that are not recorded.

Agreed. Also, it is likely that games played online where the account is known (or suspected) to belong to a professional are also considered.

John Fairbairn wrote:...
1. Out of 176 games in the GoGoD database with the position as in the lower half of the board, B is the commonest (51 cases) but A is rare (3 cases and 11th most common). The second commonest in database terms is the move shown by the variation number 1.

Second commonest is variation 1 (F4)? White wouldn't play that, but that is black's move - "black to move" shouldn't be included to your search, no? Probably this won't much change your question though, as you won't get more instances of G3 obviously.

F3 should be the most common move for white in that shape. I think/guess/kinda-remember-somewhere F3 is the older Japanese joseki while G3 is a recent Korean move. What year, which players are those 3 cases of G3?

Actually, you can get more G3 moves, because the bottom doesn't need to be totally same. You should search just with the shape at the lower-left, and manually check if the total situation relates. Because for example white can kakari the right corner (and black answers) before G3 or F3.

John,

Younggil said you should come and ask your question on his article, so that he can respond to you directly. Here's the link to the commented game again.

In retrospect, I should have pointed this out earlier. Part of the point of how we've setup Go Game Guru is that you can ask questions, so it's not just me who has 'access' - to use John's word. You do too.

Younggil says he will answer the questions about his articles whenever he has time and he knows there are a couple of people who are still waiting for answers (e.g. Uberdude).

If anyone has other questions about Go, you can send them in to Go Game Q&A, which you get to participate in through the free newsletter. Of course, we use a reputable email service and we will never share your email address with anyone. Unfortunately we won't have time to answer every question we receive, but we will answer a few questions each month.

Finally, I know I don't need to tell most of the L19x19 folk this, but do be respectful please. Some of the comments on the Kim Jiseok vs Gu Li game weren't really helpful to anyone .

John Fairbairn wrote:David, Actually I would like to hear An's reaction to the specific point about Iyama, because I don't think he has access to games by Iyama that we are lacking, and I don't think he's ever studied with him.

John, Younggil is waiting for your question. Do you still want to ask it or have you lost interest?

(We're assuming you weren't the 'John' who left a comment on that article already)

David

John, Younggil is waiting for your question. Do you still want to ask it or have you lost interest?

I'm afraid I skipped it. I was only interested as part of a wider discussion here. A private answer to such a question doesn't seem to be of much use to anybody. To me it was not a "how can I get stronger" question but more of a "how can we all view this sort of thing " one. Deep down, my question was simply about pro attitudes to databases, which I think is of widespread interest.

Also, even though I believe it's unlikely, I also sensed a possibility that An was being defensive by taking that tack, and I didn't want to risk putting him in an awkward position.

That's OK, I didn't explain clearly. Sorry about that.

You can ask public questions on our site on any article, much like this forum. Younggil wanted you to come to his site because he thinks that if he spends time answering a question he wants the answer to available to everyone who visits the site. In a way you're both thinking along the same lines of not considering a private answer as useful, but from different perspectives.

In any case, if you still want to ask your question you're welcome to. I do appreciate that you didn't want to ask an awkward question, so thanks for being considerate.

We'd like everyone to feel free to ask questions on our articles any time they like.

David

Here's the reply: http://gogameguru.com/commentary-gu-li- ... -215145317

Thanks. What I took from the reply is first that the reference for White 14 was to something different from what I inferred. As I said before, I suspected a language problem. Second, An, like most pros, is clearly sniffy about databases. Quite right, too, wnen you are a pro! But I may have given him pause for thought