Life In 19x19

Hi everyone! I'm back!
戻ってきたぜ！

Some of you veteran members may remember me. I've participated in the forums in the past but went through long intermissions of involvement. I am a Go teacher and have participated in various sections including the Malkovich-style games.

I have a new blog where I'll be featuring a lot of original Go content, some of which you may find innovative (I'm just getting started with it, but I'm hoping to make a big splash!). I am writing up extremely thorough (hopefully more thorough than ever before in English Go content) Joseki tutorials starting with the most basic variations, and coining new names for each Joseki in order to bestow names upon each and every one of them once and for all. The names are based on visualizations and structure types.


You can find my blog here: http://ygami.blogspot.ca


The original purpose for my blog was to serve as a place for me to host my stories (fantasy, romance etc.) in order to receive critique, develop a readership, and so on. But the blog evolved/mutated into my "everything" blog, which will reflect the style of the "Yukitube" Youtube channel I'm working towards. (I'm gunning for 'Yukigami' to eventually become my universal online alias/pen name - Oh, and this time, my blog won't die off after a few months or a year. This one is here to stay!)


I have rewritten my profile and signature, changed my avatar, and completely revamped my L19² account.

I am still offering Go lessons, and my teaching has improved a lot since the last time I posted lesson offers here.

I've joined the CGA newsletter team as a writer and editor, and come summer you can all expect a huge amount of Go content to appear on my blog.


As a further memory-jogger, some old blogs/social web spaces of mine were Drifting Sand, Duelist Domain, and Dreams of a Go Player, and some of my many past KGS accounts, including some very old ones, were:

Invader
gilgil
CosmicGG
LacusClyne
Alakazam
ChinOp
TheRaven
TouyaAkira

I've had so many that I can't even remember.
To be perfectly honest, I do not like KGS. I find ratings always end up either too fickle (after many games) or too flimsy, and unless I happen to automatch at the perfect time, it's difficult to find an opponent - never mind one with similar game settings in mind. Mind you, automatch was a wonderful implementation for KGS.
I am not complaining, but perhaps this is my way of partially justifying my past of constant account-hopping. I can't say I am any better now in that respect, either.

With Tygem on the other hand, it takes a while to rank up but it's a clear system based on the W/L record of your last 20 games that does not share the problems of KGS. There are more strong players there, and quite a lot more players online in general; getting a game on Tygem takes seconds.

Tygem does, however, have the problem of requiring too many games (in my opinion) for each rank-up. Even if you are overwhelmingly better than the opponents you are stuck playing when you sign up - for example, a 9d signing up with the maximum starting rank of 5d - it takes around 15 wins to rank up, or something to that extent. If you lose 4-5 games, then that's 20 games to rank up once, and if you lose more than that, it takes a lot longer. Once you rank up or down, the slate is clean again, which means the entire process repeats. There is a bonus for X straight wins at which point you can achieve a 2-rank jump, but that requires something like no losses in 15-20 games, and this can be pretty daunting a task to pull off because it is difficult to maintain concentration for so many games in a row without the weaker opponent pulling one over on you while you are wondering in your half-sleeping state why they haven't resigned 100 moves ago.

I find that WBaduk has the best rating system. Their system is based on points and seems fairly convoluted. But based on my recent experiences there, it takes only 6-8 wins to rank up if you've lost 0-3 times. I've been ranking up as long as I won at least 2/3, or something like that. It definitely feels faster. Even if the W?L ratio requirement for WBaduk and Tygem is the same, which it probably isn't, WBaduk doesn't require such a high minimum game-count.

The problem with WBaduk for me though is that it constantly gives me connection problems, and my games get repeatedly adjourned, sometimes leading to their complete disappearance even if I return after a few seconds.

KGS has the best teaching tools though, not to mention the best social community. None of the servers are perfect, and I can't really complain. But these opinionated explanations shed light on why I have a long list of KGS accounts but very few accounts on other servers. Overall, Tygem is my favourite server. It already defeats KGS in my opinion, or at least does for me, but also has a bunch of fun gimmicks such as betting. I can forgive the fact that lots of players on Tygem's Korea servers try to con wins by spamming draw requests, counting requests, and private messages.


So that's my new intro! Thank you for taking the time to read my intro and Go server rants.

Nice to meet you, nice to reunite with the new GD, look forward to my posts and self-written Go content,
and please check out my blog!

If you are looking for a teacher, please hit me up! I am very confident that I can assist you in powering up.

I remember.

Welcome back.

Your blog looks good. I'll definitely be following it for a while at least.

P.s. If you can keep up the level of detail I see in Separated in the Subway I think it will be an amazing resource.

Welcome back! I'm an intermittent member here myself, so...

I remember plenty of your content in the past. Glad to see you back and look forward to checking out what you have in store for us.

You (and everybody else) is what makes this a place I keep coming back to!

I do recall seeing you a while back. Welcome back!

Splatted wrote:Your blog looks good. I'll definitely be following it for a while at least.

P.s. If you can keep up the level of detail I see in Separated in the Subway I think it will be an amazing resource.

Thanks Splatted! I really want to make my blog interesting, user-friendly, and overall a place people want to be.
I'm glad you liked my tutorial. I plan to improve the concision and structure, making them easier to get through and more well-transitioned. Also, I will talk more about what kind of situations are best to use them in.

The version on my blog needs to be updated, the one in CGA newsletter is edited. But I'll update the blog one soon. The next one will come in early April.

Critique too will be fully accepted and appreciated!


Dante31 and schultz: Thanks! I will be sure to post some interesting stuff!

Alakazam wrote:

Splatted wrote:Your blog looks good. I'll definitely be following it for a while at least.

P.s. If you can keep up the level of detail I see in Separated in the Subway I think it will be an amazing resource.

Thanks Splatted! I really want to make my blog interesting, user-friendly, and overall a place people want to be.
I'm glad you liked my tutorial. I plan to improve the concision and structure, making them easier to get through and more well-transitioned. Also, I will talk more about what kind of situations are best to use them in.

The version on my blog needs to be updated, the one in CGA newsletter is edited. But I'll update the blog one soon. The next one will come in early April.

Critique too will be fully accepted and appreciated!


Dante31 and schultz: Thanks! I will be sure to post some interesting stuff!

Nice detail, but your punishment for the mistaken hane is wrong. There are quite a few threads on L19 where you can learn about this joseki mistake.

Uberdude wrote:Nice detail, but your punishment for the mistaken hane is wrong. There are quite a few threads on L19 where you can learn about this joseki mistake.

Looks like the correct punishment to me. I know there are other options for punishment but this one has always served me well. What do you recommend over this?

moyoaji wrote:What do you recommend over this?

Recent thread about this with further links in it: viewtopic.php?t=10053

Iirc the consensus is for the "punisher" (xD) to hane, connect and clamp.

leichtloeslich wrote:

moyoaji wrote:What do you recommend over this?

Recent thread about this with further links in it: viewtopic.php?t=10053

Iirc the consensus is for the "punisher" (xD) to hane, connect and clamp.

Following through this, Bill's post here seems to show the best results black can get out of the hane following from the two candidate moves.

It would seem to me that either result is good for white and which is better would seem to depend on the whole board position. If white needs access to the right then the hane and clamp is good. If white would like access to the bottom then the descend and connect would be good. Perhaps these can each be treated as psuedo-joseki? You can choose the one that better fits your intentions.

However, considering the fact that most players who play the hane are not going to connect the cut you can also bet on the superior result of the descend and cut variation and, if you lose that gamble you still get a fine result.

This

Click Here To Show Diagram Code

[go]$$Bc
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . a . 3 1 . . |
$$ . . . . . X O . . |
$$ . X . O . X O . . |
$$ . . . . 4 2 . . . |
$$ . . . . . . . . . |
$$ ------------------[/go]

is better for white than:

Click Here To Show Diagram Code

[go]$$Bc
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . a . 3 1 . . |
$$ . . . . . X O . . |
$$ . X . O . X O . . |
$$ . . . . . 4 2 . . |
$$ . . . . . . . . . |
$$ ------------------[/go]

As black can't block the centre so thickly and white can more safely jump out to a.

(Simple tewari says in the first diagram would you rather play 4 as shown or to right to revert to 2nd diagram).

You shouldn't play in a way that assumes your opponent will make further mistakes after they have made a mistake.

@Uberdude: I'm not good enough to judge your assertion but it's probably worth mentioning that I have seen in the second diagram described as the correct move by multiple pros*. Perhaps it depends on which is more important: the right side or the centre? Black playing on the right would suggest that that is important to him at least.


*Unfortunately I can't remember which ones but they were videos posted online so I'm sure I'm not the only one that's seen them.

@Splatted: I think this is one case where pro opinion has shifted about a decade a ago. It, as all other things in Go, is of course subject to considerations unique to each game (e.g. if black's direction is horrible after letting white connect then you would want to play the descent rather than the hane connect clamp/cut sequence). However, I think the idea is that in most cases black letting white connect underneath after the descent is not a bad result for black whereas the hane is better than descent if black allows the connection and clearly better than joseki if black blocks and white clamps (i.e. clearest punishment).

Edit: actually, probably more than a decade ago, information filters down pretty slowly in the west :p

Edit 2: Would it make sense to split this into another thread to avoid spamming Alakazam's post?

One advantage of my choice though:

If you do the hane as white, they can just prevent your connection and after you connect and then clamp, they can revert to normal Joseki but with a bad push on their part. That push exchange is indeed terrible; it's more terrible than most kyu-level players realize. But the simple descend, on the other hand, guarantees a result that is definitely bad for black. Black always had the chance to block that side (the upper part, the right side of the board) but if you compare to the wall Joseki with the double knight move shape, this one is worse for black.

I think it indeed depends on the situation. I think at the very least, the descent is very simple to carry out for low-level players as the clamp leads to great complication.

At some point I will come to a new or same conclusion about this after I take a look at it, but I don't really have time right now.

I don't think the descend is "incorrect." If anything, the descent is not only simpler and easier to carry out (and depending on the board circumstances, could be better anyways, just like the clamp variation could be better), but the clamp variation is heavily affected by stones that may exist nearby, whereas the descend version is not; its effects on black occur very locally and are less easily affected by nearby stones. With the addition of the fact that black, having played the hane, is less likely to stop making mistakes, I think the descend is ideal for low=level players in addition to not being clearly worse for strong players.

I agree that we should not assume our opponents will make mistakes, but I think that for the following reasons, it is hard to say the descend is worse, and conversely may be easier actually to say that it's better overall:

-easy to carry out
-unaffected by nearby stones
-clearly worse for black from a local standpoint, as all we need to do is compare to the normal Joseki where black purposely blocks the other way and gets that double knight move wall
-not worse, since each one provides a result that is better depending on directional values
-if you just look at the result of the descend version, and assume black cuts his losses by connecting, we can see that two of black's stones are doing almost nothing, and his pincer move is also in a pretty poor place at the end. His potential side size, compared to the regular double knight move wall joseki, is decreased.
-opinions are (apparently) mixed regarding which punishment is the best
-most players on this forum are below your level or mine, and when it comes down to it, people have different philosophies as to how to think about the way we play. While you or possibly I may aim for the most "high level" thinking and play in a deep kind of righteous sense, others may embrace their eternal amateurness (below 7d, or perhaps just below pro power) and thus care more about what wins games for them without risk of disadvantage vs. what is more...philosophically correct, or "more pro", or shows more integrity or something.

I think that overall, for our purposes the descend is better on average (meaning it can be obviously worse sometimes, based on the board).

When it comes down to it... READ OUT THE MOVES! If you can't read it, just do the descend or whatever you like. If you CAN read it, maybe still do the descend, but maybe do the hane-connect-clamp.

Eventually, players realize that the entire game is about reading. It's not about Joseki or "correct" or this and that. Whatever wins you the game is "correct", that's what I think.

Go is very psychological, and as much as some of us (including me in the past) wish to play in a sort of as-close-to-godly-as-we-can-manage way that separates the events on the board from the psychology and other factors, we can't; they're a part of the game. "Skill" is not even the right word for describing Go strength; it should be conceptualized as "winning power", at least for non-useless purposes anyway. No matter what we may think of or what excuses or reasoning we may come up with, the one who wins most is better. If, for example, poor health or tendency to blunder(which by the way, I used to experience lot, but it totally stopped once I got stronger. Some hope for you reading this??) or failure to play under time pressure causes us to lose, well - that is our "skill"! Our winning power results from mix of things. I learned this the hard way and came to terms with it a few years ago when I was training in Korea.

From that line of thinking: if we are faced with two variations here that are both correct in that they are better for white than for black, but one has a lot of practical advantages for winning and the other is "more pro" (which we don't even know yet, as apparently the consensus is divided??), which one would you pick?
Personally, unless it hinders my improvement or something, I would choose to WIN!

Alakazam wrote:One advantage of my choice though:

If you do the hane as white, they can just prevent your connection and after you connect and then clamp, they can revert to normal Joseki but with a bad push on their part. That push exchange is indeed terrible; it's more terrible than most kyu-level players realize. But the simple descend, on the other hand, guarantees a result that is definitely bad for black. Black always had the chance to block that side (the upper part, the right side of the board) but if you compare to the wall Joseki with the double knight move shape, this one is worse for black.

Are you sure those last two sentences are correct?

Click Here To Show Diagram Code

[go]$$Bc
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . 3 1 . . |
$$ . . . 7 . X O . . |
$$ . X . O 5 X O . . |
$$ . . . a 6 4 2 . . |
$$ . . . . . . . . . |
$$ ------------------[/go]

Is this result definitely bad for black? That's not clear to me yet. If it is bad is it worse than making the bad push exchange which is definitely bad? White may now tenuki which leaves black with a big follow-up of cutting at a, or perhaps continue with the rather crude walking into the double hane or more likely this:

Click Here To Show Diagram Code

[go]$$Wc
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . 4 . . X X . . |
$$ . . 1 X . X O . . |
$$ . X 2 O X X O . . |
$$ . . . 3 O O O . . |
$$ . . . . . . . . . |
$$ ------------------[/go]

White now has the aji of a ladder breaker to pull out , but if white plays bad moves as ladder breakers black could answer them he can fight or sacrifice if white pulls out 1. Also could be extend to leave different aji. If we compare that shape to the normal wall joseki of blocking the right (both are black gote):

Click Here To Show Diagram Code

[go]$$Bc
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . a . . . . |
$$ . d 5 . . . . . . |
$$ . . b . 3 X 1 . . |
$$ . X . O . 2 O . . |
$$ . c . 4 . . . . . |
$$ . . . . . . . . . |
$$ ------------------[/go]

in the former black is much thicker and all of the interesting aji points marked in the second diagram are gone. It's not uncommon for the black wall in the second diagram to come under attack or die if black overplays in later fighting on the right, that's much less likely in the first. In the first black's right side is one line smaller, which could mean less territory but could also mean less space for an invasion to make eyes so more secure, and white's corner is a line bigger. Assuming simple yose the difference here is 4 points gain for white. However, white's yose on the lower side at c is now less good, only being a monkey jump. Another consideration is in the second diagram black has more ko threats. It's hard to combine and judge all these pros and cons but my feeling so far is the first diagram is better for black than the second.

Alakazam wrote:I think it indeed depends on the situation. I think at the very least, the descent is very simple to carry out for low-level players as the clamp leads to great complication.

For sure the clamp and black resisting with descent by fighting leads to complication, and as you say later the presence of nearby (or not even particularly nearby) stones could alter this fight from being favourable for white to even or better for black. I don't actually know any sequences after the clamp, just that it is supposed to be good for white so would play good shapes/read if it came up (it'd be nice if there was a quick local success option rather than a protracted advantageous fight where it's quite easy to make a mistake bigger than black's mistake of getting into the bad fight).

Yes 10k players might block the descent and then you cut and get a great result, but to assume they block rather than connect means you have to settle for the slacker connection of the 2nd diagram in my first post. I think the tewari idea of white to play: a or b as below is very important (and simple) and a widely applicable general lesson.

Click Here To Show Diagram Code

[go]$$Bc
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . 3 1 . . |
$$ . . . . . X O . . |
$$ . X . O . X O . . |
$$ . . . . a 2 b . . |
$$ . . . . . . . . . |
$$ ------------------[/go]

May add more later...