Life In 19x19

So I've had more free time recently, and after trying my hand at chess for some time I finally come back to Go. I'm playing mostly on IGS as I can always get a game quickly with acceptable time settings there. I am currently up to 6k which means if the rank comparison table is not widely out there I've not gotten that much weaker despite the long hiatus.

As for my current study schedule, I've read about the "memorize the first 100 moves of 100 pro games to get to shodan", and while I certainly don't think it will magically be this effective I've decided I might as well try it out since I do enjoy playing over pro game. Of course, I mean a pretty weak interpretation of memorizing game, the aim is simply to be able to replay the first 100 moves of the game without looking at the diagram. Then I will do it a second and sometime third time immediately again (and if I can do that it generally mean I can easily replay it within a few hours if I don't go over another game in the meantime).

That said, since I've gotten the Genjo-Chitoku book it would be a shame to just use it to memorize game, so I've been using the above method along with more in depth study of the games, at least for those which are commented. It's also quite easier to follow and analyze the comments when you have the moves of the game in mind.

Apart from that, I'm obviously doing Tsumego because of course you should do Tsumego if you're serious about going stronger. I'm mostly doing the "lots of easy problem fast" variant for now, though I'm thinking of going over the Gokyo Shumyo in the near future. I'm also unsurprisingly re-reading the Kageyama as well as James Davies' Tesuji.

What I'm probably not doing enough is a thorough review of my games, but well, most of my wins are rather comfortable and many of my losses are due to one tactical or strategical blunder. Reviewing the first type of games feel a bit like going "wow my opponent was so terrible" again and again, and the second can be "okay I was fine here and what a moron I was making this move".

Here is a (rather extreme) example of the first type of game, where I had a very comfortable position after my opponent decided to create a group to suicide himself for my profit, and was never really in trouble and then my opponent self-destructed and half the board died and one of his few remaining group needed to live by ko.



Other than that I feel that my ability to quickly evaluate the balance of territory without precise counting is quite terrible.

Bki wrote:Apart from that, I'm obviously doing Tsumego because of course you should do Tsumego if you're serious about going stronger. I'm mostly doing the "lots of easy problem fast" variant for now, though I'm thinking of going over the Gokyo Shumyo in the near future. I'm also unsurprisingly re-reading the Kageyama as well as James Davies' Tesuji.

Doing lots of easy problems fast may be a confidence builder if you get them right. (I suspect that you are batting close to 1000. ) Aside from that, I doubt if that is helping your game much. In general, the best problems for learning are those that you get right around 50% of the time. Doing them fast is fine, but they should not be too easy.

BTW, I have tried doing problems on the computer with the compture playing the main line of the solution at a rate of 5 sec. per move. I found that if I missed the problem, after the first move I could almost always see the rest of the line in a second or two. (This was on a web site that unfortunately no longer exists, but something similar can be done manually.) Seeing the solution one move at a time instead of all at once is a plus, IMO.

----

About your game, B211 was excellent, OC. In terms of life and death, neither B201 nor B219 was necessary for the kill.

Bill Spight wrote:Doing lots of easy problems fast may be a confidence builder if you get them right. (I suspect that you are batting close to 1000. ) Aside from that, I doubt if that is helping your game much. In general, the best problems for learning are those that you get right around 50% of the time. Doing them fast is fine, but they should not be too easy.

BTW, I have tried doing problems on the computer with the compture playing the main line of the solution at a rate of 5 sec. per move. I found that if I missed the problem, after the first move I could almost always see the rest of the line in a second or two. (This was on a web site that unfortunately no longer exists, but something similar can be done manually.) Seeing the solution one move at a time instead of all at once is a plus, IMO.

Well, it does allow to assimilate some common eye gouging patterns at the very least, and generally it does develop intuition. Of course problems such as killing a bulky five or living with a ponnuki in the corner are a bit too easy to really be of any use.

Also the way I tend to do problems tend not to give a 50% success rate, since unless I set myself a time limit I tend to stubbornly read things until I find the solution and tested it against sensible defense... though of course that's the ideal and I don't always manage to be that thorough (problems without answer motivate me more for that naturally), of course, but it means I'm more likely to spend half an hour on that one stupid problem until I'm confident I got it right.

About your game, B211 was excellent, OC. In terms of life and death, neither B201 nor B219 was necessary for the kill.

It's less that B211 was excellent and more than W210 was stupid (or so desperate he didn't care anymore probably). Actually on l&d I think B127 was wrong and should have been at R10. Of course, during the game I knew I was winning and so just made sure all my groups were connected and alive rather than going hard for the kill.

And yes B201 was completely redundant. And gote.

Bki wrote:

Bill Spight wrote:Doing lots of easy problems fast may be a confidence builder if you get them right. (I suspect that you are batting close to 1000. ) Aside from that, I doubt if that is helping your game much. In general, the best problems for learning are those that you get right around 50% of the time. Doing them fast is fine, but they should not be too easy.

BTW, I have tried doing problems on the computer with the compture playing the main line of the solution at a rate of 5 sec. per move. I found that if I missed the problem, after the first move I could almost always see the rest of the line in a second or two. (This was on a web site that unfortunately no longer exists, but something similar can be done manually.) Seeing the solution one move at a time instead of all at once is a plus, IMO.

Well, it does allow to assimilate some common eye gouging patterns at the very least,

Maybe.

and generally it does develop intuition.

Maybe not.

Also the way I tend to do problems tend not to give a 50% success rate, since unless I set myself a time limit I tend to stubbornly read things until I find the solution and tested it against sensible defense.

Fine. But we are talking about doing problems quickly, with a short time limit. I would set a time limit of 10 seconds, myself.

I suggest that you would benefit from doing easy, but harder problems quickly. Why not give that a try?

Bill Spight wrote:Fine. But we are talking about doing problems quickly, with a short time limit. I would set a time limit of 10 seconds, myself.

I suggest that you would benefit from doing easy, but harder problems quickly. Why not give that a try?

Well, I'm doubtful a problem that can be meaningfully solved in 10 seconds would count as anything but easy.

Also it's not like I'm intentionally avoiding slightly harder problem. Mostly I'm working my way back through old problem sets which usually start with easier problems. The plan is to eventually have a set each of easy, medium and hard problem (what exactly should qualify as easy, medium or hard is another interesting question of course).

Also a game I played today. Again my opponent was super unreasonable and died a lot even if I didn't even need to kill to win :



W152 was overkill now that I look back on it. I think I do have a tendency to make superfluous move to make doubly sure my opponent is dead/I am alive when I'm already obviously winning. Of course in those situation where I'm so much ahead it doesn't matter but I have to make sure not to do the same in situation where the score is closer.

A game without something big being killed, though I felt comfortably ahead for most of the game.

So I've finally decided (and tried a bit) on a daily training plan. Following the advise to also do harder problems, I'm doing all three volumes (elementary, intermediate and advanced) of Cho Chikun's encyclopedia of life and death simultaneously. I do it in chunk of less than 30 minutes (doing too much problems at once lead to diminishing returns imo as you're not as focused by the end of it) each, with 60 elementary problems at less than 30 second each, 30 intermediate problems at 1 minute each, and 10 advanced problems at 3 minutes each. If I feel like doing more problems that day I will take out the Gokyo Shumyo, and solve a few number of problem to my satisfaction (not timed but until I'm pretty sure I got the solution).

Of course, since the problem sets don't have answers, it can be a bit tricky to mark my mistakes. If I didn't get the solution in the required time, obviously it's a failure. If I didn't have time to check my solution to be completely certain it actually works, it's a failure. If I'm certain my solution is right but I haven't double checked it enough? I note it separately, but for all other purpose it count as a failure.

(It makes it hard to get correct for problems where the right answer lead to a ko, of course, since I would have to check that all the plausible way to try to get unconditional death fail before that).

I not which problem I failed, and how many for each set I failed each day, as well as the accumulated failure rate (total of problems failed in this set/total of problems attempted). When the later will get strictly greater than 50%, I will start all over from the beginning.

I might also add some other rules if I think of them and they feel like a good idea.

Anyway, that's the problems. About pro games I aim at playing between 1 and 3 games each day. I'm not yet sure if memorizing the first 100 moves of a lot of games will make me significantly stronger, but at the very least I've already gotten better at memorizing the first 100 moves of a game . The first times I tried it took almost 2 hours, now I'm down to a bit more than 30 minutes (it vary according to the game, and also how much time I spend on trying some variation, playing over the rest of the games, etc...)

Playing over a game can be perfect to fit between two tsumego sets too.

As for playing the game myself, I do it when I feel like it, which at the moment means around 2-3 games a day, sometime a bit more. I will look over each games afterwards, though how much time do I spend on that vary.


So that I don't get accused of only posting my win, here a game I lost today :



I think around or were the critical moment. Taking sente to play deep within the center might have been better (I take a loss on top of course but I think I can get at least a seki there). His was well timed.

Maybe, though, that wasn't necessary and should simply have been more ambitious. Or maybe it was fine and the issue was that I didn't apply myself enough in the endgame. I think the most glaring mistake was white 165. At n13 (:b166:) instead might have been better. Also, trying to get H2 would have been good, him playing J2 was pretty big and let to the blunder that cost me L2. Maybe white 179 at L3 would have been enough?

The advanced tsumegos are quite hard. I've started it over for the third time. Sometime I get problem that I "solved" the previous time wrong, but I think most of the time is that I was a bit too optimistic and I failed to see a refutation for my play (and sometime the "refutation" I saw the second time was wrong of course). Still, some problems are not that hard to do again after understanding the trick (or when I already know that ko is the best that can be done). But other elude me after several attempts, so I've started trying to solve them separately by playing them over a board.

Let's take the first problem for example :

Click Here To Show Diagram Code

[go]$$B Black to play
$$ ------------------
$$ | . . . . O . . . .
$$ | . X X X O X . . .
$$ | . O O O . . . . .
$$ | . . . , . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]

If black is to do anything with this position, the cut is the only move. Then the next few moves are basically forced :

Click Here To Show Diagram Code

[go]$$B The first four moves are obvious
$$ ------------------
$$ | . . . a O . . . .
$$ | b X X X O X 4 . .
$$ | . O O O 1 2 . . .
$$ | . . . , 3 . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]

Now what I first looked at was the descent at a, then trying to get b in sente. But since playing a immediately remove a liberty, it's not good. But trying to capture the 3 white stones doesn't really work. So what I've come up is this :

Click Here To Show Diagram Code

[go]$$B Solution?
$$ ------------------
$$ | . . . . O . . . .
$$ | . X X X O X O . .
$$ | a O O O X O . . .
$$ | . B . , X b . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]

White can't prevent the connection at a because he then loses the semeai (if white play a though, black must first play b and only then start capturing white. Otherwise white atari his way to victory). Then, either he enclose black, but then black can live :

Click Here To Show Diagram Code

[go]$$B Black live in the corner
$$ ------------------
$$ | . 9 . 7 O 8 . . .
$$ | 5 X X X O X O . .
$$ | 1 O O O X O . . .
$$ | 3 X W , X . . . .
$$ | 4 2 6 . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]

Black get and in sente and thus can live.

Click Here To Show Diagram Code

[go]$$B Black get two eyes
$$ ------------------
$$ | . . . . O 2 . . .
$$ | . X X X O X O . .
$$ | X O O O X O . . .
$$ | . X O , X 1 . . .
$$ | . 3 W . . . . . .
$$ | . 5 4 . . . . . .
$$ | . 7 6 . . . . . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]

If white doesn't atari, black can crawl and get two eyes. If at , then black cut at and capture his way to life. is necessary, but doesn't need to be played immediately.

I think my analysis is correct, but if I failed to see something obvious that refute it, then please tell me.

Click Here To Show Diagram Code

[go]$$B Black to play
$$ ------------------
$$ | . . . . . . . . .
$$ | . O . . . O X X .
$$ | . O X X . O O X .
$$ | X O O X . . . X .
$$ | O O O X . . O . .
$$ | X X X X . O . . .
$$ | . . . . . . . . .
$$ | . O O . O . . . .
$$ | . . . . . . . . .[/go]

Another tough problem, though once you found it the solution seems to be rather simple. I tried a lot of thing for this problem, even some convoluted ways to make two eyes by threatening to cut through (which obviously didn't work).

Click Here To Show Diagram Code

[go]$$B "a" and "b" are miai
$$ ------------------
$$ | . 1 4 5 . b . . .
$$ | a O 2 3 . O X X .
$$ | . O X X . O O X .
$$ | X O O X . . . X .
$$ | O O O X . . O . .
$$ | X X X X . O . . .
$$ | . . . . . . . . .
$$ | . O O . O . . . .
$$ | . . . . . . . . .[/go]

is forced, then threaten ko at "a" and connecting at "b". If at , then black hane, white connect, black "a" for basically the same ko.

I'm rethinking my L&D training. It's fine for the elementary set (which I'm about to finish and I've been wondering), but I've been dissatisfied with the intermediate and advanced set. For the former, there's many problems that I know I can solve given the time to work out the solution, but the time run out (especially when I let my mind wander around some variation I know fail), which is frustrating. The advanced ones should probably just be solved at my leisure.

So I will probably drop those two at the moment, and keep doing the elementary set (or do another set of easy problem in the meantime). I do need to find a set of harder problems to do and an appropriate way to study them though.

I've had a pretty bad losing streak recently, even falling back to 6k on IGS, but since then I've been doing better. Hopefully that losing streak was of the sort that indicate I am getting stronger. Or maybe simply having played a bit less as a result, I've been applying myself more for each game (I feel like I would get at least a couple stone stronger if I stopped making some blunder that I feel are beneath me, but clearly aren't).

On training, I've gotten the Segoe Tesuji Dictionary, and I'm enjoying it so far. I actually like the splitting of problems by type of move rather than their purpose. This make it far more enjoyable to go through the book in order, as you encounter fighting, life and death and endgame situation. Not understanding the text bothered me a bit at first, but the problems are well done and the tactical situation quite clear so it's often obvious whether you're supposed to cut, connect, live, kill or just grab points... I'm going through all the C problems first, studying the book for 45 minute-1 hour per day, and taking as much time as I feel like on each problem. I will think more about how to proceed further once I've gotten through all 3 volumes this way.

On L&D problems, I'm going through the Elementary problem a second time. For problems I've failed to solve in the time I've given myself, I will try to pause and take the time needed to solve them rather than just go immediately over the next problem. I will also sometime take the time to thoroughly double-check my reading for a problem I'm pretty sure I have solved correctly.

Also I've gotten the Takao Shinji Joseki Dictionary but I'm not yet sure what will be the best way to study it.

From the game four posts upthread:

You played the 2-2 point, which allows black this sequence:

Click Here To Show Diagram Code

[go]$$Bc Position at move 70
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . X . . . . . . . . . . . . . . |
$$ | . . O O X . . . . . . . . O X . . . . |
$$ | . . O X . . . X . O . . . O . X . . . |
$$ | . . O X . . . . . . . . . . . . . . . |
$$ | . O X X . . . . . . . . . . . X . . . |
$$ | . O O X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . X . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . O X . . . . . . . . . . . . . . |
$$ | . X X O O X X . . . . . . . . . . . . |
$$ | X . X X O O . X . . . . . . X . O . . |
$$ | . X O X . . . . . O . . . . X . O . . |
$$ | 6 O O X . O . . . , . . . X . X X O . |
$$ | 5 . O X X O . 2 . O . X . X X O O . . |
$$ | 7 W O O X 1 . . . . . . . X O O . . . |
$$ | . . 4 3 . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Now if you tenuki, he can play at A1, and have a seki, which he can turn into a bent four at his leisure if he ever finds himself with favorable ko threats. So you must play at A1,starting the ko yourself. And you must win it. If he has two good threats he can kill your corner. ( Uberdude will probably post here to tell me that I am wrong in some detail, but the point remains that you are allowing complexities in making life )

As a general rule, the 1-2 points are the most efficient ways to make eyes in the corner, and should be at the top of your short list when looking to make life.

In this instance, it works better than the 2-2:

Click Here To Show Diagram Code

[go]$$ Position at move 70
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . X . . . . . . . . . . . . . . |
$$ | . . O O X . . . . . . . . O X . . . . |
$$ | . . O X . . . X . O . . . O . X . . . |
$$ | . . O X . . . . . . . . . . . . . . . |
$$ | . O X X . . . . . . . . . . . X . . . |
$$ | . O O X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . X . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . O X . . . . . . . . . . . . . . |
$$ | . X X O O X X . . . . . . . . . . . . |
$$ | X . X X O O . X . . . . . . X . O . . |
$$ | . X O X . . . . . O . . . . X . O . . |
$$ | . O O X . O . . . , . . . X . X X O . |
$$ | . . O X X O . . . O . X . X X O O . . |
$$ | W a O O X . . . . . . . . X O O . . . |
$$ | . b . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

The point pair a/b are miai, ensuring two eyes without an inconvenient ko.

In summary: look for the 1-2 points when making life.

For reference:

Click Here To Show Diagram Code

[go]$$Bc White lives
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . X . . . . . . . . . . . . . . |
$$ | . . O O X . . . . . . . . O X . . . . |
$$ | . . O X . . . X . O . . . O . X . . . |
$$ | . . O X . . . . . . . . . . . . . . . |
$$ | . O X X . . . . . . . . . . . X . . . |
$$ | . O O X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . X . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . O X . . . . . . . . . . . . . . |
$$ | . X X O O X X . . . . . . . . . . . . |
$$ | X . X X O O . X . . . . . . X . O . . |
$$ | . X O X . . . . . O . . . . X . O . . |
$$ | 2 O O X . O . . . , . . . X . X X O . |
$$ | . . O X X O . W . O . X . X X O O . . |
$$ | . O O O X B . . . . . . . X O O . . . |
$$ | . . . 1 . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Click Here To Show Diagram Code

[go]$$Bc Variation, Seki
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . X . . . . . . . . . . . . . . |
$$ | . . O O X . . . . . . . . O X . . . . |
$$ | . . O X . . . X . O . . . O . X . . . |
$$ | . . O X . . . . . . . . . . . . . . . |
$$ | . O X X . . . . . . . . . . . X . . . |
$$ | . O O X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . X . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . O X . . . . . . . . . . . . . . |
$$ | . X X O O X X . . . . . . . . . . . . |
$$ | X . X X O O . X . . . . . . X . O . . |
$$ | . X O X . . . . . O . . . . X . O . . |
$$ | 2 O O X . O . . . , . . . X . X X O . |
$$ | 1 . O X X O . W . O . X . X X O O . . |
$$ | 3 O O O X B . . . . . . . X O O . . . |
$$ | . 5 a 4 . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Note: White could start a ko with a play at a.

----
Edited for correctness.

Indeed, I was mixing my L+1 groups at the time of the game. The 2-2 points is the vital point for the L+1 group with the leg on the long side. It's a mistake that I think that I fixed since then but still I do appreciate this being brought up.

In this particular situation though, if B try a seki now, then after in Bill's variation B6 isn't a guaranteed eye anymore. Then white atari at E9 threaten B9 next which would kill his surrounding group. That ponnuki in the center reduce his central prospects to basically nothing. Then provided I can get R13 I think I would be ahead in the game. Later in the game it was even more impossible for him to go for the seki since A4 threaten to kill the surrounding group.

A game I played today.



I usually don't approach right as with but here I felt approaching low would let him develop the lower side too easily. With the game move I get some profit on the side and still leave the possibility for an invasion or reduction later. I do wonder whether I could have tenukied with to approach the top... Maybe if he had played a one point jump with instead since that would make it harder for him to mostly kill the 3 stones with two moves.

I'm not sure about the top right. at Q12 might have been trouble. But the game result was reasonable. attaching on top at E15 might have made more sense with the top. Then if I can get sente to play L15 my position there is rather good. was what I played instead of a one point jump since I felt this had a better relation with the L17 group (especially once I get L15)

invited of course, but I do think I ended up ahead in in the exchange that followed. I think or a move with the same idea, was more or less forced. Going deeper wouldn't feel likely to work well though. Maybe a cap at K5 would have been enough though? was needed to prevent the S8 clamp which would be unbearable (the cut doesn't work yet though). Either of W110 or W112 are dubious, but I completely missed B113 during the game. Still because of his side group's weakness I do think how the game went didn't too badly for me since I ended in sente.

The game was ended by a blunder, but even if it hadn't I would probably still be ahead as long as I played reasonably well.