Life In 19x19

Needo wrote:I am pretty sure that the app is more likely to reuse problems that I have missed. I am just now up to 650 problems in the pool now. Up until problem 900 I was only drawing from a pool of 350. I added more problems to the pool because I did remember the solution to a many of the problems and I wanted to force myself to read.

When I was learning go I also avoided looking at problems with solutions that I remembered. I now think that was a mistake. In a real game, memory is, if anything, at least as effective as reading, and supports it.

BTW, if you remember a solution to a problem, you can use that memory to strengthen your visualization. First, close your eyes and visualize the problem diagram. (Not as easy as it may sound.) Then mentally visualize the solution, move by move.

It is also possible that the 300 problems that you added to the pool are too difficult for you right now.

Problem and visualization exercise.



Black to play and kill.

The main line has 5 moves, and the main failure sequence has 6 moves. But there are other relevant variations that have over 10 moves, not all alternating. For instance, in the main failure sequence, what if White passes or tenukies at move 4? OC, in a real game that is unlikely to happen, but visualizing those sequences is good visualization practice.

Bill Spight wrote: You mentioned that you gave gotten past problem 1000. IMO it is not easy to come up with that many tsumego problems at your level. Maybe so, since today there are many problems available online, but when I was learning go that was not the case. I did fewer than 500 tsumego before reaching shodan. Maybe you should start over and do the easiest 500 problems again.

The app keeps track of how many problems have been attempted. Most of those problems I saw multiple times. I suspect that by the time of writing that 1st post I only saw somewhere between 250 and 350 unique problems. I also want to thank you for pointing out the empty 6 in the corner problem. I am going to use that problem for practice during work tomorrow. If I remember correctly, that is an unsettled position in the Davies Life and Death book. I don't want to look until I get back to town, so that I can use the exercise. I also want to see if it makes any difference on the edge and if having 1 extra liberty for black makes any difference.

Not all the news was bleak. I did have time for a game today before picking my kid up from driver's ed. The life and death problems may have helped me murder a group, make a small reduction, and allow me to tenuki earlier than what I would usually be comfortable with. I should look back at that game to make sure I didn't leave the fight too early. I should also look to see if that reduction could have been fended off.

There are some easy sets of problems here:

https://tsumego-hero.com/sets

(sort the sets by difficulty)

Needo wrote:I also want to thank you for pointing out the empty 6 in the corner problem. I am going to use that problem for practice during work tomorrow. If I remember correctly, that is an unsettled position in the Davies Life and Death book. I don't want to look until I get back to town, so that I can use the exercise. I also want to see if it makes any difference on the edge and if having 1 extra liberty for black makes any difference.

(Emphasis mine.)

Good. That's the spirit!

While you're at it, you might see what difference 2 extra liberties make for the corner group.

Hi Needo,

you might see what difference 2 extra liberties make for the corner group.

After that, a very natural next shape to study, LGroup:

Click Here To Show Diagram Code

[go]$$
$$ +--------------
$$ | . . . . . . .
$$ | . . . O X . X
$$ | . O O O X . .
$$ | . X X X X . .
$$ | . . . . . . .
$$ | . X . . . . .[/go]

I want to thank JLT for that website . I like the way it makes you cool down for a day if you fail that problem. That has helped slow me down and not play an impulse move. Ed I am afraid that I can't keep track of all of the moves in my head with the L group. Using a crutch, I will have to use a program and explore all of the decision tree for that group.

I don't know if this has happened with anyone else on this forum. But, I have become a little obsessed with solving problems to the extent that I would rather do tsumego than play an actual game. I am sure this will pass.

It is cool to find out that you only need i extra liberty to freeze an incomplete killing shape if it has 1 shared liberty, and you need two extra liberties to freeze an incomplete killing shape if its remaining liberty won't be shared.

I would also like to find out what conditions are required to make an unsettled position symmetrical (the first move to kill is in the same location as the first move to live) verse you can't kill a group by playing where its first move to live would be. (you have to play somewhere else) This is just like learning to do math again .

Hi Needo,

I have become a little obsessed with solving problems to the extent that I would rather do tsumego than play an actual game. I am sure this will pass.

I wish I had this 'problem'!

can't keep track of all of the moves in my head with the L group. ... use a program and explore all of the decision tree for that group.

Even better is using actual stones on a real board.

Playing on a board is certainly preferable than doing so on the computer imo, though I tend to use it mostly for tough problems, though sometime I will play over the variations of an interesting problem after solving it.

As for the L group, I think it's a good candidate to train your visualization ability. The position is not hard to recreate in your mind whenever you feel like working on it. Then you consider the various ways white could try to live and try to see the variations unfold in your mind as you kill him... Then you add hanes, then do the L+1 groups, L+1 groups with hane, etc...

It won't come at once but if you practice it it will help solving problem and reading deeper in general.

Needo wrote:Using a crutch, I will have to use a program and explore all of the decision tree for that group.

I don't know if this has happened with anyone else on this forum. But, I have become a little obsessed with solving problems to the extent that I would rather do tsumego than play an actual game. I am sure this will pass.

For a small tsumego like the L group, it is possible to explore all the decision tree mentally, but for a more complicated one, nobody can do that. Here is how I proceed (I am not saying this is the best way, I am just a kyu player and have seen people solve tsumegos with much less effort than me...)

1) Try to find a reasonable move for . It could be a move that reduces space, or reduces liberties, or that looks like a vital point, or you may recognize a known technique or a known shape...

2) Then try to find a reasonable main line ...

3) If you think that works, explore the tree a bit more. At least, look at what happens if is played at .

4) If your turns out to be wrong, go back to stage 1) and try another candidate for .

5) If none of your candidates works, then either is not among your candidates so you must try something "unnatural", or was among your candidates but you misread some sequence ... so you must go back to the beginning and explore the tree more thoroughly. You may want to proceed by elimination, discard first moves that are definitely wrong so that you can concentrate on other candidates.

jlt wrote:

Needo wrote:Using a crutch, I will have to use a program and explore all of the decision tree for that group.

I don't know if this has happened with anyone else on this forum. But, I have become a little obsessed with solving problems to the extent that I would rather do tsumego than play an actual game. I am sure this will pass.

For a small tsumego like the L group, it is possible to explore all the decision tree mentally, but for a more complicated one, nobody can do that. Here is how I proceed (I am not saying this is the best way, I am just a kyu player and have seen people solve tsumegos with much less effort than me...)

1) Try to find a reasonable move for . It could be a move that reduces space, or reduces liberties, or that looks like a vital point, or you may recognize a known technique or a known shape...

2) Then try to find a reasonable main line ...

3) If you think that works, explore the tree a bit more. At least, look at what happens if is played at .

4) If your turns out to be wrong, go back to stage 1) and try another candidate for .

5) If none of your candidates works, then either is not among your candidates so you must try something "unnatural", or was among your candidates but you misread some sequence ... so you must go back to the beginning and explore the tree more thoroughly. You may want to proceed by elimination, discard first moves that are definitely wrong so that you can concentrate on other candidates.

I object to your 2-3). It's fine to guess black's moves, but you can't really call your solution correct if you didn't check it worked against any of white's defenses. Of course, thankfully it's obvious that most of white's possibility fail to work, and often many black moves are forcing so you only really have to consider one or two of white's moves. But failing to consider all the way white can try to resist is blindly guessing and wishful thinking, not solving a problem.

As an addition to your step 4, having read out the variations for a promising but ultimately failing can help you find better candidate moves.

Bki wrote:

jlt wrote:(...)
2) Then try to find a reasonable main line ...

3) If you think that works, explore the tree a bit more. At least, look at what happens if is played at .

I object to your 2-3). It's fine to guess black's moves, but you can't really call your solution correct if you didn't check it worked against any of white's defenses.

Sure, I agree that you should check "everything", but nobody does that when the tree has thousands of paths. Some branches are "obviously" discarded, but what you think is obvious may be wrong (otherwise you would solve every tsumego correctly).

In my point 3) I said At least, look at what happens if is played at .
I find that a large percentage of misreads are due forgetting to check that.

jlt wrote:

Bki wrote:

jlt wrote:(...)
2) Then try to find a reasonable main line ...

3) If you think that works, explore the tree a bit more. At least, look at what happens if is played at .

I object to your 2-3). It's fine to guess black's moves, but you can't really call your solution correct if you didn't check it worked against any of white's defenses.

Sure, I agree that you should check "everything", but nobody does that when the tree has thousands of paths. Some branches are "obviously" discarded, but what you think is obvious may be wrong (otherwise you would solve every tsumego correctly).

Most of "obvious" wrong moves are really obviously wrong though. If the last black move threaten to make two eyes in a to-live problem, or make a nakade shape in a to-kill problem, anything that doesn't answer that threat can be immediately discarded. And lot of pruning can be done with experience. If you see a L-group (or 6 in a row, or some common false eye pattern) then you stop reading that variation.

The reason we don't solve every tsumego correctly are either due to failing to find the tesuji or impatience (or self-imposed time limits).

It's true that sometime you can't read out all the plausible defenses in a reasonable time, but that simply mean that the problem is above our level for the moment, not that it is impossible.

In my point 3) I said At least, look at what happens if is played at .
I find that a large percentage of misreads are due forgetting to check that.

I don't know, I think the most common misread is not noticing a shortage of liberty a few move deep within a variation

Bki wrote: I don't know, I think the most common misread is not noticing a shortage of liberty a few move deep within a variation

I agree that it's also a common cause of misreading. A good habit is to start by counting liberties of all chains before making any attempt.

jlt wrote:

Bki wrote: I don't know, I think the most common misread is not noticing a shortage of liberty a few move deep within a variation

I agree that it's also a common cause of misreading. A good habit is to start by counting liberties of all chains before making any attempt.

Well worth repeating.