has your reading speed gotten better from doing tsumego? reading being, according to the chapter of tesuji by davies, confirming ALL possible white responses to your move fail before checking the answer?

i ask because about 6-7 weeks ago i realized my awful reading was holding me back, it is far below my ranks on KGS and Tygem (still on graded go problems for beginners vol 2 for petes sake and i cannot even touch the tesuji book because those problems take 5-10+ minutes) so i made it my priority, i do a minimum of 20 problems a day, usually 50, and i cannot read any faster, i can guess the vital point quicker and so i spend less time reading out wrong answers, but it takes me just as long to read as it always has. reading itself is an absolute chore and exhausting mentally. i cannot read in games because i already run out of time as it is (meanwhile my opponents spend barely any time at all)

primarly, the concept of tsumego is confusing and unclear to me. what exactly is the point of it? what should i be getting from it? i thought it was reading speed but since that hasn't improved its made me wonder.

there is a page on sensei's library where people time themselves doing different books. for GGPB vol 2 there is a 1 dan who had an average speed of 5 seconds per problem. i do not believe he could have possibly read out all of whites responses to every problem at a speed of 5 seconds per problem, but thats what tsumego is right? i have to believe that time consists of guessing the vital point in perhaps .5-1 second, then taking 3-4 seconds to flip to the answer and then back to the next problem. or maybe im wrong and that's the kind of speed your reading can get to. i don't know. i don't know what to expect. so much of tsumego seems like some orally passed down tradition that isnt in writing anywhere. it took me a while to even wonder, should i not be doing these problems on a board, but in my head? then it took that chapter by davies to tell me that i had to check every possible response by white. i still wonder if "correct" reading means to be able to imagine and retain all the stones visually in your head, which seems impossible to me, but what do i know? tsumego comes with essentially no instructions, nor really, "here's how to make the most of these" and instead "just do them, you'll see why". my frustration being i have yet to see why, improvement wise, and its driving me crazy

"Raw reading" gets a bit easier, but stays pretty exhausting. What really gets better is your shape-knowledge, which allows you to prune a lot of reading.

For example, here's a tsumego which will take you about a second if you know the corner shape: (white to play)

Click Here To Show Diagram Code

[go]$$W
$$..XXXXX.|
$$..XOOOX.|
$$..XO.OOX|
$$..XO.OX.|
$$..XO....|
$$--------+[/go]

If you don't know the shape, you'll have to read about 7 moves deep. And even that will be easier if you're familiar with these sorts of oiotoshi shapes.



Nah, the problems in that volume are very basic for a dan player. He (or she) will have seen most of them (in variations maybe) many times and solved similar ones again and again.

Just like the problem above, a dan player won't have to "read out" any variations here, he (or she) just knows the shape.

Well, as a trivial example: you know that a three point eye shape lives or dies according to whose turn it is, and straight four unconditionally lives, right? Now: to a total beginner that is not trivial, but requires reading!

By doing tsumego I have learned to adapt multiple, more complex heuristics like that as effortlessly as the above rule. See leichtloeslich's post for an example. I don't have to read, I just know "instantaneously". This is likely what the 1d at senseis also meant: you don't have to read as far because the result is already obvious after e.g. finding the first move. I have also learned to better visualize the stones in my mind, and to just read faster. It is a bit of each of those three, why doing tsumego is good, I would say. Just do a load of it. Doing those problem pdfs at Tatsuki's helped me a lot, they don't have answers so you really have to read, not just peek at answers. Also helpful at some point to really start reading about basic shapes at e.g. Senseis. I wish I had started doing that earlier than I did.

Regarding visualisation of stones: I can visualize perhaps about 5x4 area quite effortlessly. (give or take one row eiher way). For positions which cover a larger area, it seems to me I focus on that kind of area, storing situation elsewhere with mental notes like "that wall has three libs and a weakness like that" . YMMV.

A lot of time, the basic shapes of easy tsumego are absorbed into your memory. The work you did long ago to understand all the different lines of the basic shapes means that you know them by heart and when you see the shape again, you know exactly what to do without having to think about it all over again and there are no surprises that can be pulled out because you know what to expect for all circumstances. For harder tsumego, many times it's an exercise in how to reduce the problem to basic shapes you already know. Then you look around to see if there are surprise lines which require further thoughts. So basically the more tsumego you do and understand, the bigger your personalized database becomes in your brain. And the easier it becomes to see such shapes in your own games so you can use the knowledge to move beyond beginner levels.

_________________
http://tchan001.wordpress.com
A blog on Asian go books, go sightings, and interesting tidbits
Go is such a beautiful game.

Don't be discouraged, it's quite normal for an 11k to find GGPfBv2 challenging. I think your reading is probably on schedule.

Don't expect to notice your reading getting faster. It's like breathing. People who start exercising daily don't notice their lungs expanding, do they? You'll gradually notice that volume 2 is starting to seem boring and volume 3 is starting to seem do-able, but that happens slowly. You'll eventually hit milestones where you're laughing at someone's reading mistake and then realize you used to make the same mistake.

By the way, you say that you can see the vital point when you do the problems so you spend less time reading out wrong answers. As a general rule even when the first variation you try works, you should spend a bit of time reading the failure variations to make sure you can find the refutation of the mistake. If you find the refutation, you practice your reading and, often enough, see some interesting tricks that are the real point of the problem. If you can't find the refutation, that may mean that there are multiple ways to live (of which one may be better than the others), or it may also mean that you haven't found the theme of the problem yet, and that there is some strong resistance that you haven't found yet. (Possibly you do this already, but one interpretation of your answer is that you jump to the next problem as soon as you find any solution that seems to work.)

By the way, if you could post a game where you got into time trouble, it might help us see what kinds of situations take you too long to read out, and diagnose the problem (if any).

I've only recently got through that book myself and I think the biggest thing to come out of it isn't reading speed or accuracy, though I expect they've improved a bit, it's just that I'm beginning to read situations out more even if it is still spotty as to when I do that or just lob a stone down.

Also you say you run out of time, what time settings do you typically play on?

PeterN

I disagree slightly. I was able to solve the problem very quickly, but not by "recognizing the shape". There is also the matter of "general principles" involved and this example problem was a case of "if there are two ways to make life, each failing by one move, and these two ways share a move in common. then alive".

Thus R1 makes "three in a row with the move" because R1 is sente against the corner (if black uses the next move to stop the three in a row the corner dies). Once you see that also see that white could start with T2 and then R1 after black captures or captures if black connects (the move in common doesn't always have to be the first move.

But while teaching us to read better, problems also have aspects related real games that don't usually get addressed when solving problems. Besides deciding live, dead, or seki, in the real game context need also to be able to read "dead in isolation but represents X forcing moves" (how many ko threats would this dead group provide).

Click Here To Show Diagram Code

[go]$$W
$$..XXXXX.|
$$..XOOOX.|
$$..XO.OOX|
$$..XO.OX1|
$$..XO5432|
$$--------+[/go]

Got it!

Click Here To Show Diagram Code

[go]$$W
$$..XXXXX.|
$$..XOOOX6|
$$..XO.OOX|
$$..XO.OX1|
$$..XO5432|
$$--------+[/go]

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

A bit more on the problem.

Click Here To Show Diagram Code

[go]$$W Solution 1
$$..XXXXX.|
$$..XOOOX.|
$$..XO.OOX|
$$..XO.OX1|
$$..XO.3.2|
$$--------+[/go]

Click Here To Show Diagram Code

[go]$$W Solution 2
$$..XXXXX.|
$$..XOOOX.|
$$..XO.OOX|
$$..XO.OX.|
$$..XO.1..|
$$--------+[/go]

One of these solutions is generally better than the other one. Which one is it?

Edit: Oops. It is possible to have whole board positions where one solution is correct, and other positions where the other solution is correct. I have edited the question accordingly.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Yes. On the last point see http://senseis.xmp.net/?GoProblemsTheFudgeFactor



Fine. Focusing on your weaknesses is good.



Why so many? If you spend two minutes on a problem, 50 problems will take you over an hour and a half. And then you still are unlikely to understand the ones that you missed. Is your purpose to do some number of problems? Or is your purpose to understand them?

BTW, as your reading improves, you are unlikely to spend less time reading, but you will read deeper and more accurately.



If you are not reading, what are you doing?



Every game of go is a learning opportunity. Reading is a way of exploring the possibilities of a position, and learning about it. A tsumego problem is an opportunity to learn about a significant position in go, and to practice learning about it.

Speed comes from recognition and understanding. By and large you do not have to work at it, it comes naturally. If you know a position thoroughly, you will see the key sequences and their consequences quickly.

If you miss a problem, then you do not understand it thoroughly. Just noting the answer and saying, OIC, does not mean that you understand it thoroughly. Instead of moving on to the next problem, spend a few minutes thinking about the problem, about what are the key stones, points, and plays, about why the order of play matters, or if it does. Think about why you missed the problem and why your answer was wrong. Take a moment, close your eyes, and visualize the problem position in your mind, then visualize the key sequences, both successes and failures. Once you have understood the problem, then move on to the next one. Better to understand one problem than to zip through fifty.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

the reason i take so much time when playing games, even though i don't read much, is because i am trying to decide what the biggest move is and sometimes it can take me a minute or two before i decide and these add up.

the reason i do so many tsumego is i've heard it's best to do as many as possible. i also don't get them wrong and go OIC and immediately move right on to the next lol. like i said i read them all the way out so i rarely get one wrong but if i do try and figure out why it was wrong and what my mistake was.

I think I see what you're saying. But when you try to decide which move is biggest, I can imagine three things you can be doing.

1. Reading out local sequences that follow from your initial move (either now, or later).
2. Reading out the local sequence that would follow if you don't get around to playing here.
3. Reading out global sequences that follow from your initial move.
4. Comparing -1- and -2- to get a sense of the swing in points involved.
5. Assessing the board at the end of the sequences you're looking at (either to figure out who's leading, or to make a qualitative judgment like "After this and that, I'll have to invade and live on the top")
6. Evaluating how a certain line will affect connections, ladders, peeps, sente, and endgame elsewhere on the board.

Our main weapons are tesuji, tedomari, and a fanatical devotion to the Honin- wait where was I? right. So obviously 1-3 are just reading, pure and simple, 4 and 5 aren't the same as reading, but they do piggyback on solid reading, and 6 is a combination of the two. So when you get into time trouble because you're looking for the biggest move, you should think of it as a form of reading or an aspect of reading.

hmmm, I'm confused, are you doing tsumego on the board?

no not anymore, this was like 8 months ago when i first started doing problems

ok so wait i have been thinking about this. correct me if this new interpretation is wrong.

tsumego is about reading, but it is not so much about brute force reading which is what i once believed -- being able to read out responses to all variations. that is important but, just because you did that successfully does not mean you are done with the problem and ready to move onto the next one, but you also then must look and study the shapes that made your moves possible and try to identify those shapes, proactively rather than passively\unconsciously (because not many shapes have sunk in this way)

If you want to make sure your move works you must do a kind of brute force for all the opponents moves.
Once I have the/a solution (and I find the problem/shape interesting to work on) then it is a good exercise to verify that there are no other move.

When you actively want to understand the "nature" of the shape that made the move work, you can change the position a bit: add or remove a liberty and/or put a hane in place and see what this change.

Probably this is more efficent for local shapes than for wholeboard problems. It is very helpful for understanding the L+1 shapes or the carpenter's square.

If you want to messure the absulute reading speed maybe reading out a ladder would be best for, because you cannot gain speed by skipping variations or guessing the vital point.

Brute-force is necessary only if a) one does not know any applicable short-hands or b) no short-hand applies (rare). In particular, to achieve an optimal(!) result (such as 'unconditional life for the defender' in a local life+death problem about exactly one group) , a player must, at his turn during the reading, find AT LEAST ONE fitting move; the still unread alternatives can be pruned (unless also side conditions, such as 'maximal territory' and 'minimal number of remaining ko threats', must be fulfilled). To continue the example for the attacker, at his turn, he must find at least one move preventing the player's unconditional life. If, however, unconditional life is not the solution, then EVERY move must be read (unless other short-hands can be applied) to verify this.

Additionally, as one gets better, it is sometimes necessary not just to live, but to pick the best way to live out of multiple options.