Filthy casual training

skydyr · Post by **skydyr** » Fri Feb 06, 2015 2:46 pm

Some thoughts:

As an aside, I don't want to discourage you from posting your progress in problem collections, etc., but I wonder if the volume of those posts discourages discussion of your more discourse and game oriented posts.

tentano · Post by **tentano** » Fri Feb 06, 2015 4:56 pm

First off, thanks for that review.

Several of the things you mentioned didn't seem obvious to me, although especially your suggestion for 116 seems like something I SHOULD have seen.

Since it's a bit old, I wonder how you feel about my play a month later. Is it more or less the same? Does it seem to have a little more bite?

WORST game of today:

This contains sincere derp. Made a very large plunge in the top left quadrant. I really should have resigned, it deserves the title of worst game.
I think move 127 should have been spent saving that group.

BEST game of today:

Not sure it is THE best, but it's a clear demo of how often it's all about a single big capture. It's not worse than the other wins by much, if it is.
I felt very much in control for last 50 moves or so.

I seem to have a set routine. 1) mess up the opening a little bit, 2) ???, 3) big capture to win or lose. It feels a bit artless, but not so artless that I would give up on winning games with it.

Looking at these two specific games, I also notice I spent less time on them than I thought. I was sure I spent around 15 minutes of my time, like I usually do. Wonder if I should make myself slow down a little more. The derp in the worst game should have been avoidable with a little more forethought, after all.

Second, I've actually been thinking about the value of those progress updates.

They really do help me stay on point, but they also seem a bit spammy, and I can't imagine they're the world's most interesting reading. I'd either solve it with a second thread (such extravagance! :oops:

) or maybe a different frequency (weekly?). Somehow, even if nobody actually cares or reads, pushing them here makes me feel a little bit more accountable.

skydyr · Post by **skydyr** » Fri Feb 06, 2015 7:56 pm

tentano wrote:Second, I've actually been thinking about the value of those progress updates.

They really do help me stay on point, but they also seem a bit spammy, and I can't imagine they're the world's most interesting reading. I'd either solve it with a second thread (such extravagance! ) or maybe a different frequency (weekly?). Somehow, even if nobody actually cares or reads, pushing them here makes me feel a little bit more accountable.

I agree completely, and in that respect, it can be quite valuable.

As for the games:

Honestly, in the second game, black had a won game and threw it away with a series of weak, passive, and otherwise slack moves. Being able to take advantage of these is good. Not playing into a situation where you rely on your opponent playing this way is better.

You seem to have the bloodlust, so I would recommend playing games where you explicitly aim to kill nothing and win on points. That doesn't mean you don't attack, but you need to have a goal for each and every attacking move that does not include killing something. Killing should be a last resort, though it may at time become necessary because your opponent is being that unreasonable. In general, it's easier to live than to die, so you have to neglect a group to some degree before it becomes killable, even.

I will tell you, one of the things I fear most in a game is making a kill early, because against anyone worth their salt it's too slow to remove all the aji and it will come back to life and haunt you in the worst way, or you will spend the rest of the game stepping around kos, etc. that you can't afford to fight because of the aji, while your opponent will be making things as complicated as possible to try and come back. Additionally, you'd be surprised how big a group you can lose in exchange for sente and some outside moves and still come out ahead. I've had games I won because I sacrificed a 40 point group early on. Some of the outside stones got caught in a semeai to kill and I consolidated half the board in the process of letting my opponent capture it.

As you get stronger, also, games will be knockouts less and less frequently, so you'll have to be able to keep it up for all 10 rounds and win on points.

In terms of reviews, though, it can be more helpful to show games you lose, and sometimes particularly those you lose small.

tentano · Post by **tentano** » Sat Feb 07, 2015 3:50 am

Thanks for those reviews. Seems like you pretty much share my self-diagnosis. Even the ??? part. :oops:

It's also very interesting to see I still make relatively simple tactical mistakes. I should be able to do better there with only a little effort.

I'm happy that you don't generally see that I'm making impossible invasions, but rather that opponents leave the door open for me to make them. That means I should naturally get to the point where evenly ranked opponents no longer leave me that much room, and I had better become as good at the opening as they are. You (and others) have been quite clear that my opening proves I don't know what to choose, and I don't feel very effective in the first 20 moves or so.

Avoiding a big kill, though. That's an interesting challenge. I'll try that a few times this week. I expect I'll feel like I'm letting go of perfectly capturable groups, but that's what makes it interesting.

My problem in selecting recent games I lost by a small margin is that I didn't actually have any. They're all in the same vein as those two, where something big dies and it settles the game.

So, in addition to the no kill ~~pet shel~~ go game, I should go for a smaller margin of victory. These other purposes also come to mind: How do you practice yose if you nearly always force a resignation? How do you practice judging the balance of points if you just fire the artillery until someone submits?

You've got to admire a game where it's completely natural to be winning most of your games and you still find a whole slew of faults with how you play. It guards against complacency, but it also guards against submissive acceptance of limitations. There are things I could do, to play better.

So this week, I'll try to create at least two games which meet these specific criteria:

1.Ends with scoring.
2.Less than 15 points difference.
3.No kills of groups larger than 10 stones.

I'll post them in the review section, next saturday, whichever two come closest to both looking good (to me) and meeting these criteria.

Bill Spight · Post by **Bill Spight** » Sat Feb 07, 2015 4:49 am

tentano wrote:I'm happy that you don't generally see that I'm making impossible invasions, but rather that opponents leave the door open for me to make them.

In the first game the invasion on the right side with :b39:

looks premature. Also, even though it is on a vital point (S-09), the better way is to start with the peep on Q-08. Then if White connects at Q-07, Black S-09 is good.

tentano · Post by **tentano** » Sat Feb 07, 2015 9:11 am

Weird how it seems obviously good if I spend only 10 seconds reading it out, but I didn't actually think of it at the time. Thanks for the tip.

tentano · Post by **tentano** » Mon Feb 16, 2015 8:20 am

Code: Select all

+-------+-------+-------------------------------------------------+--------+
| Pass  | Prob# |  Title                                          |  Count |
+-------+-------+-------------------------------------------------+--------+
|   1+  |  101  |  501 Opening Problems                           |   501  |
+-------+-------+-------------------------------------------------+--------+
|   1*  |  p98  |  Essential Life and Death 2                     |  much  |
+-------+-------+-------------------------------------------------+--------+
|   2*  |    1  |  Graded Go Problems for Beginners Volume 3      |   421  |
+-------+-------+-------------------------------------------------+--------+
|   2*  |    1  |  1001 Life and Death Problems                   |  1001  |
+-------+-------+-------------------------------------------------+--------+
|   2+  |    1  |  Graded Go Problems for Dan Players Volume 1    |   300  |
+-------+-------+-------------------------------------------------+--------+
|   2*+ |    1  |  Cho's Elementary                               |   900  |
+-------+-------+-------------------------------------------------+--------+

So far, so struggle.

501OP is not an easy ride for me. I get between 1-3 right on each page, and 1 more often than 3. I think I got all 4 on only one page. Many of the problems I get wrong don't seem more obvious after seeing the solution. There's clearly a wide gap between me and understanding this well.

Some of the easier problems I solved by reading it out. This doesn't strike me as understanding the opening, but at least some of it makes sense.

For each problem, I first try to be honest with myself by asking, "What would I do in practice?" then I try to solve it by the principles they give in the book. Then see if the hint under the diagram changes my expectation of the correct answer. Only after all three steps fail to hit on a correct answer do I count it as a failure.

And that still leaves me with around 40% failed. Maybe this is simply too hard for me?

I keep playing games with large captures and resignations, too. Apparently my bloodlust is pathological, since I clearly have some sort of aversion to peaceful endings.

skydyr · Post by **skydyr** » Mon Feb 16, 2015 11:09 am

tentano wrote:I keep playing games with large captures and resignations, too. Apparently my bloodlust is pathological, since I clearly have some sort of aversion to peaceful endings.

A sudden burst of insight (or madness): Stop doing life and death problems for a while, because they focus your game on living and dying too much. You might want to consider looking at a book of opening problems instead, if you want problems, or a book on attacking, or joseki. You'll keep improving, but ideally with a different focus to round out your game more.

Part of the capture problem, though, is the level you play at. Your opponents aren't any better than you, and they (and you) likely misjudge situations like when to tenuki from a group, or what constitutes a reasonable invasion or reduction and what you should expect from it. You might want to try playing a set of games against slightly stronger players (I've heard 3 stone games recommended for improving, or at least about that rank difference). They'll know when to punish and when to invade or not better than your normal opponents.

I'm always available on DGS if you're interested (this goes for everyone). Just mention where I might know you from if you challenge me and I can't recognize your username. I will confess, though, I probably don't take enough time with each move there (take advantage!).

tentano · Post by **tentano** » Wed Feb 18, 2015 6:27 am

I find it hard to really fault the focus on killing when I still manage a 55-65% win ratio every week for the past 6 weeks or so.

Curiously, my performance when giving handicap is insane (70% win ratio for this month so far) while when taking handicap it's about 50% (but very rare, so I'm on 4 losses and 3 wins). It doesn't paint a clear case for heedless folly anyway.

I'm quite content to allow myself to enjoy the moment until I come up a rank or two and it gets hard again.

I'm also a bit unsure if DGS is the right format for me. I play a lot of moves within two or three seconds, and I worry I might lose focus in the middle of a "fast" sequence which takes hours or days (!). On the other hand, there's not much time pressure if I can just sit back, calm down and retake focus. Even if that takes a day.

Ah, well, I'm curious now...

tentano · Post by **tentano** » Wed Feb 25, 2015 8:38 am

Code: Select all

+-------+-------+-------------------------------------------------+--------+
| Pass  | Prob# |  Title                                          |  Count |
+-------+-------+-------------------------------------------------+--------+
|   1   |  101  |  501 Opening Problems                           |   501  |
+-------+-------+-------------------------------------------------+--------+
|   1   |  p98  |  Essential Life and Death 2                     |  much  |
+-------+-------+-------------------------------------------------+--------+
|   3   |    1  |  Graded Go Problems for Beginners Volume 3      |   421  |
+-------+-------+-------------------------------------------------+--------+
|   2   |    1  |  1001 Life and Death Problems                   |  1001  |
+-------+-------+-------------------------------------------------+--------+
|   2   |    1  |  Graded Go Problems for Dan Players Volume 1    |   300  |
+-------+-------+-------------------------------------------------+--------+
|   2   |    1  |  Cho's Elementary                               |   900  |
+-------+-------+-------------------------------------------------+--------+

1.Book pass complete!
Finished GGPfB3 for the second time.

I wouldn't call it a smooth ride, but it's much better than the first time.

Demanding perfect certainty for each problem is very demanding, but it's a big difference to how I did it before. I went over this book from start to finish at least once, before my hiatus. It used to be fully impossible to get them all right.

The same pattern as I had with EL&D1 and LUR2 holds. The second pass feels less like lunatic mode, a lot of at-a-glance solving. It makes me stop and examine carefully, because I'm terrified I might be getting complacent. It's only hard mode, now, and I'm really just jumping at shadows.

Another thing I noticed with GGPfB3, the answers. Since I have to exhaustively prove to myself that each move is correct, I find the answers a touch on the sparse side. They're not wrong, but they are often very poor at explaining why something is correct. My DDK self didn't find them very explanatory.

Even now, for some of these problems, after the first two correct moves, I spent non-trivial time (30-90 seconds?) checking out possible ways of resisting the correct answer, before accepting it as correct. "After :b3:

white cannot make two eyes." is not really obvious at a glance in many cases.

There are limitations on space, since this is printed material, but often I wonder if it wouldn't have been helpful to include a few more moves. Not for me now, but for me back then, and anyone else at that level. Discovering what kind of moves are correct is enormously helpful to get you to broaden your thinking, but "Study and find them for yourself!" isn't everyone's preferred method.

Despite all that, this is now apparently the easiest book in my lineup. Two or three more passes and it won't be in my lineup anymore.

2.Progress reports
I'm not entirely sure, but I think it does slow down my work rate when momentary dips in tsumego processing rate aren't reported. There just isn't that little nagging feeling that I should make time, instead of letting it slide.

I don't know if the previous regime of daily posts is best, though. I need something different. Something less onerous, but not less effective.

I've got just the thing: time-shifted batch reporting! All the functionality of daily reports, without the spam! Pseudoaccountability technology makes a giant leap into the century of the Frui-- I mean, OUT of the century of the Fruitbat!

3.DGS
Because of skydyr, I now have a DGS account. Since one game only seemed a little dull, I added created ten zero-handicap games open to all comers. Takers range from 5k to 19k. That should count as 2-3k KGS to 16-17k KGS. Probably.

The furthest along is only 41 moves in, so far. It's a MUCH slower pace than I'm used to. This idea that you can just sleep on your next move, it's novel. I find the games rearing up in the back of my head while doing something else. Very strange way of playing.

It definitely feels like another category, so it's not at all strange that the ratings are not the same as on KGS. Playing on DGS feels halfway between doing tsumego and actually playing. It feels unnatural that I can spend 5 minutes on a single move, even if it's not a particularly important one.

I guess the biggest mental strength for DGS is ability to be patient and go over a situation until the next move feels completely right. There's no time constraint to counterbalance this. At least, if I've got all day, one single move doesn't seem like it would put me under any pressure. I can clearly deliver thousands of reasonably 4k-like moves in a single day, after all.

4.On 501OP
I don't feel entirely ready to digest this book as deeply as I want to. About a quarter of it is fully indigestible, which is disappointing, but hopefully also temporary. There isn't a lot of incomprehension one can maintain with continuing study.

So I'll take a different tack vs the usual tsumego. I'll just go over it, and let it slide that I really don't understand some of it. The three quarters which are perfectly sensible should offer some benefit, at least. Hopefully, they will provide some sort of mental framework with which to understand the rest.

PeterN · Post by **PeterN** » Wed Feb 25, 2015 11:50 am

Not having the wrong moves explained does lead at least to the amusing situation of going through the book the earlier times and being completely stumped as to why your answer won't work, and then on a later try realising it's fairly obvious why it won't work.

The one thing to remember with the GGPfB series is that the rank they state it for lies... Lies I tell you! :blackeye:

PeterN

tentano · Post by **tentano** » Wed Feb 25, 2015 12:44 pm

About that stated rank ... my copy of GGPfB3 has "20k to 15k" on the front, and "12k to 6k" on the back. It's quite clear there's something fishy going on there.

skydyr · Post by **skydyr** » Wed Feb 25, 2015 3:07 pm

tentano wrote:About that stated rank ... my copy of GGPfB3 has "20k to 15k" on the front, and "12k to 6k" on the back. It's quite clear there's something fishy going on there.

Starting and ending ranks?

tentano · Post by **tentano** » Fri Feb 27, 2015 3:35 am

In order to highlight the contrast with GGPfB, Baduktopia's approach deserves some attention.

This problem is from EL&D2:

Click Here To Show Diagram Code: [go]$$B $$ | . . . . . . . . . . . . . $$ | . . . . . . . . . . . . . $$ | . . O . . . . . . . . . . $$ | . . . . . . . . . . . . . $$ | . . O , O O O O . O . . . $$ | . . O X . X X . O . . . . $$ | . . O X . . . X X O O . . $$ | . . . . . . . . . . . . . $$ +--------------------------[/go]

This problem appears as the first diagram, it's a bit of a thinker, but for those who don't have the discipline to not look very slightly to the right (or down, in the case of this post), the answer is right there on the same page.

Click Here To Show Diagram Code: [go]$$B $$ | . . . . . . . . . . . . . $$ | . . . . . . . . . . . . . $$ | . . O . . . . . . . . . . $$ | . . . . . . . . . . . . . $$ | . . O , O O O O . O . . . $$ | . . O X . X X . O . . . . $$ | . . O X b c . X X O O . . $$ | . . . a . . 1 . . . . . . $$ +--------------------------[/go]

The GGPfB approach would just give :b1:

and say "No matter what white tries, black gets two eyes."

In EL&D2, it's given along with several potential white responses, which each get several diagrams dedicated to showing why white might try it, and why it doesn't work. The message is simple: this looks scary, but you can counter this.

Click Here To Show Diagram Code: [go]$$B $$ | . . . . . . . . . . . . . $$ | . . . . . . . . . . . . . $$ | . . X . . . . . . . . . . $$ | . . . . . . . . . . . . . $$ | . . X , X X X X . X . . . $$ | . . X O . O O . X . . . . $$ | . . X O . a . O O X X . . $$ | . . . . . b . . . . . . . $$ +--------------------------[/go]

What's this then? The same problem is shown inverted, with a bunch of diagrams spent on showing what happens if you do it wrong. They even spend a diagram on a wrong response to the wrong response. I'm hoping for a wrongception in the later two books.

So that's one single problem, over three pages, for a total of 18 diagrams. Most problems aren't shown that elaborately, but it's quite common to have three diagrams about the same problem. It feels a little unwieldy and I often catch myself thinking "yes, yes, that was obvious".

What if it's not that obvious, though? After reading out the first diagram, you'll see nothing new in the other 17 diagrams, but if your reading failed due to a blind spot, or you simply don't understand why your different answer fails, it can be helpful to see the result.

This sort of demonstration isn't aimed at people who can just read it out perfectly, but it is aimed at people who want to learn how. It's showing plainly what you "ought" to be able to see if you can read it out.

There are almost (I don't know why there are exceptions, but there are a few) no letters or numbers or other markings in the diagrams. Just some stones with a simple instruction written above each diagram, like "kill white" or "save black". The reader is always black. An inverted problem always signifies you switch between killing and saving.

The theme of every page is clearly mentioned on top. It's a very strong hint about what kind of move you should be trying to find. If you're supposed to be looking for a ko, you will be told on a per diagram basis, too.

There are no answers in this book, either. I really don't know what I'd do if I could still not see whether something is correct or not, after all those diagrams. I really treat them as answers. You might want to use a card to cover them when you're going through this book.

The only ones which aren't a case of the blatantly obvious are the little "tests" every once in a while, which refer back to specific pages. There are actual answers available, for the especially hard of learning, or the highly insecure, which you could download off the Baduktopia website.

One other thing which I should really mention, is that a lot of problems have several solutions. The foreword freely admits this, since it's not really a problem as such. If you ignore the pervasive hints, you can find perfectly workable alternatives in many cases.

The goal of this book is more about teaching you specific moves than trying to drill the one sacred correct answer, so that's why the authors haven't contrived their problems in such a way that only one move could work for each of them.

The end result is a far gentler approach to tsumego, which is a little wasted on me. I still appreciate this format, since it adds something completely different which may work very well for some people. I really wonder how much this could have helped me earlier on.

Obviously, it's not actually worthless to me now, either. It just lifts the curtain to things I've already seen with all the extra diagrams. I'm too lazy to extract all the first diagrams (without any solution or hints) into a list when they're already printed in the book.

Bill Spight · Post by **Bill Spight** » Fri Feb 27, 2015 8:02 am

tentano wrote:About that stated rank ... my copy of GGPfB3 has "20k to 15k" on the front, and "12k to 6k" on the back. It's quite clear there's something fishy going on there.

Kano (or his ghost writer) did not establish the difficulty of the problems empirically, but guessed. A beginner's thought process is just about as mysterious to a pro as a pro's is to a beginner.

Life In 19x19

Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training

Re: Filthy casual training