Malweth’s Metablog

Sorry if I’m just rehashing an old point, but this is how Tsumego should be studied…

(White to play)
<%image(20060424-SP32-20060424-080718.gif|228|325|)%>

There’s one point in this problem that sticks out more than any other… if you’ve done a lot of problems and especially tesuji, you probably know what that point is!

Now… here’s the wrong way to answer this problem:

<%image(20060424-SP32-20060424-081523.gif|227|324|)%>

You choose W1 and your handy Tsumego program gives you B2 (or worse just says “correct” and moves on). Yay, you’ve gotten the correct vital point. Every problem has many vital points… finding the correct one (using this method) is usually a matter of luck (though some spots may look better than others - this trains you to figure out what the problem creator was thinking).

Once you’ve been given B2, you don’t have the original problem… you have a brand new problem with brand new vital points. Don’t think that this is a good thing! You’ve just made the problem easier by one level of reading, only through a bit of intuition and a lot of luck! Note that building Tsumego intuition is good, but this can be done without cheating yourself out of the rest of the problem - this also doesn’t apply to VERY simple “center of three” problems that involve only basic vital points - though these problems must also be understood.

The correct way to solve this problem is to KEEP YOUR HANDS OFF THE MOUSE/STYLUS!!! It’s pretty easy to see that black has 2-3 good answers and 2-3 more mediocre answers to White’s first vital point. What happens when W plays after B2 for each of those possibilities? What if W starts elsewhere - after all, B’s response looks pretty solid! How can W make two eyes here?

After I got this far (in my head, of course), I was trying to capture the center black stone for an eye… It certainly looks like the problem’s all about that! But the tesuji doesn’t quite work, so where are the possible eyes? There are really only three places eyes might be found… and one of them is already yours:

<%image(20060424-SP32-20060424-082613.gif|226|323|)%>

Next… I start thinking about threats. If W plays to the left of two, what is threatened? White connecting above 1, of course… which means Black must play here (think tic-tac-toe… If there are Xs in two corners, with nothing between, where must O play?) That brings us here:

<%image(20060424-SP32-20060424-082906.gif|230|328|)%>

Now it’s pretty easy… there’s still one more step, but there’s really only two threats on the board… capturing the center Black stone from either side. One of those makes a partial eye elsewhere, the other does not. Choose the superior move, and Black has the option of breaking the eye at B12 (note that the UL eye is still safe) in which case W captures to make eye no. 2. Or Black connects and W finishes the eye as shown below:

<%image(20060424-SP32-20060424-083401.gif|226|327|)%>

The game of Go is a martial art. The dependancy on mind is equivalent to the dependancy on body. Each person’s natural skill at the game is different, but a poor mental image is as bad as an unhealthy body.

I used to be a swimmer with an average of about an hour of practice a day (or more). This isn’t to say I was olympic (I don’t have near enough raw talent for that), but I was in shape. A college career and a few jobs later has made me a true American (though at least recoverably so). My new goal is to practice my martial art daily - strength must grow with tsumego, but inner strength is needed as well.

I’ve taken some Tai Chi (enough for me to do damage to myself without more classes). Unfortunately classes had to end (my teacher’s lease ran out) - but I realize I must continue daily practice until I can find a new class to join. I’m not a complete idiot, but this martial art is hard to learn out of a book. Luckily I’ve gotten a lot of the early basics already and know what to think about as I practice on my own and try to figure out the 10 or so forms that I hadn’t learned yet (I got up to about 70 on the Yang long form, but much of the rest is similar, and I’ve followed along in class).

First and foremost, this is NOT a goal! I vow to complete (and attempt to remember without using notes) the entire loosening exercises (qi gong) and tai chi form. I also vow to complete an entire practice before I eat anything for the day (ergo the goal is to complete practice in the morning rather than late at night).

I have been on a whole grain kick, but I’d also like to push this even further… this might be harder because, although I’m not a daily fast food eater like much of this country, I don’t think I could give it up forever! More on that after I’ve implemented the above phase of my training.

Here’s another example of “hidden” tsumego elements. This one is a bit more obvious, but it can still be tricky.

First, a bit of intro is necessary… I consider “basic” tsumego to have one simple element that defines the shape of the answer. For example, “center of three” is my name for playing at the center of a three spaced eye. Many of these names are recognizable, defined on senseis. For example, a “Bulky 5,” and the “hana-roku” (flower six or rabbit six)

Here’s an example of one of these problems… the goal is to understand what type of shape this problem involves… Black to play

<%image(20060408-SP32-20060408-201414.gif|196|224|What kind of problem is this?)%>
Figure: Black to Play

of course it helps if you know what the answer is…. see below:

<%image(20060408-SP32-20060408-201617.gif|195|230|Answer)%>

So, you may have expected that the core of this problem is not a bulky-5! Nope… see the continuation for the answer:

<%image(20060408-SP32-20060408-201806.gif|195|230|Continuing...)%>

In case you didn’t recognize it, this is not a seki. It’s a “Bent 4 in the corner” which kills the white stones. The object of the Bent-4 is that White cannot play first, so Black can fix all ko threats before initiating the following sequence:

<%image(20060408-SP32-20060408-202218.gif|195|230|Bent 4 continuation...)%>

After 6 and 7, Black can play 8. White captures, and B begins the ko:

<%image(20060408-SP32-20060408-202342.gif|195|230|Ko, but...)%>
Ko, but no threats should be on the board!

Here’s an example of a tsumego problem that holds hidden meaning.

How many solutions are there to this problem (White to play):

<%image(20060408-SP32-20060408-195129.gif|167|234|White to play)%>

—

Stumped? There is only one solution. The correct solution is:

<%image(20060408-SP32-20060408-195458.gif|167|227|Only solution.)%>

“Why is this the only solution? Doesn’t the hane work?”

Yes… the hane works. In fact there’s no ko. Either way, White is dead… the difference is subtle, but if you contemplate the problem enough times, you will understand what it is. It’s like a koan — except there is an answer. The answer is shown in the next two diagrams:

<%image(20060408-SP32-20060408-200031.gif|193|226|Diag. 1)%>
Diag. 1 (Black 4 at White 7)

<%image(20060408-SP32-20060408-200054.gif|161|228|Diag. 2)%>
Diag. 2

The difference is in the number of ko threats. This can make all the difference in your games!

Go is like building blocks. You start out with two blocks and you can make a tower. Start adding in more blocks in varying shapes and amazing creations can result. Life and death isn’t one of those blocks… but the shapes involved are!

There are hundreds and thousands of different shapes that generate or take away eyespace. One of the intermediate techniques is the “J-Group,” which involves a shortage of liberties to take away an eye in sente. An advanced application of this takes away one of two eyes in sente or both eyes in gote. This means that a group that has a potential of three eyes can be reduced to only one!

These building blocks are important to dermine the health of your groups. One of the biggest complaints I’ve always had with this is when it’s more “open ended” - for example, your group is surrounded and you have a very loosely defined area… there are many more paths that can be taken to reduce your eyespace. This can be considered the most advanced application of Life and Death - something we become better at as our reading becomes better.

The question I have is where the difference between the previous example and advanced tesuji problems lie. Personally, I think there is no difference. In tesuji there are other common, simple shapes that become more advanced with tougher applications. A cut is simple when attaching and extending, but a cut of a loose group can be more complicated. With regards to loose Life & Death: the shapes are different, but the process is the same.

I find that this really does relate to actual games. I play frequently on turn based go servers, but I find that I play faster on those servers than I do in face to face or even real-time online games! Why don’t I sit back and ask myself the questions I should be asking (can XYZ be cut off? Is ABC in trouble and need to live? Will I be cut off soon?) I trust my intuition and end up losing games that other people decide to sit back and actually analyze. This isn’t to say that “playing for fun” isn’t ok, but the strive we have is to improve our play, not to become automatons…

To some degree I believe that fast play is essential. It is a test of our abilities under the stress of time constraints. If I can’t read through a solution in an adequate time (10 seconds or so in most games) than I need to fall back on intuitive play. This should not always be the case, though. When the situation warrents it, and when the time is available, the situation should be read out to the best of my ability. This falls into line with my theory on tsumego:

There are three types of tsumego, but to make things simple you can even split it into two types:

Type I: problems that, according to your ability, range from “my god that problem is so easy I don’t even have to think” to “hmmm… tricky, but you can’t fool me that easy” (e.g. 5-15 second problems).

Type II: problems that, after a minute or three, have you pulling out your hair saying “I must be a fricking moron, but that problem is totally impossible.”

These problems should be mixed! Preferably approximately 90% (Type I) to 10% (Type II). Both problems should be read out completely before even thinking about looking at the solution (and Type II problems should have the solution under lock and key until you’re positive you know what the answer is). I’m toying with the thought that these problem types should be mixed. In other words, you should complete ten Type I tsumego, then spend a max of 1-2 min on a single Type II problem. Read out as much as you can on the Type II, but GIVE UP on it (don’t look at the solution) and move on. The purpose of these problems isn’t to answer them! The process of trying as many things as you can think of is most important (knowing the answer from its book solution will prevent this process).

Therefore, my L&D/tesuji workout includes:

Type I) Childrens Korean Go Institute Problem Book 3+4 (section 2, L&D)
Type II) Segoe’s Tesuji Dictionary Volume lower (3rd Volume on Sensei’s)
Other) Fujisawa’s Tesuji Dictionary Volume 1-3

Fujisawa’s is “other” because, firstly it’s on order, and secondly I believe it is supposed to be a treatise on tesuji with examples more than a problem book… we’ll see when I get it (look for the review)!

So I finally broke down and started a blitz account - mainly because I have no time to play lately and no desire to play a long drawn out game! You know what the problem with KGS is, though? ? ? ?

Ok, I’ve probably said this a million times, but KGS has a major problem - ranks. I know there are lots of blitz players out there, and a blitz game takes a minimal amount of time (especially for those who play a million of them a day). So why are people so damned afraid to play a “?” account? It’s a blitz game. If I win, my rank goes up. If you lose it probably doesn’t affect you at all. If I lose it still HELPS ME GET A RANK.

I offered a match to a number of people who had blitz games up, but all were declined. I left my game up for a while with only one hit — and he was new to Go (supposedly - he played a quite good opening) and was only playing his 3rd game (at least on that account - the others were with bots). Unfortunately I had to keep giving him time and still won W+T.

I just don’t understand what the problem with “?” is. KGS was always touted as the “friendly” server, but when it comes to actually playing games it misses the “friendly” mark completely. I guess that once I’m rated I won’t have such a big problem… as it was, I had to play a bot to at least upgrade my rank to “13?” instead of just “?” - maybe that will work as a starting point.

I personally think that KGS should get rid of the “?” competely. Imagine there’s no “?” after your rank (you can see it but no others, perhaps). This means that by playing a ranked bot you could get an initial ranking (or just set your own, novel idea IGS & others) and start at [13k], no question mark. Other players would be more willing to play you, not knowing you wont affect their ratings at all (not until you have enough games under your belt) which means you’ll have a much better way of GETTING more games to gain a solid ranking.

What are the drawbacks? The only one I can see, from the point of view of solidly ranked players, is that they won’t know just how much a game will affect their ranking. From my point of view this isn’t a drawback. I’ve broken away from the incessant need for a higher ranking (yeah I want shodan and above, but that’s a long term goal that I’ll set at AGA rated tournaments). The whole concept of online rank is the thorn in Go’s paw. Blitz isn’t much better, but playing games with long time control is more emotional. If you lose, you’ve lost all your games for the day - it’s hard to pick up a new opponent and play again. You haven’t wasted the last hour and a half (on the contrary…), but if you played a poor game it feels like a waste. In blitz they’re all poor games and you don’t learn as much, but it’s much easier to pick up after a loss and play again…

… perhaps it sounds shallow, but the choice is between playing one long game poorly or one bad blitz game and one good one. This choice often turns into the choice between playing two blitz games or not playing at all. (Where I stand now, I also think blitz will help me regain my fast opening - I lost it somewhere in the “territory is important” realm, and I can’t seem to find it again).