dutchie's journal

[EdLee]

Hi dutchie,

Almost a pass.

W almost passed with , so that's good for you.
( AlphaGo would have some good candidates here. )
Tricky: if J13 is enough for B to win, that's one possibility...

Did you consider G16 hane first, before here ?

Did you read E16 first, before here ?

The hane exchange ( - ) may be OK, but did you consider M17 first, before ?

Strange.

Of course, sente is important -- which is why you shouldn't have connected on --
you could've tenuki'd.
It's a good idea to get a rough assessment of the board, before .
( My feeling is B is comfortably ahead. If the final score was only B+1.5, then something bad happened from here. )

Your idea here is OK, but your shape is thin. ( e.g. C13 is still big. )

If you played here yourself on , you would also be nicely connected.

Did you read F12 ?

Did you read L10 atari ?

N14.

Did you read L9 ?

101 Did you read L9 ?

[Schachus]

Ed Lee is right that something must have gone wrong and it was the sequence in the middle starting at f5.
I would like to point out though, that this couldnt have happenened if you had earlier played some throw ins in order to take your opponents eyes and liberties away. It wasnt strictly needed, since you kept is group dead, but it would have given you an easier time, if you reduce his liberties, so that capturing races in case of a cut like f5 are less scary. In fact, for example at move 130(the same trick applies in easier form already at move 127 and 129), you can even capture in a ladder using those throw ins(!), though ins, I thought this was nice:)

go to move 131.
also check move 147

[dutchie]

It's been a while, and I have nowhere near stuck to my playing goals, but I'm back! Played a couple of games tonight, enough to go from 15k? to 13k on KGS. The first was very interesting, featuring a triple ko (that was definitely played incorrectly by both of us!). The second was a bit more of a steamrollering, finishing in a 188 point win for me. Who knows how far off my start-of-year target of 5k by January I'll be, but maybe it's achievable? I'll certainly be glad when I've moved up the ranks enough to have an easier time finding games.

[dutchie]

Another long gap between posts. Looking back on my goals from last year, they probably would have been achievable if I'd stuck to my plans. Still, I think I made reasonable progress overall.

What have I been doing? I've worked my way through GGPB vols 1 and 2, and made several starts with volume 3. I've found that some of the problems are beyond me and have been restarting it before I get too frustrated, although it's been a while since I've done any from it recently. I've also been doing the odd puzzle with the Tsumego Pro app on my phone when I have some time to kill. For whatever reason, I feel like I have a bad habit of just guessing moves on that because of the immediate feedback so it doesn't feel as useful as the books.

I also bought Bruce Wilcox's Contact Fights book/tool/thing based on lots of recommendations aroudn L19 a couple of days ago and have worked through the Elementary and Novice sections. And wow, it really does feel worth the money already! I played a few games trying to put it into practice, and after a few losses it's really started to feel like it's coming together and making sense. I've now managed to get what seems to be a fairly stable 9k rating on KGS, only 9 months after my initial target. I'm definitely really happy with that. And I feel like I'm enjoying playing a lot more than I did before too: no longer finding the losses as frustrating, and it's getting much rarer for them to be decided by a stupid misread in endgame or anything like that.

So for the foreseeable future I want to get back into GGPB3 and just keep playing! Maybe I'll be able to catch up on my goals from last January and be 5k before autumn comes back?

[dutchie]

I've played a couple so far, but I'm still yet to beat an 8k. Here's tonight's game, with a few variations and comments:

[mitsun]

Shortage of liberties is dangerous. In your variation with double-hane, W cannot cut, because B has a killer answer -- a tesuji which frequently appears in problem books. I am sure you can find it if you look. The key to looking is to notice the shortage of liberties.

was the critical win/lose fight of the game. There is not enough room for B to live outright in the corner, but W is not strong enough on the outside for a clean kill. at B2 would give B a ko for life, which comes up often after invasions like this. You might also consider invading at C4 instead of C3, since you have some strength above if W allows you to run that way.

[dutchie]

Thanks mitsun. I can see that tesuji now, but I'm not sure if I'd spot it in a game without someone pointing out it's there! On the other hand, I can't see W's move to prevent B from living outright in the bottom left. Can you (or someone else) show some variations?

[dutchie]

Here's a game vs another 9k where I spent most of it thinking I'd lost, to be amazed when KGS told me I'd won by 7.5. I've gone through and reviewed it, and there's a nice little L&D puzzle at the end that I think could have been seki if I didn't defend my cut.

[Knotwilg]

Conclusion from this game: it looks like you are great at playing big points and often play away from a local fight (tenuki) to do that. Sometimes too soon for my taste, but most semi-beginners tenuki not nearly enough, so keep the good habit. However, develop your sense of weak groups and how to attack them (in particular): 1) look for cuts 2) surround 3) reduce their space 4) play at their vital point

12: your sense of direction is good, but I believe something else is called for here: mutual strength in local combat; and so I favor another move
28: I would play the same move as the one I had in mind for 12, also because the direction of your influence is not so interesting here
36, 38 (and subsequent): I believe there's a much more urgent area than this
52: somehow, neither player recognizes the potential left here; only very late the move that would catch most players' eye is played
62: when you push through like this, both players hurt their own stones; now which one's stones affected are more important here? This is regardless of not seeing the atari.

76 to 98: after the loss of your cutting stones, you start playing very well. By 100, the damage is largely undone.
100: However, I would not miss out on an opportunity for sente here, locally double sente.
104: You know shape, you're in for progress
106: you play large points in early endgame, you're in for progress

126: with the wall you built, the move that has been asking to be played since 52 is now screaming



(BTW, you are right that the position in the end would result in seki if you don't respond, but this is due to lack of ..., otherwise Black would go for a ...)

[dutchie]

Thanks as always for the thoughts, Knotwilg.

My routine's been a bit out of sorts recently, and certainly things like tsumego and playing have fallen by the wayside a little. But I was with some family in a board game café over the weekend and I talked us all into a little informal rengo and taught them the basics on a 9x9, which has rekindled my interest. I have bought and read the first couple of chapters of Attack and Defense, and splashed out on a goban and stones for myself. More excitingly, I've also signed up for my first tournament, the 21st Bar Low in London in a couple of weeks. I think I'll try to get to a club session or two in nearby Winchester beforehand to get some more face-to-face practice in, subject to nothing coming up on Wednesday evenings.

I played a KGS game against an 8k just now, and annoyingly I lost on time thanks to the Shin KGS interface not beeping at me when I was using up byo-yomi periods. It's nice to be able to play in a web browser with a somewhat more modern interface, but the lack of sound for time is really problematic, seeing as I tend to get lost in thought at the best of times. The game wasn't looking great, but I think it was pretty interesting whatever the result.



I think things to work on here are time management, reading/L&D, and making sure I'm not just playing moves without even considering possible responses (eg move 123)

[dutchie]

Today's game was with 2 stones handicap against a 7k, which I lost after a big group of mine died, and featured numerous other L&D mistakes. I definitely need to get the tsumego habit going again.

[dutchie]

Last night I went to the club up in Winchester for the first time, and it was great! It was in a very good pub close(ish) to the bus station, and I played a game against one of the members, an 8k, and also the club secretary is going to pass on email addresses for some other players closer to me in Southampton. Unfortunately both of us liked to take our time over moves, so the endgame was a little rushed as the pub was closing, but it was a very close game. My opponent recorded it and sent me the SGF, so I've reviewed it below. My theme (once we got into the midgame around the time of the centre invasion) was to play moves on the side of overplay rather than conservatively, and I think it went pretty well. Any further comments from people more knowledgeable than I are welcome.

And just 10 days to go until the tournament! It's nice to be confident that my estimated rating is pretty close to what I seem to be playing at.

[Uberdude]

- move 13: yes, h17 normal locally, but play lower side fine too. White doesn't need to add a move, so it's not like you need to rush to punish him for being naughty.
- move 15: blocking is slow, you let white get 3-3 in sente; either extend to r9/q10 to help your group, or tenuki and be prepared for white to attack with a pincer.
- move 19: when the 4-4 already has the extra stone, and they have lower side extension, knight's approach is probably not a good idea (unless you are a super strong bot). d3 attach is a standard idea: you either get a chunk of the corner to help you settle along lower side, or go into corner. 3-3 invasion also possible, the outside isn't so big with a nice reduction point at k4 for later.
- move 27: yes pincer first so white is under pressure. Also o4 rather than o5, lean on the strong group to be more forceful as you want to get out in sente, not gote (or if they ignore you want a big follow-up, o5 doesn't have one). Your problems here stem from the r4 block mentioned before.
- move 35, not really clear what this is doing, you can even go as far as f18.

And just 10 days to go until the tournament!

Bar-low? I'll be at the Candidates which runs parallel.

[dutchie]

- move 13: yes, h17 normal locally, but play lower side fine too. White doesn't need to add a move, so it's not like you need to rush to punish him for being naughty.
- move 15: blocking is slow, you let white get 3-3 in sente; either extend to r9/q10 to help your group, or tenuki and be prepared for white to attack with a pincer.
- move 19: when the 4-4 already has the extra stone, and they have lower side extension, knight's approach is probably not a good idea (unless you are a super strong bot). d3 attach is a standard idea: you either get a chunk of the corner to help you settle along lower side, or go into corner. 3-3 invasion also possible, the outside isn't so big with a nice reduction point at k4 for later.
- move 27: yes pincer first so white is under pressure. Also o4 rather than o5, lean on the strong group to be more forceful as you want to get out in sente, not gote (or if they ignore you want a big follow-up, o5 doesn't have one). Your problems here stem from the r4 block mentioned before.
- move 35, not really clear what this is doing, you can even go as far as f18.

Thanks for the comments, useful things to think about!

And just 10 days to go until the tournament!

Bar-low? I'll be at the Candidates which runs parallel.

Yep, I guess I'll see you there!

[Knotwilg]

225: Black has an even better move.