My week in numbers (was "Tiny steps towards shodan")

[RBerenguel]

Yes, these were the books!

Sadly, the online catalog from my local go bookstore is outdated... I can't have Maeda's Tsumegos or Rin's Come up to shodan, two books I really wanted

I got 1001 L&D problems, I've alread gone through 40 or so. Nice book!

[RBerenguel]

Got some time today, and played two games, one as white 7k against an 8k, and another with two stones against a 6k.

Played horribly. I paste them below. I have not reviewed them yet, not even replayed to see the big blunders. Although in general, bad DoP decisions in the first, and failing to end with a group in the second. If anyone has further comments, I'll be very thankful.

http://files.gokgs.com/games/2011/12/8/RBerenguel-tooryanse.sgf

http://files.gokgs.com/games/2011/12/8/igoholic-RBerenguel.sgf

[mitsun]

A few comments on the LL position in the second game:

Move 18 (M5) is way too cautious, giving your opponent an opportunity to defend below where he is extremely weak. Your wall to the right is a source of strength which must be to be used to attack! Invade immediately at J3 or K3 and split W into two groups. Once W defends at L4, this golden opportunity is gone. Invading now at J3 loses most of its effectiveness, as W has a perfect response to stay connected. Just think how much better your attack would be around here if W did not have the L4 stone.

Over the next few moves, B played an invasion sequence which did not work, ending up losing three stones and giving up quite a few points. Resolve never to let this happen in future games! Try to read out this sort of position before you invade. If you think the invasion works, then of course go ahead and play it. But if you can read that it does not work, you can make other plans and avoid a big loss.

Reading out this sort of position, so you can play invasion sequences which actually work (according to your best reading), is one of the best ways to get stronger. If nothing works immediately, maybe the best plan is to play elsewhere for now and later prepare patiently for an invasion which does work.

For move 22, consider the B hane at H4 (the game move). In this position, W just cuts at H3 and B has no way to save the J3 stone, so W stays connected and strong. It takes quite a bit of reading to reach this conclusion, but try to put in that effort. Playing a sequence like this when it does not work incurs a large loss, compared to leaving the situation alone.

Since the first plan does not work, let's consider a different plan at move 22: bump at H3, planning to crawl at G2 if W connects above. But wait, after you read ahead a few moves into that sequence, the result still does not look good (if B keeps crawling to live, while W keeps blocking above, the B stone at D4 gets cut off). So don't play that sequence either.

A much better plan is to analyze why these sequences do not work, then look for a useful move which would make them work. For example, B could look for an opportunity to play E3 while the bottom side is still unsettled. This dual purpose move would help secure corner territory, and it would also threaten to activate the J3 stone in a way that does work.

[RBerenguel]

Hi mitsun,

thanks a lot! This is something I have to work on... Even if I can read decently tsumego-tesuji problems, when I'm playing I tend to lose focus too easily and just rely on (bad) intuition. Something to work on A LOT!

[Numsgil]

Hi mitsun,

thanks a lot! This is something I have to work on... Even if I can read decently tsumego-tesuji problems, when I'm playing I tend to lose focus too easily and just rely on (bad) intuition. Something to work on A LOT!

If you keep drilling on the problems, eventually they'll be so easy for you that you'll actually see them in-game

There's something about an actual game vs. a position presented as a problem that makes players put blinders on. As far as I can tell it's universal.

[RBerenguel]

There's an Spanish saying about this "Mal de muchos, consuelo de tontos" (rough uninspired literal: the fact that it affects a lot of people only reliefs the dunces)... At least I know I'm not alone with that, but I hope to get over it soon, I feel pretty dumb about it

[RBerenguel]

Today and yesterday I played two games against Igowin HD (quicker and more relaxing than firing KGS.) Won both games (by resignation, I was winning by more than 40 points in both), raising my estimated rank in the app to 5k. I wonder why the app sometimes refuses to defend about-to-die groups.

I did 5 of the problems at the end of the first chapter in The Endgame. I got them decently well, would have at least made the second best choice in all. But I still have trouble realising what really is sente and what really is gote (and worse, reverse sente). I guess I'll improve with re-readings.

I have also finished reviewing Yi Ch'ang-Ho's tesuji vol 1, got 23 problems wrong (out of... 129 or so?) Tomorrow I'll check what each tesuji was about and why I misread. I'll try to spot patterns in these failures.

[oren]

Today and yesterday I played two games against Igowin HD (quicker and more relaxing than firing KGS.) Won both games (by resignation, I was winning by more than 40 points in both), raising my estimated rank in the app to 5k. I wonder why the app sometimes refuses to defend about-to-die groups.

It probably refuses to defend in order to play like a 5k.

[RBerenguel]

I'm supposedly worse than 5k and I do try to defend my almost dead groups

[oren]

I'm supposedly worse than 5k and I do try to defend my almost dead groups

I'm 5k and tend to have a lot of dead groups. You might be doing something wrong.

Anyways, computers just play different. I wouldn't try to analyze it too much.

[RBerenguel]

Won't analyze, it's just puzzling If he leaves the groups in the open, go for the blood!

[RBerenguel]

Seriously Igowin HD (iPad 1), what's up with you? AI Level 5k, and I was black

[SoDesuNe]

It runs on an iPad :p

[RBerenguel]

It should be adjusting the level nevertheless.. It's not like the iPad does not have a processor (and Igowin's answers were pretty fast nevertheless)...

[Sur]

Don't play computer go, it just messes you technique. Computer is nowhere near human when it comes to go.