Life In 19x19

I still hope Yang Dingxin can get his first international title!

I'm rooting for Shi Yue
I don't understand his style, but I find it really interesting to watch. Maybe the veteran can finally win his second international title against the young up-and-comer!

Final;
2nd game,
Yang Dingxin defeated Yang Shi Yue by resign.

Yang Dingxin vs Shi Yue - 1:1

So we get a decider! The 2nd game featured a popular modern joseki at top left (and I played in my last tournament game) which ends with white getting a big corner and black a wall plus ponnuki. I was somewhat surprised Shi then 3-3 invaded at the top right as the resulting top side seemed pretty large and efficient. Is this a case of a bad 3-3 invasion? it's all the rage these days but usually that's because the resulting wall is either inefficient or attackable (or teasable with p14 peep) later but that didn't happen this time and it just made a huge territory. Or was 3-3 ok but subsequent reduction or lack thereof the issue?



For comparison here's a game I played as LZ recently on Fox where it killed the ponnuki group later (of course it could have lived in gote but that was painful so the human decided to tenuki to kill another group so LZ duly kill it):

So LZ201 (with low playouts) thinks the 3-3 was fine (-<1%, q14 high approach another interesting idea, don't often see those, f12, r3 other choices), but that o3 approach was -5% and it was urgent to reduce:

Uberdude wrote:So we get a decider! The 2nd game featured a popular modern joseki at top left (and I played in my last tournament game) which ends with white getting a big corner and black a wall plus ponnuki. I was somewhat surprised Shi then 3-3 invaded at the top right as the resulting top side seemed pretty large and efficient. Is this a case of a bad 3-3 invasion? it's all the rage these days but usually that's because the resulting wall is either inefficient or attackable (or teasable with p14 peep) later but that didn't happen this time and it just made a huge territory. Or was 3-3 ok but subsequent reduction or lack thereof the issue?

If we look at the game from the vantage point of AI, then B has a fairly clear advantage, and W should reduce. However, I think maybe this was Shi Yue's strategy, to give B what he wanted but to keep the game close enough to still win. It didn't work out, but clearly at the end the game was still quite close. I think this sort of "deceptive" strategy can work in human games where mistakes can be made, and where overconfidence can be leveraged as a strategic ploy. The AI always seems to have all of the answers to anything, so of course that doesn't work against an AI.

Uberdude wrote:So LZ201 (with low playouts) thinks the 3-3 was fine (-<1%, q14 high approach another interesting idea, don't often see those, f12, r3 other choices), but that o3 approach was -5% and it was urgent to reduce:

Screen Shot 2019-02-13 at 13.50.51.png

I also tried this with network f438268e with a lot of playouts, and it initially thought of the high approach(which I also thought was interesting), but it eventually settled on either approach or the 3-3 invasion to the bottom right corner. It also felt that the 3-3 invasion on the top right was relatively okay(1-2%). Maybe it's because the M11 reduction is so good.

Final;
3rd game,
Yang Dingxin defeated Yang Shi Yue by resign.

Yang Dingxin is new LG cup Champion.

Yang Dingxin just sets an interesting record: after winning game 2 yesterday he was promoted to 8-dan (because of winning enough games as 7-dan). And now he is a professional 9-dan after winning this major title (due to China's special promotion rule). He was 8-dan for only one day.

I really liked Yang's reduction of lower left corner, and in sente to boot! (Though he did soon come back to save the tail stone). Seemed quite AI like to me.