Life In 19x19

LZ thought there were some mistakes in the early fights around 10%, but it's quite rare to have almost 50% ones in pro games.

Iyama beat Yamashita in the 4th Kisei game to take a 3-1 lead.

Yamashita faced kadoban in the 5th game, and after the first days' play (just 49 moves) bots had Iyama with a nice lead of >70% depending which one you asked. But Iyama made a critical mistake in the middlegame (solid extend instead of knight move) which meant Yamashita got a nice squeeze to turn the game around and a big ko trade later couldn't change the result. so Iyama leads 3-2 now.

Uberdude wrote:Yamashita faced kadoban in the 5th game, and after the first days' play (just 49 moves) bots had Iyama with a nice lead of >70% depending which one you asked. But Iyama made a critical mistake in the middlegame (solid extend instead of knight move) which meant Yamashita got a nice squeeze to turn the game around and a big ko trade later couldn't change the result. so Iyama leads 3-2 now.

Don't you mean it is 4-1 now and the title match is over?

I read Uberdude as saying that Yamashita took a lead and Iyama was unable to take the game back. In that case, Iyama lost game 5, and the series is now 3-2.

DoubleSente wrote:I read Uberdude as saying that Yamashita took a lead and Iyama was unable to take the game back. In that case, Iyama lost game 5, and the series is now 3-2.

Oh, so sorry - I followed this game with swapped colors in my mind!!
I was feeling very sorry for Yamashita blowing the game after such a strong opening, now I need to revise my feelings

sorin wrote:

DoubleSente wrote:I read Uberdude as saying that Yamashita took a lead and Iyama was unable to take the game back. In that case, Iyama lost game 5, and the series is now 3-2.

Oh, so sorry - I followed this game with swapped colors in my mind!!
I was feeling very sorry for Yamashita blowing the game after such a strong opening, now I need to revise my feelings

My revised feelings: Yamashita masterfully turned the game around and got a well deserved win!

I am not necessarily a big fan of Yamashita, but I think Iyama won enough titles already, and it is time for him to pass the baton back to the older generation.

sorin wrote:Oh, so sorry - I followed this game with swapped colors in my mind!!

Couldn't you tell black was Yamashita from the unconventional high move 7?

Uberdude wrote:

sorin wrote:Oh, so sorry - I followed this game with swapped colors in my mind!!

Couldn't you tell black was Yamashita from the unconventional high move 7?

Both black 7 and 9 seemed unconventionally high indeed, I thought Iyama is so relaxed he can afford to play around and experiment a bit

Iyama beat Murakawa Daisuke in the first game of the Judan title match today in a pretty convincing fashion. Next up: another chance to win the Kisei vs Yamashita with 6th game on 7th/8th.

Is this going to be Yamashita's losing move, so early at the beginning of day 1?
He played atari and lost 10% right there, and it seems to go downhill from there for him. Maybe too early to say?

Quick analyse by lizzie with leelazero 201, black(iyama) is leading, around 63%.

sorin, rather than taking the 40.4 winrate from only 1.8k playouts I'd suggest playing that atari and giving plenty of playouts for a more accurate assessment (particularly as a ko so LZ needs playouts to get through the ko and judge outcome of trades). Elf v2 thought it was a 30% mistake, and the bot they use on Yike 鹰眼 "Hawkeye" gave a similar 45 -> 18% white winrate judgement. Having said that Yamashita won the previous game from a position the bots judged as similarly bad, though there is the question whether this position has a similar scope for a human like Iyama to mess up.

Going back a few moves, all those bots slightly disliked Yamashita's attachment on the shimari, Elf suggesting this very nice nozomi (a shape I've noticed many bots like and I've been trying to promote up my policy network recently, see forum/viewtopic.php?p=227323#p227323), attach a crosscut combo. This is too good for white who bullies black into overconcentration whilst making good outside shape, so 4 should wedge but white is still good.

Click Here To Show Diagram Code

[go]$$W
$$ +---------------------------------------+
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , O . . . . , . . . . . X . . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . O . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . 1 . . . . |
$$ | . . . O . X . . . O . 3 4 . . 2 . . . |
$$ | . . . . O X . . . . . X 5 . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+[/go]

Gosh, I just checked the game and Yamashita pulled off another reversal from a dismal position (the Yike bot had Yamashita down at 5%). Iyama had 2 weak groups which he should have been able to settle by exploiting the bad aji in Yamashita's attacks and keeping miai for safety, but one little inaccuracy (move 199*) and the group died in ko. Usually it's Yamashita making mistakes vs Iyama in overtime. 3-3 now, decider is on 14-15 March. This would be quite the comeback for Yamashita if he can pull it off, not just from game scores but the losing positions he is winning from when facing kadoban.



* Yike bot said that's where the turning point was, I had a little explore with LZ on low playouts and can see some variations in which blocking from above instead is better because black can force with f11 and h10 causing some shortage of liberties aji in the bamboo joint giving connection aji through e11, though the a12 placement is sometimes a surprise so more thorough analysis needed to conclude black would indeed have survived blocking that way.

Here's my analysis with LZ 207 about why block other side for 199 was better. Conclusion, it is: having the stone at g13 comes in very useful 40 moves later when black wants to counterattack against white's middle group in a semeai he can then win by 1 liberty, or creating the shortage of liberties in the e10 bamboo combo-ing with e11. There are some lines where white does kill the left group but that involves ignoring a move in the middle so black captures a bunch of stones there in sente and wins anyway. The intuition of these bots to see block above better in just hundreds of playouts (<1 second) when it depends on ~40 move sequences of delicate use of aji is impressive. I expect Iyama could do it too if he had enough time and concentration though.