(;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Chinese]SZ[19]KM[6.50]
PW[Park Junghwan 9p]PB[Master]C[Meng Tailing 6p

"Crawling on the 2nd line in the opening; Master once again overturns human understanding"

Translated by Daniel Hu 4d

Today we don't have a guest, but I will be talking about a game with Park Junghwan.

This game is after we understood Master was unbeatable, so lets see how Park managed this.]
;B[pd]
;W[dd]
;B[qp]
;W[dq]
(;B[nq]C[It particularly loves the large knight's enclosure.

My two games with it, it played this

Its opening isn't fixed]
;W[qf]
;B[nc]
;W[rd]
(;B[co]C[It is particularly fond of the low approach in combination with the large knight's enclosure]
(;W[qc]C[park opted to resist, very normal]
;B[gq]C[and this is very normal]
;W[ep]
;B[go]C[In this situation B raises it's head to the centre and it works well with the large knight formation]
(;W[en]C[so Park changes his move.

If he didn't understand Master so well, he wouldn't have played here, but the knight press, as we all play there

It;s hard to say whether this is good or bad. But a computer like Master has no emotions, but humans do, and I think this shows how Park's feelings were a little unsettled

]
;B[cl]C[The purpose of the jump is to take sente]
;W[kc]TR[pd][rd]C[This seems like a comfortable pincer to make but notice that compared to a normal 3-3 shape, the triangle exchange is good for B. So W can't reach the same efficiency as two moves in the corner.

Then let's look at the bottom left, it is 4 vs 3, so compared to B first taking the corner, it is more efficient. 

I find these are very difficult questions to answer.

At the least, it seems normal.

Now the next area becomes abnormal]
;B[pf]
;W[pg]
(;B[qe]C[This move most people don't dare to play as W's next move is clearly to atari

This is a really shocking move]
(;W[of]C[of course Park ataried, demonstrating humans' judgement of value]
(;B[re]
(;W[rf]C[This move needs some consideration as B is clearly not still going to descend]
;B[pe]C[W has two cuts but you can't protect them, given that you clamped, you have to connect under.

]
;W[se]C[It still looks like this shape isn't good for B, but it's next choices I feel are even more surprising to humans than R15]
;B[qg]
;W[pc]
;B[ne]
;W[qh]
;B[rg]
;W[sf]C[but this means a shape that totally overturns our understanding of value appears]
;B[rh]
;W[qi]
;B[ri]C[If B wants to live, B has to crawl like this]
;W[qj]
;B[rj]
;W[qk]
;B[rk]
;W[ql]
;B[rl]
(;W[qm]C[B has 6 moves already on the 2nd line.

If this was a student, child, they would certainly we told off.

But master played like this

Crawling on the 2nd line is said to be "living is worse than dying"

In hindsight, perhaps W shouldn't have continued extending. This move is a little problematic.]
;B[oc]C[This is truly very painful]
;W[rb]C[To avoid hurting L17, directly making life]
;B[og]
;W[ph]
;B[ng]C[A simple extend]
;W[rm]C[Of course, W makes pain for B]
;B[sg]
;W[sd]
;B[sl]C[B has to submit.

B's efficiency is so low, and W has a huge wall outside it. And even S7 was sente. It seems like a very satisfactory result for W.

But firstly it was Master who played this, making it very difficult for you to say anything

Secondly lets compare this to if W just blocked S15. That diagram where W ends by protecting at Q12. There B's upper group is not yet settled, and W has potential on the right. Now W has also gained value on the right, but W isn't alive.

Perhaps this is the key point. That this W group isn't alive is why Master thinks this is ok to play.

That this W group that seems thick, is really a group that is very thin, and very easy to attack.

Then this is a totally different judgement.

If W is still to be attacked, we can't say W got something valuable here. When the top right corner became so small, and B got so solid at the top.

So the analysis depends on the strength of the right white group. Whether it is positive or negative in value, and if positive, by how much.

Truly it's hard to say B is bad. Only time will tell.]
;W[jp]C[Truly though B is so small on the right, B has a solid 4 points and the top group has a solid 8 points.

And another corner enclosure, while W's corner is 5 points, so B's territory on the right side is far ahead of W's. And L17 is a little soft, so it is still a matter of the value of the right white group

Clearly Park says he can now use the group as thickness, invading here]
;B[no]C[This move is so simple]
;W[hp]
;B[gp]
;W[jm]C[B's 3 stone do also look a bit heavy]
;B[ni]C[And this move!

so simple

When you talk of attacking, you normally talk about how to seal the group in. This is a job that takes a lot of effort.

But this move is so slow, leaving so much room for W to run]
(;W[mm]
(;B[hn]C[But perhaps this is the true goal of O11, to probe W's response before choosing where to run with this B group.

O11 wasn't really to directly attack, but to get out more easily with this group

This is a very high level of play, what we normally call a xingqi diaozi (melody of stone flow)

This is a connected line of thought, to jump O11 and then to kosumi

This move aims at M6.

What to do? Should W protect? Or how? No pro would want to protect as it looks very inefficient]
;W[el]C[Just simply jumping. This doesn't look like a great move, but what else can you play?]
;B[ln]C[But it is easy to choose where to play for B]
(;W[in]
;B[kl]
;W[hm]C[so W plays to deal with this side first, saying that W is already alive on the right, and truly W does have a lot of eyespace,

but as W is so heavy, it will inevitably make a big impact on this fight.]
;B[gm]
;W[hl]
;B[gl]C[this is why W didn't dare hane]
(;W[hk]
;B[gk]
;W[hj]
;B[ek]C[but now, not only has B pushed through between the two white groups, but this attach drastically reduces the value of E8.

Somehow with the O11 H6 combination, B very simply changed direction to seal in the bottom left, taking the initiative this game.

Master in hard to notice spots can bring our top players to collapse. Its understanding of go is so much stronger than ours that we won't have any chance when playing against it.

Starting from K4, a move with plenty of fighting spirit, now we find W with several groups all not alive.]
(;W[cm]C[W can't accept being simply sealed in.]
;B[bm]
;W[dl]
;B[ck]
;W[dk]
;B[dj]
(;W[cn]C[W is worried B will atari D7 to make a good exchange, but this is too crude]
;B[bn]
;W[ej]
;B[ci]C[so simple]
(;W[fi]
;B[er]C[B's play is very simple saying W's position is already split into pieces, E9 originally seemed to be meaning to seal in for a wall, now B settles the left and B just wants to live and the bottom to be happy.

I think this shows how B thinks the position is good for it.]
(;W[do]
(;B[ei]C[a very nice combination]
(;W[fj]C[It is imperative to connect here, and B's good move makes a good exchange]
;B[jr]C[B is in no rush to exchange C5 with F9, seeing that it is its right anyway.]
(;W[eh]
;B[cp]
;W[mf]
;B[nf]
;W[fc]C[This shows that Park thinks the game is not going well.

W isn't even protecting the F9 cut which concerns the G10 cut too

So how does Master attack W?

There are many options]
(;B[gc]C[But Master plays here. Just after W protects, B attaches, a very flexible way to play.

This is a move humans wouldn't even think about especially  when there are so many ways to attack. Attaching immediately after a protection.

But it is very hard to respond. B is probing to see W's response.]
;W[gd]C[Park chooses above]
;B[hc]
(;W[fd]C[but W thinks this is his area and doesn't want B to live. But I think this is W's losing move. Afterwards, the strength and weakness totally reverses]
;B[mk]C[Master plays so simply]
;W[om]C[W says there are 3 places to make eyes - above, on the right, and B's weaknesses

But it seems like Master didn't think about killing W.]
;B[jd]C[now W is in trouble. W can't just die as all the previous efforts (to attack) will be wasted.]
;W[kd]
;B[jf]C[But as B is too solid, there isn't much  W can do]
;W[kf]
;B[jg]
;W[id]
;B[jc]
;W[je]
;B[ie]
;W[ke]
;B[hd]
;W[if]
;B[he]
;W[jb]C[it looks like W's shinogi was very pretty but B isn't bothered]
;B[pn]C[W has to make life as the group is too big to sacrifice in an exchange, but it is difficult to live by your own strength so what to do?]
(;W[lm]
;B[km]
;W[mn]
;B[jn]C[but this is a very clear method]
;W[jo]
;B[kn]
;W[mo]
;B[io]
;W[im]
;B[kp]C[Master may give up a lot in the endgame, but it isn't going to let W connect]
(;W[np]
;B[mp]
;W[op]
;B[oo]
(;W[oq]
(;B[pp]C[if B still goes here, ]
;W[or]C[this is sente]
;B[mq]
;W[qr]C[there might be a ko, though it doesn't look like it works for W either, but there is still the possibility])
(;B[mq]
;W[qq]
;B[pp]
;W[pq]
;B[rq]
;W[rr]
;B[rp]C[there is still the 1st line connection. This game has gone far away. When B captures at J17, the W top group is still in trouble as well. The G10 cut is still severe. And W doesn't seem to have much territory anyway]
;W[lp]
;B[lo]
;W[nr]C[This sente means W doesn't immediately die.]
;B[mr]
;W[po]
;B[qo]
;W[qn]
;B[sr]
;W[qs]
;B[or]
;W[pr]
;B[ns]C[W gets to make a multi-step ko. Clearly W needs to spend more moves to live, but Park tenukies anyway]
;W[ic]
;B[kh]C[But B answers!

W already profited a lot, but there isn't much significance to affect the result]
;W[ho]
;B[gn]
;W[di]
;B[cj]
;W[pm]
;B[fk]
;W[fr]C[only move]
;B[fq]
;W[eq]
;B[dr]
;W[cr]
;B[gr]
;W[cq]
(;B[hg]
;W[ih]
;B[ig]
;W[gf]
;B[hf]
;W[on]
;B[po]C[But playing solidly does have some logic which is that it is very difficult for W to play the ko on the bottom right]
;W[ps]
;B[rs]
;W[ss]C[a ko in B's favour]
;B[sq]
;W[nn]
;B[rs]C[Though this ko takes two moves to finish, so some chances appear, not only is the ko in B's favour, but when B wins it, W has to make life.

And in terms of territory if B wins the ko, B has 30 points more, W is losing by a lot.]
;W[nj]
;B[mj]
;W[nl]
;B[cg]C[B doesn't even bother with the ko]
;W[ce]
;B[os]C[B doesn't capture but atari here, like B already doesn't know how to play anymore

(a phrase Meng often used to say someone is too happy with their position they can play anywhere)]
;W[ss]
(;B[so]C[Maybe in the computer's mind, this move has less variations, and so is closer to victory. We don't know exactly how, but the bot has a habit of leaking water in the endgame, winning by less and less. 

Basically the endgame is insignificant, the gap is too large.

this ko is still a one-step ko and B has so many more threats.]
;W[rn]
;B[rs]
;W[jq]
;B[kr]
;W[ss]
;B[bq]
;W[bp]
;B[bo]
;W[br]
;B[ap]C[wasting a threat]
;W[ar]
;B[rs]
;W[hq]
;B[hr]C[B could have gotten this in sente earlier]
;W[ss]
;B[fo]
;W[eo]
;B[rs]
;W[lh]
;B[ki]
;W[ss]
;B[em]
;W[fm]C[we can see B's threats are too many]
;B[rs]
;W[li]
;B[lj]
;W[ss]
;B[gj]
;W[hh]
;B[rs]
;W[jj]
;B[kj]
;W[ss]
;B[fg]
;W[gi]
;B[rs]
;W[bf]
;B[qr]C[B+R


This is a standard gap between Master and humans.

Its just how it is.

Despite being Korean number 1, Park didn't show any chances against Master. We go up to fight and soon we go down.

See you next time.])
(;B[rn]C[This is actually much better for points]
;W[so]C[the difference is much bigger than it looks as B can end the ko at T7]))
(;B[hh]C[not even protecting in sente, just asking to win by a little bit])))
(;W[mq]
;B[lp]
;W[oq]
;B[pp]
;W[nr]
;B[pr]C[W living here but not with the big group is not ok]))
(;W[mp]
;B[np]C[like this it feels like W is already dead, so W needs to find B's weaknesses]))
(;W[mn]C[W should have pushed first]
;B[lo]))
(;W[hd]C[another game]
;B[fb]
;W[ec]
;B[ic]
;W[je]
;B[kb]C[just simply making life and taking points. If like this, maybe the meaning is that W's point snatching becomes pointless]
;W[cg]C[Though I think the difference between the two sides wouldn't be much here, W made a lot of points in the centre, perhaps this is how W should play.

This seems like W can continue fighting.]))
(;B[nl]C[B could try to kill W, it looks hard to live, especially if W doesn't want to lose too much on the outside.]))
(;W[nj]C[W didn't come back to protect this group. This will certainly live, but then it will be a matter of counting territory, and B clearly has more, so W didn't choose this.]))
(;W[fk]
;B[gj]
;W[gi]
;B[fj]
;W[ek]C[W would be cut of in sente]))
(;B[cp]C[If B directly extends, B is scared W will capture]
;W[fk]
;B[ei]
(;W[fj]C[This exchange is no longer possible, but would be very good for B])
(;W[eh])))
(;W[dr]C[you can't do the normal block]
;B[do]
;W[eo]
;B[eq]
;W[dp]
;B[jr]C[Though it looks like W split B's position, B is no issues living, but W needs to, and it seems that it is W that hasn't got anything of value.]))
(;W[ei]C[W would really like to play here but]
;B[fk]C[this is sente]
;W[]
;B[fi]
;W[fj]
;B[gj]
;W[gi]
;B[fh]C[It is very hard to escape with the 3 stones now as well as the G11 stone.]))
(;W[ej]
;B[dm]
;W[em]C[Park feared having to play here]
;B[ci]))
(;W[dl]C[it seems like W should have played here directly as C7 becomes a bad exchange, making a big difference to the fight]))
(;W[gk]
;B[hk]
;W[fk]
;B[jl]
;W[hq]C[even if B dies]
;B[km]
;W[io]
;B[ik]C[if these are sente]
;W[im]
;B[nl]C[The right side is a lot bigger]))
(;W[lm]C[You can't block]
;B[jn]C[W can't let B kill the two stones]
;W[mn]
;B[lo]
;W[mo]
;B[mp]
;W[in]C[I'm guessing B would just make the exchange]
;B[io]
;W[kn]
;B[im]
;W[ko]
;B[lp]C[actually this is still W collapse as though W gets to connect up, W still doesn't have eyespace and the shape is still thin, 

but B already got a ponnuki and the option to take the territory and connect at K3]
;W[jl]
;B[jq]
;W[ip]
;B[ir]))
(;B[gm]C[normally we would talk about jumping out, but ]
;W[hn]C[this is comfortable]
;B[gn]))
(;W[ml]C[in hindsight perhaps this would be slightly better, and B wouldn't so easily find W's weaknesses]))
(;W[og]C[since B has crawled so much on the 2nd line, R7 is no longer as big

so why did Master play like this?

viewers, do you think this is a valid way to play?

Personally I express my doubts. But its go you have no choice but to respect.]
;B[qm]C[I doubt B would play here, as the top is urgent.]))
(;W[ob]C[I think maybe it should be here

I don't where B will play]
;B[pe]C[since if you still connect, ]
(;W[ng]C[B's shape is not as good]
;B[oc]
;W[nb]
;B[mb]
;W[pc]C[2 cutting points])
(;W[oc]
;B[nd]
;W[ng]C[or this, P18 is much better placed than Q17]))
(;W[pc]C[This is a natural shape too]
;B[pe]C[B's shape looks a bit strange, but where next?]
(;W[oc]
;B[nd]
;W[od]C[If you cut]
;B[oe]
;W[ne]
;B[nf]
;W[me]
;B[og]C[though it looks like the two black stones above are dead, you have to remember that this position was originally W's corner with a pincer, W spent a lot of moves here, and it should get more than B, but B's ponnuki is very valuable and works very well with the bottom right corner])
(;W[nd]C[or here?]
;B[og]
;W[md]
;B[ph]C[to be honest, it seems similar, as long as B gets two moves in a row on the right, at the least you can't say you have punished R15, you can't say B is bad.])
(;W[og]C[there's nothing else to play as connecting lets black get good shape]
;B[ne]C[The corner still has issues])))
(;B[pe]C[And if you connect you are in "Russian square" shape, certainly to be told off]))
(;W[re]C[what's the problem with blocking?]
;B[qg]
;W[rf]
;B[of]C[The direct extension]
;W[ph]C[why did master play R15? since if directly P14, W might just Q12, did it fear R12?

I don't know why it risked the P14 atari to aim for this result. Clearly its judgement and that of humans is different.]))
(;B[of]
(;W[ph])
(;W[qh]C[because this is more solid?])))
(;W[dn]C[If here]
;B[cn]
;W[dm]
(;B[cl]C[It only crawls once before jumping, very interesting.

We are used to crawling an extra time]
;W[cp]
;B[dl]
;W[fn]
(;B[jp])
(;B[cf]
;W[fc]))
(;B[cm]
;W[dl]
;B[ck]C[like this, W is more solid, and so can tenuki]
;W[lc])))
(;W[ck]
;B[fp]C[my game went here]))
(;B[fc]C[In a human tournament you would expect this]
;W[cf]
;B[qh]
;W[qc]
;B[qe]
;W[re]
;B[pf]
;W[pg]
;B[qg]
;W[rf]
;B[og]C[This has become a classic opening and it has been played for a long time]
;W[hc]C[Master has played two game like this, as black and instead of the normal human move]
(;B[pj])
(;B[nj]C[It protected here])
(;B[oj]C[or here])))
(;B[np]C[sometimes here or elsewhere]))