Vulnerable groups : memorised or re-analysed ?

judicata · Post by **judicata** » Tue Jan 11, 2011 12:57 pm

First, to end speculation on the game Dusk Eagle posted, here are the first several moves of that game (I've removed the opponents name, although the entire game is publicly available):

It appears the sequence was played pretty early, and looks very good for white.

Don't be embarassed--it will prevent improvement. As others have said, don't place too much emphasis on the corners. Although the corners are the easiest place to create territory, "capturing" some points in the corner does not come without a cost. Besides, if you really want to make territory in the corners you open in (i.e., the corners you play in first) open on the 3-3 point.

I would also encourage you to play some real people--don't let embarassment hold you back!

Finally, as far as a "sandbox" you can experiment posting a game here in this thread--you won't break anything.

hyperpape · Post by **hyperpape** » Tue Jan 11, 2011 1:14 pm

Chaa006 wrote:I'll try : I just don't like feeling embarrassed But is there a sandbox on the forum in which one can practice uploading games, as this would be my first attempt at such a thing ?

Just in case you hadn't seen it, here are instructions: viewtopic.php?f=14&t=833.

Chaa006 · Post by **Chaa006** » Tue Jan 11, 2011 1:15 pm

judicata wrote:First, to end speculation on the game Dusk Eagle posted, here are the first several moves of that game (I've removed the opponents name, although the entire game is publicly available):

It appears the sequence was played pretty early, and looks very good for white.

Don't be embarassed--it will prevent improvement. As others have said, don't place too much emphasis on the corners. Although the corners are the easiest place to create territory, "capturing" some points in the corner does not come without a cost. Besides, if you really want to make territory in the corners you open in (i.e., the corners you play in first) open on the 3-3 point.

I would also encourage you to play some real people--don't let embarassment hold you back!

Finally, as far as a "sandbox" you can experiment posting a game here in this thread--you won't break anything.

OK, many thanks. All advice heeded and taken on board. I have played real people in the past (many of them : Brian Chandler, Stuart Dowsey and Alan Fairbourn were probably the highest graded players I ever had the privilege of playing against), but at the moment I am playing table tennis against real people and Go against the computer when I am stuck indoors !

I have attached my most recent game against MFoG, basically because I can conceal my embarrassment at my awful moves behind the satisfaction of having won

Hyperpape : many thanks for the link; it arrived just after I'd pressed "Submit" but I think I have done it correctly ...

daniel_the_smith · Post by **daniel_the_smith** » Tue Jan 11, 2011 2:17 pm

Chaa006 wrote:P.S. Daniel-the-Smith : I have just looked for "Five might live but the sixth is dead." in Go Proverbs Illustrated, but can't find it; obviously there must be a very large number of Go proverbs, so not all could possibly be included, but I wonder if you remember where you came across that one ?

Here's one place: http://senseis.xmp.net/?path=GoProverbs ... xthWillDie

Proverbs aren't meant to be too deep, IMO; the main point is that you don't have time to make a bunch of separate groups live individually.

---

A situation where you have to keep track of all your opponent's sente moves against your groups is inferior to one where you don't, all other things being equal. The solution isn't to start keeping better track of these moves; the solution is to stay out of positions where your opponent has them at all. (A few are unavoidable, probably, but you should take great pains to minimize them...)

Another thing to think about: sente moves like that are very likely to be unignorable ko threats. It doesn't take too many of them to get into a position where you both know your opponent will win any ko. At that point, any sticky situation becomes favorable to your opponent, since they just have to make ko.

Chaa006 · Post by **Chaa006** » Tue Jan 11, 2011 2:32 pm

daniel_the_smith wrote:Here's one place: http://senseis.xmp.net/?path=GoProverbs ... xthWillDie
Proverbs aren't meant to be too deep, IMO; the main point is that you don't have time to make a bunch of separate groups live individually.(snip)

Thank you. I was thinking of treating myself to a copy of Milton Bradley's New Go Proverbs Illustrated as a late Christmas present, but maybe I should save my money and just read the Sensei's proverbs online

(Not so easy to do in bed, 'though).

Dusk Eagle · Post by **Dusk Eagle** » Wed Jan 12, 2011 1:20 pm

From what I can tell, one of your problems (and it's a common one among weaker players, including myself when I started) is that you struggle to recognize where the big points are on the board. That's what I've focused on most in this review. None of your groups died or got too weak in this game, so I couldn't focus on vulnerable groups all that much, but hopefully this will be of use to you anyway.

[sgf-full](;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Japanese]SZ[19]KM[0.50]
PW[The Many Faces of Go]PB[Philip Taylor]WR[18k]BR[19k]DT[2011-01-11]RE[B+Resign]VW[]
;B[cp]
;W[pq]
;B[qd]
;W[ec]
;B[cc]
;W[qo]
(;B[rq]C[This is a clear case of putting too much value on the corner.For one, you can't live in here: you don't have enough room (see variations for white's next move). However, even if you could live in here, you would have only 2-5 points total while white gets overwhelming influence.

To put it another way, there is a proverb that says "The second line is the route to defeat." Playing on the second line before the endgame will often give you a very low position with very few points while your opponent gets overwhelming influence. See variation 3 for what I mean.

See: http://senseis.xmp.net/?Crawling]
(;W[mq]LB[jp:B][jq:A]C[Nothing move by MFoG, if it doesn't kill the corner, it should at least extend to A or B, or play elsewhere.]
;B[eq]C[Good.]
;W[qj]
(;B[oc]LB[de:A]C[I think 'A' would be bigger, but this isn't terrible.]
;W[qg]
;B[qm]TR[qd][qm]SQ[qg][qj]C[This is a stone that is bound to become weak, while you provoke white to kill your corner stone at the same time. Both of your triangled stones are too far away from the white stones to put them under a severe attack, and your latest move will have no base once white blocks off the corner. It will be a very weak, useless stone.]
(;W[hc]LB[de:A]C[:O! Locally, white should have played 'A', but I think the variation is bigger.]
;B[lc]
;W[iq]
;B[gq]LB[de:A][dj:B][ro:C]C['A', 'B', 'C' are much better points. This is too small.]
;W[kq]LB[ce:C][de:A][cf:B][ci:H][cj:G][ck:F][io:D][rp:E]C[Soooooooooo passive and small. Don't emulate play like this. All of the marked points are 10 times better than MFoG's play here.]
;B[ro]C[Good move: connect your weak stones to make them strong. This is why white should have prevented that itself.]
;W[qp]
;B[rp]
;W[jc]
;B[cj]
;W[dm]
;B[cm]
;W[cl]
;B[bm]
;W[el]
;B[dk]
;W[ek]
;B[dl]
;W[ej]
;B[bk]C[What are you afraid of? Check if white is actually threatening anything.]
;W[ci]
;B[em]TR[dm]C[Since this stone is not a cutting stone, it is not all that valuable to capture or for white to save. That top-left corner is still bigger.]
;W[cf]
;B[ce]
;W[de]
;B[dc]
;W[df]
(;B[dd]C[What does this do? It only strengthens white (aji keshi) and makes bad shape for you (empty triangle). ]
;W[ed]
;B[pp]C[(This isn't a ladder-breaker)]
;W[qq]
;B[ee]
(;W[ef]
;B[fe]
(;W[ff]
;B[ge]
;W[gf]
;B[he]
;W[hf]
;B[ie]
;W[if]
;B[je]
;W[kd]
;B[ke]
;W[ld]
;B[le]
;W[md]
;B[mc]
;W[me]
;B[mf]
;W[nf]
;B[lf]
;W[ne]
;B[ng]
;W[og]
;B[of]
;W[oe]
;B[pf]
;W[pg]
;B[nh]
;W[qn]
;B[rn]
;W[rm]
;B[rl]
;W[di]
;B[bi]
;W[bh]
;B[ah]
;W[ag]
;B[ai]
;W[bg]
;B[ae]
;W[be]
;B[bd]
;W[cd]
;B[ac]
;W[ce]
;B[eb]
;W[db]
;B[cb]
;W[da]
;B[ea]
;W[po]
;B[op]
;W[oq]
;B[oo]
;W[nq]
;B[om]
;W[mg]
;B[lh]
;W[mh]
;B[mi]
;W[li]
;B[lg]
;W[mj]
;B[nj]
;W[lj]
;B[nk]
;W[pk]
;B[ol]
;W[pl]
;B[pm]
;W[kh]
;B[ki]
;W[kj]
;B[ji]
;W[jh]
;B[ih]
;W[ii]
;B[jj]
;W[ij]
;B[jk]
;W[ik]
;B[jl]
;W[il]
;B[lk]
;W[jm]
;B[kk]
;W[im]
;B[jg]
;W[dn]
;B[cn]
;W[en]
;B[fm]
;W[fn]
;B[gl]
;W[gn]
;B[hh]
;W[hm]
;B[gk]
;W[gi]
;B[hn]
;W[gj]
;B[ho]
;W[gm]
;B[ei]
;W[dj]
;B[eh]
;W[fg]
;B[ch]
;W[dh]
;B[lo]
;W[mo]
;B[pn]
;W[lp]
;B[lr]
;W[ko]
;B[ln]
;W[kn]
;B[lm]
;W[ck]
;B[bl]
;W[bj]
;B[aj]
;W[ak]
;B[al]
;W[aj]
;B[do]
;W[gh]
;B[lq]
;W[jp]
;B[nn]
;W[hp]
;B[gp]
;W[go]
;B[ep]
;W[gd]
;B[kc]
;W[jd]
;B[nd]
;W[od]
;B[nc]
;W[id]
;B[hd]
;W[gc]
;B[ib]
;W[ic]
;B[jb]
;W[kb]
;B[lb]
;W[pc]
;B[pd]
;W[qc]
;B[qf]
;W[rd]
;B[rf]
;W[rc]
;B[pi]
;W[qi]
;B[qh]
;W[rh]
;B[ph]
;W[rg]
;B[sf]
;W[si]
;B[pb]
;W[ob]
;B[nb]
;W[ra]
;B[qb]
;W[oa]
;B[na]
;W[km]
;B[pa]
;W[qr]
;B[np]
;W[mp]
;B[mn]
;W[hg]
;B[jf]
;W[fl]
;B[hr]
;W[ir]
;B[hs]
;W[is]
;B[fr]
;W[hi]
;B[kg]
;W[kl]
;B[ll]
;W[rr]
;B[sr]
;W[hb]
;B[ka]
;W[re]
;B[rb]
;W[sa]
;B[se]
;W[sb]
;B[sd]
;W[sc]
;B[qa])
(;W[eb]C[Can your corner live? If you're not so sure, are you really prepared to play like this? End of comments.]))
(;W[fe]C[Try reading out the ladder again. Read it all the way through. It's tricky at the end, but the result will surprise you. Perhaps you'll even catch other players making this same mistake, and then you can slaughter them

.])
(;W[be]C[If you can live (which is a bit of an 'if') it will be terribly small. If you die, white doesn't need to worry about your cut just now.]))
(;B[eb]
;W[fc]
;B[be]C[A better way to play.]))
(;W[rp]C[Now what?]))
(;B[de]
;W[hc]
;B[ci]TR[de]C[This is a big point for both your top stones and your bottom stones. Note that because your triangled stone is high, it wants another move along that side in order to feel "happy." If your triangled stone was on the third line, it wouldn't be as good. Meanwhile, along the bottom-left, you're starting to form a large area. White can still invade, but then you can attack white and profit from chasing MFoG around the board.]))
(;W[rp]C[This should kill.]
(;B[qq]
;W[pp]
;B[pr]
;W[or]
;B[qs]
;W[rr])
(;B[pr]
;W[or]
;B[qr]
;W[oq]
;B[rs]
;W[sr]))
(;W[qr]
;B[qp]
;W[pp]
;B[ro]C[Black will start pushing on the second line, which is a horrible, terrible result.]
;W[qn]
;B[rn]
;W[qm]
;B[rl]
;W[ql]
;B[rk]
;W[rm]
;B[sm]C[If you can't feel how bad this result is for black, look at it every day until you do.]))
(;B[eq]LB[dj:A]C[A good move: encloses the corner. In the future, you can follow up with a move at 'A'.])
(;B[de]C[Another good move. Picture the difference between you playing here and white playing here.]))[/sgf-full]

Chaa006 · Post by **Chaa006** » Wed Jan 12, 2011 1:52 pm

Thank you, Dusk Eagle : I very much appreciate your review. I am aware of (some of) my weaknesses, and when I was playing humans regularly I learned to avoid a few of them, but these days I am very much out of practice and am having trouble even getting back to 17 kyu against MFoG (I'm winning more than I'm losing now, but until I start seeing the occasional 16 kyu rating by the program I will know that I am still very very rusty !). I've just finished another game against MFoG, and again it resigned, but I still lost a couple of small groups through complete stupidity on my part. I will now stop playing and work through your review. Thanks once again.

Later : OK, worked through the review, which was very helpful. I will work through it a few times more to make sure I haven't failed to absorb something. But a comment, after some introspection : it is not that I don't recognise that the moves I make are frequently small; it is more that I lack the confidence to play the big move, because experience has taught me that I lack the skill to make proper use of such a move subsequently. I think that this is one of the down-sides of playing against a computer with its skill level set to match one's own : it is very easy for an 18 kyu human to play small moves against an 18 kyu computer yet still come out on top; whereas if I were to set the computer to play at shodan level, I would have to play a completely different class of game. Little moves would be punished mercilessly, and only well thought out big moves would have any chance of succeeding. I would still lose, of course -- that goes without saying -- but it might help me to up my game. What do you think ?

Much later : well, I decided to investigate your variant in which you play W14 at S4; I set MFoG to play at shodan level, then tried various approaches in the T1 corner until I found one that allowed me to defend the corner against MFoG, and then used that as a base from which to develop. I had to considerably increase the time allowed to prevent White from losing on time, and finally set it to 3 hours/player, of which White used about 2 (I used about 1). The game ended with a win by White of course, but the margin was nowhere near as great as I anticipated: I lost by 85,5 points. But most interesting to me was the fact that I learned so much more by playing against MFoG at shodan level; when it plays at 18 kyu, which is where I estimate myself to be, it makes so many apparent mistakes that no "big moves" are needed to defeat it : it can be defeated simply by small, tactical, moves. But at shodan level, these little moves just don't work. I also began to appreciate the real benefit of playing sacrifice stones at cutting points.

So, thank you once again for your analysis, and also my thanks to Magicwand & Violence for their comments, which followed your message. I now intend to play a number of unranked games against MFoG set to play at shodan level, to see just how much I can learn by playing against a much stronger opponent.

Magicwand · Post by **Magicwand** » Wed Jan 12, 2011 1:53 pm

strong dan level knows exactly how safe your group is.
we do not need to re-analize (<--that is funny) on every move because we already expect opponent's attack before it happens and finished reading prior.
we do not memorize all possible shape of go. we read.

Violence · Post by **Violence** » Wed Jan 12, 2011 3:22 pm

Magicwand wrote:strong dan level knows exactly how safe your group is.
we do not need to re-analize (<--that is funny) on every move because we already expect opponent's attack before it happens and finished reading prior.
we do not memorize all possible shape of go. we read.

^What he said.

When we have a group, even if it looks solid, a lot of times, I think of the potential attacks that can happen. "If only I had this stone... if only this was here..." etc. Then as you play with that in mind, you can come up with a way to induce the conditions for a strong attack. It's a lot of reading, and some creativity.

Likewise, with my own groups, I try to think of the ways my opponent can attack them, and come up with a response in my head. Effectively, it's like bracing for a hit before it comes. You want to have the plan before the move comes. I'm sure we've all been surprised by moves before, and one of the things that hurts the most is a punch that comes from an unexpected direction.

This all reminds me of a quote Mr. Gus Price(Tabemasu) said to me as he watched me playing online.

Me: This group looks so weak... but I've never seen it before, I dunno how to attack it...

Gus: How horrible, you're going to have be creative, why would you ever want to do that at all???

I ended up losing the game.

Life In 19x19

Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?

Re: Vulnerable groups : memorised or re-analysed ?