Page 1 of 1
A teaching game 13x13
Posted: Wed Jan 16, 2019 12:00 pm
by Knotwilg
A few months ago I gave a lecture on Alpha Zero to introduce Monte Carlo and other machine learning techniques in our learning society at work (the software department of digital mapping company and navigation software provider TomTom).
One of the attendants knew the game but had only played a few times, enough though to buy himself a board and stones, which he today brought to the office. We played 13x13, which was quite right for his level and the time we had. We played a no komi (teaching) game. In retrospect I missed a few opportunities to set up a problem at his level. The aji I left was too hard for him to exploit.
Anyway, the game led to three lessons:
1) when attacking, avoid contact plays but pressure from a small distance
2) make your stones work efficiently, maximizing liberties, influence, pressure ...
3) exploit/don't leave an excess of cutting points
(;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Japanese]SZ[13]KM[0.00]
PW[Dieter]PB[Daniel]RE[W+26.00]
;B[dd]
;W[kj]
;B[ej]
;W[jc]
(;B[jj]C[Lesson 1: when you want to put pressure on a stone, don't play a direct contact move, because both stones have 3 liberties and the opponent can move next.
See variation]
;W[jk]C[Black's liberties are down to 2]
;B[ij]
;W[ik]C[While Black is already gasping for air, White is building secure territory in the corner.]
;B[hk]
;W[hl]
;B[gk]C[Properly connecting your stones and making them stronger by joining their forces.]
;W[ki]LB[kk:A]C[White operates more freely, backed up by the space in the corner.
While doing so, White has checked that the cut at A represents no danger
See variation 1]
(;B[ji]LB[jh:3][ii:2][hj:1]C[Lesson 2: your stones should indeed aim for connection, but also for efficiency.
Before, this chain had 3 liberties. After, it still has liberties. There's hardly any gain in influence towards the centre, nor in strength. There is a mild increase of pressure on the White stones, but as White can move next, he can increase his advantage comparatively more.
See variation 2]
;W[jh]C[Black's liberties are again down to 2, while it is White increasing his influence towards side and centre.]
;B[hi]C[A very good tactical solution, connecting all the stones in a way that gives "eyes" as well (a higher level concept in the context of eventual life for your stones)]
;W[cf]C[White takes the initiative to play elsewhere on the board, leaving the corner unattended.
This is a risk, because White has 2 cutting points. Let's see if Black can exploit them in the variation
(Lesson 3: exploit an excess of cutting points)]
(;B[ic]C[(lesson 1)]
;W[jd]C[A less aggressive response, still White has the better of this exchange]
;B[id]
;W[je]
;B[ie]
;W[bd]C[Being connected on a large scale, White goes for action elsewhere
(still ignoring the excess cutting points)]
;B[cg]C[Lesson 1 but ...
This is a good opportunity to say the contact play is not always bad per se and depending on tactical arguments it can sometimes be a surprisingly good move.
There are no clear cut rules for playing well in Go!]
;W[bg]
;B[ch]
;W[bh]C[White is low, but since Black's forces are somewhat overlapping in the centre, White is ok with that]
;B[bi]C[Tactically Black is doing well: he knows when to block and when to give way]
;W[cd]
;B[df]C[very good: Black closes off the centre from this side, while tucking white into the corner
Also ...]
(;W[bf]C[White fixes his weakness but feels submitted to it]
;B[ce]
;W[be]
;B[de]
;W[db]C[find breathing space]
;B[eb]
;W[cb]C[a sneaky move]
;B[jf]C[Black takes the initiative to close off the centre. This shows gamesmanship]
;W[kf]
;B[if]
;W[kh]C[White takes time to fix his weakness]
;B[ih]
;W[gl]C[We're at the so called "end game" now, where all groups are stable and it's all about increasing/decreasing your/the opponent's territories]
;B[fl]
;W[ec]C[White returns here to intrude Black's territory]
;B[fb]
;W[fc]
;B[gc]
;W[dc]
(;B[fd]LB[gb:5][ed:3][gd:4][dg:2][ci:1]C[Lesson 3: beware of too many cutting points]
;W[jb]
;B[ib]
;W[bj]
(;B[cj]C[A good sense for danger. We discussed the variation, seeing that Black cannot cut]
;W[ai]
;B[ci]
;W[bk]
;B[bl]C[The first real tactical error. This simply does not work]
;W[ck]
;B[dk]
;W[cl]
;B[dl]
;W[cm]
;B[dm]
;W[ed]
;B[ee]C[Lesson 3]
;W[gd]C[Exploiting cutting points]
;B[fe]
;W[gb]
;B[hc]
;W[fa]
;B[gm]C[Black now executes perfect endgame]
;W[il]C[forced]
;B[jg]
;W[kg]C[forced]
;B[ja]
;W[ka]
;B[ia]
;W[kb]C[forced]
;B[hm]C[last point]
;W[im]
;B[fm]C[giving back the initiative when there's nothing more to do]
;W[ea]
;B[]
;W[]
;B[]C[White wins by 26 points thanks to
- taking advantage of premature contact plays
- a more efficient usage of his stones
- exploiting cutting points])
(;B[ai]
;W[ci]
;B[cj]
;W[di]
;B[bk]
;W[dg]))
(;B[gd]C[Sometimes, taking one point less, is better for overall strength]))
(;W[jg]C[If White plays elsewhere]
;B[bf]
;W[ce]
;B[ag]C[Black can capture 2 stones]))
(;B[kh]
(;W[lh]
;B[kg]C[The centre stone is cut off, and White loses his efficiency])
(;W[kg]
;B[lh]
;W[lg]C[White can capture these stones but ...]
;B[kk]C[When he cuts here too, the position crumbles]
;W[lk]
;B[kl]LB[lj:B][il:A]C[White can't prevent both captures at once])))
(;B[kk]
;W[kl]
;B[lk]
;W[ll]C[This leads nowhere for Black, who's already down to 2 liberties and no breathing room is to be found at the edge.
This is a mental exercise White performed before playing move 12])
(;B[jh]LB[kh:A][ji:B]C[If Black jumps, his influence towards the centre is much bigger. Also the threat of tucking White into the corner, by playing A, represents more pressure on the white stones.
Greater efficiency induces more risk: White could cut at B. But this can be checked to be manageable. It should, given the amount of Black stones present in that area.]))
(;B[ij]C[Keep distance. Now the opponent cannot take advantage of the contact play to weaken your attacking stone by removing its liberties. Still the pressure is on because the space in which that stone can operate is reduced. The same is true for the attacking stone, but it is backed up by the supporting stone on that side.]))
Re: A teaching game 13x13
Posted: Wed Jan 16, 2019 12:07 pm
by Bill Spight

How about the counter hane at L-05?
Re: A teaching game 13x13
Posted: Wed Jan 16, 2019 1:35 pm
by Bill Spight
Small improvement in the endgame.
(;SZ[13]FF[4]AP[GOWrite:3.0.15]CA[UTF-8]GM[1]ST[2]GN[ ]RU[Japanese]PB[Daniel]FG[259:]KM[0.00]PW[Dieter]RE[W+26.00]PM[2]
;B[dd];W[kj];B[ej];W[jc]
(
;C[Lesson 1\: when you want to put pressure on a stone, don't play a direct contact move, because both stones have 3 liberties and the opponent can move next.See variation]B[jj]
;C[Black's liberties are down to 2]W[jk]
;B[ij]
;C[While Black is already gasping for air, White is building secure territory in the corner.]W[ik]
;B[hk];W[hl]
;C[Properly connecting your stones and making them stronger by joining their forces.]B[gk]
;C[White operates more freely, backed up by the space in the corner.While doing so, White has checked that the cut at A represents no dangerSee variation 1]LB[kk:A]W[ki]
(
;C[Lesson 2\: your stones should indeed aim for connection, but also for efficiency.Before, this chain had 3 liberties. After, it still has liberties. There's hardly any gain in influence towards the centre, nor in strength. There is a mild increase of pressure on the White stones, but as White can move next, he can increase his advantage comparatively more.See variation 2]LB[jh:3][ii:2][hj:1]B[ji]
;C[Black's liberties are again down to 2, while it is White increasing his influence towards side and centre.]W[jh]
;C[A very good tactical solution, connecting all the stones in a way that gives "eyes" as well (a higher level concept in the context of eventual life for your stones)]B[hi]
;C[White takes the initiative to play elsewhere on the board, leaving the corner unattended.This is a risk, because White has 2 cutting points. Let's see if Black can exploit them in the variation(Lesson 3\: exploit an excess of cutting points)]W[cf]
(
;C[(lesson 1)]B[ic]
;C[A less aggressive response, still White has the better of this exchange]W[jd]
;B[id];W[je];B[ie]
;C[Being connected on a large scale, White goes for action elsewhere(still ignoring the excess cutting points)]W[bd]
;C[Lesson 1 but ...This is a good opportunity to say the contact play is not always bad per se and depending on tactical arguments it can sometimes be a surprisingly good move.There are no clear cut rules for playing well in Go!]B[cg]
;W[bg];B[ch]
;C[White is low, but since Black's forces are somewhat overlapping in the centre, White is ok with that]W[bh]
;C[Tactically Black is doing well\: he knows when to block and when to give way]B[bi]
;W[cd]
;C[very good\: Black closes off the centre from this side, while tucking white into the cornerAlso ...]B[df]
(
;C[White fixes his weakness but feels submitted to it]W[bf]
;B[ce];W[be];B[de]
;C[find breathing space]W[db]
;B[eb]
;C[a sneaky move]W[cb]
;C[Black takes the initiative to close off the centre. This shows gamesmanship]B[jf]
;W[kf];B[if]
;C[White takes time to fix his weakness]W[kh]
;B[ih]
;C[We're at the so called "end game" now, where all groups are stable and it's all about increasing/decreasing your/the opponent's territories]W[gl]
;B[fl]
;C[White returns here to intrude Black's territory]W[ec]
;B[fb];W[fc];B[gc];W[dc]
(
;C[Lesson 3\: beware of too many cutting points]LB[gb:5][ed:3][gd:4][dg:2][ci:1]B[fd]
;W[jb];B[ib];W[bj]
(
;C[A good sense for danger. We discussed the variation, seeing that Black cannot cut]B[cj]
;W[ai];B[ci];W[bk]
;C[The first real tactical error. This simply does not work]B[bl]
;W[ck];B[dk];W[cl];B[dl];W[cm];B[dm];W[ed]
;C[Lesson 3]B[ee]
;C[Exploiting cutting points]W[gd]
;B[fe];W[gb];B[hc];W[fa]
;C[Black now executes perfect endgame]B[gm]
;C[forced]W[il];B[jg];C[forced]W[kg];B[ja];W[ka];B[ia];C[forced]W[kb]
(
;C[last point]B[hm]
;W[im]
;C[giving back the initiative when there's nothing more to do]B[fm]
;W[ea];B[tt];W[tt]
;TB[dg:ig][dh:hh][di:gi][dj][ef:hf][ek:fk][el:em][fj:hj][gd:he][ii]C[White wins by 26 points thanks to - taking advantage of premature contact plays- a more efficient usage of his stones- exploiting cutting points]B[tt]TW[aa:da][ab:bb][ac:cc][ad:ah][aj:ak][al:bm][eb:fb][jl:mm][kc:me][kk:mk][la:mb][lf:mj]
)
(
;B[ga];W[ea];B[hb];W[ha];B[kk];W[kl];B[ga]
(
;W[gj];B[hj];W[ha];B[ah];W[ag];B[ga]
(
;W[fb];B[hm];W[im];B[fm];W[ha];B[lk];W[ll];B[ga];W[tt]
;TB[dg:ig][dh:hh][di:gi][dj][ef:hf][ek:fk][el:em][fj:gj][gd:he][ii]C[*** White wins by 25 pts., 1 pt. better for Black.]B[ha]TW[aa:da][ab:bb][ac:cc][ad:af][ah][aj:ak][al:bm][eb][jl][jm:mm][kc:me][kk:mk][la:mb][lf:mj][ml]
)
(
;W[hm];B[fb];W[fm];B[em];W[gm];B[gb]
;TB[dg:ig][dh:hh][di:gi][dj][ef:hf][ek:fk][el][fj:gj][gd:he][ha][ii]C[*** White wins by 25 pts., 1 pt. better for Black.]W[eb]TW[aa:da][ab:bb][ac:cc][ad:af][ah][aj:ak][al:bm][im:mm][jl][kc:me][kk:mk][la:mb][lf:mj][ll:ml]
)
)
(
;W[fb];B[hm];W[im];B[fm];W[ha];B[lk];W[ll];B[ga];W[gj];B[hj];W[ha];B[jm];W[jl];B[ga];W[tt]
;TB[dg:ig][dh:hh][di:gi][dj][ef:hf][ek:fk][el:em][fj:gj][gd:he][ii]C[*** White wins by 25 pts., 1 pt. better for Black.]B[ha]TW[aa:da][ab:bb][ac:cc][ad:ah][aj:ak][al:bm][eb][jl][kc:me][kk:mk][km:mm][la:mb][lf:mj][ml]
)
)
)
(
;B[ai];W[ci];B[cj];W[di];B[bk];W[dg]
)
)
(
;C[Sometimes, taking one point less, is better for overall strength]B[gd]
)
)
(
;C[If White plays elsewhere]W[jg]
;B[bf];W[ce]
;C[Black can capture 2 stones]B[ag]
)
)
(
;B[kh]
(
;W[lh]
;C[The centre stone is cut off, and White loses his efficiency]B[kg]
)
(
;W[kg];B[lh]
;C[White can capture these stones but ...]W[lg]
;C[When he cuts here too, the position crumbles]B[kk]
;W[lk]
;C[White can't prevent both captures at once]LB[lj:B][il:A]B[kl]
)
)
)
(
;B[kk];W[kl];B[lk]
;C[This leads nowhere for Black, who's already down to 2 liberties and no breathing room is to be found at the edge.This is a mental exercise White performed before playing move 12]W[ll]
)
(
;C[If Black jumps, his influence towards the centre is much bigger. Also the threat of tucking White into the corner, by playing A, represents more pressure on the white stones.Greater efficiency induces more risk\: White could cut at B. But this can be checked to be manageable. It should, given the amount of Black stones present in that area.]LB[kh:A][ji:B]B[jh]
)
)
(
;C[Keep distance. Now the opponent cannot take advantage of the contact play to weaken your attacking stone by removing its liberties. Still the pressure is on because the space in which that stone can operate is reduced. The same is true for the attacking stone, but it is backed up by the supporting stone on that side.]B[ij]
)
)
Re: A teaching game 13x13
Posted: Wed Jan 16, 2019 3:12 pm
by Vio
Wow if you did tell all this it's by far a lot for someone who just played a few games.
If he did follow all with shiny eyes he is already a promising player!
Re: A teaching game 13x13
Posted: Wed Jan 16, 2019 5:10 pm
by Knotwilg
Vio wrote:Wow if you did tell all this it's by far a lot for someone who just played a few games.
If he did follow all with shiny eyes he is already a promising player!
Well I don't usually overexplain and I don't know exactly how many games he had played. From this game I would think he's already past beginner stage. He's a smart guy too so I reckon he can handle the information.
But you're right in general: the review shouldn't contain too much information.
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 12:31 am
by Vio
Well, it was my very first feeling, and it was comparing it with how I would have handle it myself. I usually restrain myself as much as possible to comment on someone else way of teaching as it is very hot topic, especially on Internet with someone you aren't his friend yet! But here it is not like usual as your post is not oriented to your student but to us spectators so I felt invited.
When b for kakari did a contact move my feeling would be more mixed, the move is quite not that bad considering the lvl of my opponent, in the right direction first, and with a nice global view. OC it's not considered the optimal move to come, and it has a beginner taste that I would share your will to show, but it's not in my too bad move category. Indeed I think you created the 3_3 cut weakness in an attempt to sanction the contact move. And B over concentration will remain something a bit hard to explain too.
Because I feel we are more and less both same strength it's a bit hard to give a variant, I was thinking to go the path of "hane on the head of 2 stones" could be safer and a good didactic too, but eventually I would maybe just not try show anything yet and wait for a bigger failure to explain. Later in my own way I would feel very embarrassed to leave the cut weakness at 3_3 because I am playing psychologically, using the fact that my opponent knows by now I am strong so he will not try to cut and at same time I give him wrong idea of what is working or not. Well you had the very nice attitude to show him afterwards what he could have done.
In fact I am maybe a bit more relax and lazy, waiting to sanction bigger mistake with a low ddk. And I would have been more near your choice with a low sdk.
First try on my side to comment on pedagogic choice, so forgive me if you feel any rudeness, that is OC not in my attempt.
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 1:11 am
by Knotwilg
Vio wrote:Well, it was my very first feeling, and it was comparing it with how I would have handle it myself. I usually restrain myself as much as possible to comment on someone else way of teaching as it is very hot topic, especially on Internet with someone you aren't his friend yet!
Same here

Vio wrote:But here it is not like usual as your post is not oriented to your student but to us spectators so I felt invited.
And you were.
Vio wrote:When b for kakari did a contact move my feeling would be more mixed, the move is quite not that bad considering the lvl of my opponent, in the right direction first, and with a nice global view. OC it's not considered the optimal move to come, and it has a beginner taste that I would share your will to show, but it's not in my too bad move category.
You and Bill have convinced me that the contact play in the opening is not as bad as I told the guy, but my decision to talk about it was inspired by his playing the attachment 3 times. When I see a pattern of a (more or less) bad move, I consider it worthwhile to mention it.
eventually I would maybe just not try show anything yet and wait for a bigger failure to explain.
Absolutely. However there weren't any bigger failures, besides two tactical mistakes that he already nodded to in acknowledgment during the game.
Later in my own way I would feel very embarrassed to leave the cut weakness at 3_3 because I am playing psychologically, using the fact that my opponent knows by now I am strong so he will not try to cut and at same time I give him wrong idea of what is working or not.
You are right. Leaving such a complicated thing to figure out (at his level) is more bullying than teaching. That's what I said in the OP: I should have created easier things for him to take advantage of.
First try on my side to comment on pedagogic choice, so forgive me if you feel any rudeness, that is OC not in my attempt.
Thanks for the debate. That's the point of the forum!
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 1:50 am
by Vio
O my bad, I did miss bill analysis, I hope I didn't overlap too much!.
As low sdk, I had myself the hard way to lose to a low ddk with leaving a cut like yours and I can tell you it's hard to be pedagogic then with a bit of a sarcastic wind in the air...
O one more thing, I am not used to try to create things my opponent could use at his level (not to say I don't generate that by my own failure),maybe it's a path I could explore for better teaching. I am so focused to try to play my best moves!
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 4:48 am
by Elom
Thank you for this interesting post!
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 5:53 am
by Elom
A few thoughts on some points

.
Finding Patterns using abstract reasoning—
Can beginners use pro games to find patterns? (Pro mode)
Leave them to play games themselves and show you the patterns they find? (Amateur mode)
Playing up to yourself—
Start of with a few teaching games at even (Amatuer mode). Then suggest a handicap to the weaker player to play at full strength (pro mode)—their choice though.
If they want to becom pro yet don't want a handicap, they could have more to learn than go techniques... This comes from a memory of a time I went to learn table tennis. I only wanted to try things out-- getting good wasn't in my thoughts, so asked for my experienced opponent to keep serving at his full strength.
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 6:21 am
by Vio
I don't have so good memories from beginners refusing to take handicap: I had that happen a few times especially with chess players ("we don't do that") It ended in something too humiliating with the newcomer running away and me deeply sad to lose a player.
I had same problem with players who want to understand everything in their first game, spending hours of thinking the unthinkable.
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 7:14 am
by Knotwilg
Vio wrote:I don't have so good memories from beginners refusing to take handicap: I had that happen a few times especially with chess players ("we don't do that") It ended in something too humiliating with the newcomer running away and me deeply sad to lose a player.
I have let that go. It's true that an even game won't teach them anything because we can cruise on autopilot and their inefficiencies will accumulate to a big loss without us having to apply effort. You learn more from the even battle forced by the handicap. BUT ... if they resist handicap, forcing them into it will close the door for all learning, as they will be occupied with the humiliating feeling of having to take handicap.
Beware when you do have people who take handicap. If you play to win and win, it's adding injury to insult.
Vio wrote:I had same problem with players who want to understand everything in their first game, spending hours of thinking the unthinkable.
This is easier: I just tell them to play and we'll figure it out together later.
I've kept a blog of my teaching activities:
https://senseis.xmp.net/?DieterVerhofst ... xperiences
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 7:17 am
by Bill Spight
Vio wrote:O my bad, I did miss bill analysis, I hope I didn't overlap too much!.
All I did with regard to the attachment was to suggest a counter hane for

.

In the SGF I showed an improvement for Black at the very end. Instead of gaining 1 pt. on the bottom, Black could have played a ⅔ pt. sente on the top first and later won the final ko.
Re: A teaching game 13x13
Posted: Thu Jan 17, 2019 9:06 am
by Vio
Knotwilg wrote:Vio wrote:I don't have so good memories from beginners refusing to take handicap: I had that happen a few times especially with chess players ("we don't do that") It ended in something too humiliating with the newcomer running away and me deeply sad to lose a player.
I have let that go. It's true that an even game won't teach them anything because we can cruise on autopilot and their inefficiencies will accumulate to a big loss without us having to apply effort. You learn more from the even battle forced by the handicap. BUT ... if they resist handicap, forcing them into it will close the door for all learning, as they will be occupied with the humiliating feeling of having to take handicap.
Beware when you do have people who take handicap. If you play to win and win, it's adding injury to insult.
With handicap I feel pretty safe. Firstly I am very happy if they win, we can put a stone away.
Then I can play relax only correct moves, no hamete besides maybe the last game before changing handicap if they proved some nice progression,
As said before I just focused on showing how their most terrible moves are.
Thanks for the link to your blog, I will check that.
Happy teaching!
Re: A teaching game 13x13
Posted: Fri Jan 18, 2019 10:41 am
by Elom
Thank you for such a wonderful blog!