More of these in https://senseis.xmp.net/?DieterVerhofst ... OfMistakes
They contain 3 mistakes each of my last 3 games (will add more) which either lost big in terms of winning percentage or points, usually both. It's instructive to see that these big swings still occur in the middle game or early endgame, occasionally late in the endgame and then due to bad reading, losing a group or having to sacrifice a tail. Opening mistakes occur too but typically range in the 1-2 points department. Mistakes in choosing the bigger boundary play rarely matter either.
This means improving technique (move selection in combat) and reading will still make the big difference for me and I assume most players at my level (1-2 dan).
Knotwilg's practice
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
-
kvasir
- Lives in sente
- Posts: 1040
- Joined: Sat Jul 28, 2012 12:29 am
- Rank: panda 5 dan
- GD Posts: 0
- IGS: kvasir
- Has thanked: 25 times
- Been thanked: 187 times
Re: Knotwilg's practice
I noticed that it was not actually a killing move, but still a very good one.
If white resists it is less good for white.
If white resists it is less good for white.
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
-
kvasir
- Lives in sente
- Posts: 1040
- Joined: Sat Jul 28, 2012 12:29 am
- Rank: panda 5 dan
- GD Posts: 0
- IGS: kvasir
- Has thanked: 25 times
- Been thanked: 187 times
Re: Knotwilg's practice
Maybe there was a problem with your second diagram? I think you intent this, which looks right to me.
and if white tries harder, for example.
and if white tries harder, for example.
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
Re: Knotwilg's practice
(last diagrams posted were indeed correct - I made a mistake)
Taking another course now, I've checked my last 25 games on OGS. I won 16 and lost 9. The remarkable finding is that I was leading in the opening on 23/25 occasions. In 5/25 I took that lead straight into victory. In the other 11 victories there was a swing in the lead once or twice (exceptionally more) and I took it back, or those 2 times I trailed in the opening I swung it into my favor. In the 9 losses, 5 times my opponent took my lead away from the middle game in 1 swing and 3 times in 2 swings.
If we take this as a representative sample, I can say I'm almost always better in the opening (23/25), allow the lead to be taken away in the middle game more often than not (20/23), but half of those times (9/20) I lose the game the other half I take it back.
This means I should not invest so much in my opening and can count on fighting spirit later in the game, but should pay more attention to the early middle game.
Another interpretation is that I play the opening more or less like AI does, fairly territorial with emphasis on the corners, but fail to mimic the way AI handles the influence my opponents usually get. That corresponds to an impression I have always had, which is I'm not good at dealing with opponent moyos, when they form. I could now take two approaches: improve that, or switch back to more balanced or even influence taking openings and avoid the burden of dealing with moyos.
Taking another course now, I've checked my last 25 games on OGS. I won 16 and lost 9. The remarkable finding is that I was leading in the opening on 23/25 occasions. In 5/25 I took that lead straight into victory. In the other 11 victories there was a swing in the lead once or twice (exceptionally more) and I took it back, or those 2 times I trailed in the opening I swung it into my favor. In the 9 losses, 5 times my opponent took my lead away from the middle game in 1 swing and 3 times in 2 swings.
If we take this as a representative sample, I can say I'm almost always better in the opening (23/25), allow the lead to be taken away in the middle game more often than not (20/23), but half of those times (9/20) I lose the game the other half I take it back.
This means I should not invest so much in my opening and can count on fighting spirit later in the game, but should pay more attention to the early middle game.
Another interpretation is that I play the opening more or less like AI does, fairly territorial with emphasis on the corners, but fail to mimic the way AI handles the influence my opponents usually get. That corresponds to an impression I have always had, which is I'm not good at dealing with opponent moyos, when they form. I could now take two approaches: improve that, or switch back to more balanced or even influence taking openings and avoid the burden of dealing with moyos.
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
Re: Knotwilg's practice
While I'm still completing my collection of mistakes at https://senseis.xmp.net/?DieterVerhofst ... OfMistakes
Next the plan is to play 10 games without making any https://senseis.xmp.net/?Slonection
Next the plan is to play 10 games without making any https://senseis.xmp.net/?Slonection
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
Re: Knotwilg's practice
I've created a new SL track on my mistakes for the year 2022. The first 20+ such mistakes show the following
Most substantial errors are failure in recognizing the value of a move, compared to a move elsewhere.
A recurring subcategory is missing an opportunity to seal in a big group in sente, meaning that group will either have to live small or perish if the opponent doesn't answer locally. On the other hand, sealing in a group in gote, instead of a big move elsewhere, is a mistake I made too: the opponent can take the big point without major punishment.
So that is my major goal for the next games, to see and correctly assess such opportunities. Other examples in the category of positional judgment:
- preferring central influence over a big corner
- overstating the threat of a cut, as both groups were safe or could resist - the counterpart of a slow connection
- a single minded influence move, aiming at developing a moyo without pressuring an opponent group
- recognizing a killing pattern, but overestimating the value of the kill versus the value of being forced into that kill from the outside
- allowing the opponent to force an already healthy group into low development
- playing a (big) corner reinforcement (enclosure), rather than an expanding the shape and health of an important group of stones, which could be turned into rubbish if neglected
- playing a lukewarm defensive move, while it was possible and easy to decide the game with another kill
- trying to kill with insufficient backup, allowing an escape into a complex fight, rather than surrounding and threaten to kill next
The second category of mistakes is when the value of local play is understood but poorly executed. Two subcategories: not seeing or knowing a tesuji and poor to no reading.
A third category of mistakes is more related to awareness. Technically speaking it's also "making a wrong choice" but here it's more not even being aware of the options.
Most substantial errors are failure in recognizing the value of a move, compared to a move elsewhere.
A recurring subcategory is missing an opportunity to seal in a big group in sente, meaning that group will either have to live small or perish if the opponent doesn't answer locally. On the other hand, sealing in a group in gote, instead of a big move elsewhere, is a mistake I made too: the opponent can take the big point without major punishment.
So that is my major goal for the next games, to see and correctly assess such opportunities. Other examples in the category of positional judgment:
- preferring central influence over a big corner
- overstating the threat of a cut, as both groups were safe or could resist - the counterpart of a slow connection
- a single minded influence move, aiming at developing a moyo without pressuring an opponent group
- recognizing a killing pattern, but overestimating the value of the kill versus the value of being forced into that kill from the outside
- allowing the opponent to force an already healthy group into low development
- playing a (big) corner reinforcement (enclosure), rather than an expanding the shape and health of an important group of stones, which could be turned into rubbish if neglected
- playing a lukewarm defensive move, while it was possible and easy to decide the game with another kill
- trying to kill with insufficient backup, allowing an escape into a complex fight, rather than surrounding and threaten to kill next
The second category of mistakes is when the value of local play is understood but poorly executed. Two subcategories: not seeing or knowing a tesuji and poor to no reading.
A third category of mistakes is more related to awareness. Technically speaking it's also "making a wrong choice" but here it's more not even being aware of the options.
-
marvin
- Dies in gote
- Posts: 51
- Joined: Thu Oct 11, 2012 1:36 pm
- Rank: EGF 1 dan
- GD Posts: 0
- Has thanked: 65 times
- Been thanked: 5 times
Re: Knotwilg's practice
Thanks for composing the problems, I am enjoying them 
.
Some mistakes could be also labelled as "shape", for example, problem 21. There I would play the correct move immediately because it is important from the local shape point of view.
.Some mistakes could be also labelled as "shape", for example, problem 21. There I would play the correct move immediately because it is important from the local shape point of view.
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
Re: Knotwilg's practice
As "Sensei's Library Recent changes addicts" will have noticed, I'm refactoring the 2021 Mistakes page into a path of subpages with problem and solution. While doing that editing work, I revisited the (missing) solutions, some of which already eluded me but most of which have stuck. In 2022 I'm continuing the exercise. It continues to follow distinct patterns.
1. While the corners-sides-center paradigm that has been strengthened by AI insights, remains valid, that is not where the difference is made at my level. All major swings happen in the middle game, or in the end game due to a blunder/oversight
2. A couple of patterns are about cutting & connecting
- don't make slow connections, instead defend in a more active, threatening or points making way, or even play elsewhere altogether; a kind of anti-honte
- likewise don't make lukewarm threats to cut the opponent, to which the reply would be a slow connection
- don't create heavy groups of cutting stones, while the groups you cut are relatively safe or strong
- many mistakes are not about whether to cut or connect or not, but HOW, so bad technique or reading
3. surrounding and the related concept of "bullying"
- surrounding the opponent IN SENTE, is the best kind of move you can make, trumping big moves in the corner
- surrounding the opponent IN GOTE, on the other hand, allows the opponent to take a big point elsewhere and your follow-ups will not be killer moves but taking some points while threatening to kill
- while not quite "surrounding in sente", there's often big value in bullying/harassing the opponent, forcing them to make a capture or escape less efficient
4. in 2021 I have been following a KataGo inspired territorial strategy often leading to big frameworks for the opponent. I often made bad choices in invading, when & where; here I could no longer live up to the advantage KataGo was assigning to my position; AI have an intuition for invasions which is way over my head and which is hard to learn from them
5. reading and awareness
- more generally, there are sequences that I could just not come up with; this is the usual criticism with AI but only a minor portion of the mistakes made
- at the other side of the spectrum, there are sheer blunders or bad moves made under time pressure
- related to that, some of the mistakes are due to a certain sequence affecting the states of groups in the vicinity, while I'm not aware of it
Since 2022 I have been playing a more influence oriented opening, with 5-4 points and pincers. The point loss of that choice hardly ever goes beyond 1,5. In the middle game, where the bigger swings occur, that influence works mostly to my advantage. I would say an influence game is easier to play, provided there is middle game fighting. If the opponent leaves all that influence alone and keeps taking points on the sides, a central moyo strategy may result from it and that becomes a gambit for both.
I'm also under the impression my mistakes are shifting from technique/reading towards strategy: i.e. from how to play towards where to play. I'm still misjudging urgent for non-urgent and vice versa, playing the big move too soon or too late.
1. While the corners-sides-center paradigm that has been strengthened by AI insights, remains valid, that is not where the difference is made at my level. All major swings happen in the middle game, or in the end game due to a blunder/oversight
2. A couple of patterns are about cutting & connecting
- don't make slow connections, instead defend in a more active, threatening or points making way, or even play elsewhere altogether; a kind of anti-honte
- likewise don't make lukewarm threats to cut the opponent, to which the reply would be a slow connection
- don't create heavy groups of cutting stones, while the groups you cut are relatively safe or strong
- many mistakes are not about whether to cut or connect or not, but HOW, so bad technique or reading
3. surrounding and the related concept of "bullying"
- surrounding the opponent IN SENTE, is the best kind of move you can make, trumping big moves in the corner
- surrounding the opponent IN GOTE, on the other hand, allows the opponent to take a big point elsewhere and your follow-ups will not be killer moves but taking some points while threatening to kill
- while not quite "surrounding in sente", there's often big value in bullying/harassing the opponent, forcing them to make a capture or escape less efficient
4. in 2021 I have been following a KataGo inspired territorial strategy often leading to big frameworks for the opponent. I often made bad choices in invading, when & where; here I could no longer live up to the advantage KataGo was assigning to my position; AI have an intuition for invasions which is way over my head and which is hard to learn from them
5. reading and awareness
- more generally, there are sequences that I could just not come up with; this is the usual criticism with AI but only a minor portion of the mistakes made
- at the other side of the spectrum, there are sheer blunders or bad moves made under time pressure
- related to that, some of the mistakes are due to a certain sequence affecting the states of groups in the vicinity, while I'm not aware of it
Since 2022 I have been playing a more influence oriented opening, with 5-4 points and pincers. The point loss of that choice hardly ever goes beyond 1,5. In the middle game, where the bigger swings occur, that influence works mostly to my advantage. I would say an influence game is easier to play, provided there is middle game fighting. If the opponent leaves all that influence alone and keeps taking points on the sides, a central moyo strategy may result from it and that becomes a gambit for both.
I'm also under the impression my mistakes are shifting from technique/reading towards strategy: i.e. from how to play towards where to play. I'm still misjudging urgent for non-urgent and vice versa, playing the big move too soon or too late.
-
rhwocns
- Dies in gote
- Posts: 34
- Joined: Wed Oct 28, 2020 11:51 pm
- Rank: 3k in a local town
- GD Posts: 0
- Has thanked: 6 times
- Been thanked: 5 times
Re: Knotwilg's practice
kvasir wrote:
Knotwilg wrote:I guess this is the answer
kvasir wrote:
Last edited by rhwocns on Wed Feb 09, 2022 2:38 pm, edited 52 times in total.
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
Re: Knotwilg's practice
This was a strange but satisfying game at the OGS 1k-1d level. I won by 80 points after a pretty straightforward piece of fighting. I continue playing the 5-4 point to have a more influence oriented opening. In this case my opponent allowed a shimari at 23.
In the light of the discussion elsewhere I'd indicate the following as either "dan moves" or moves I've probably acquired since evaluating my games with KataGo:
23 - playing away because of a slow opponent move (gote)
31 - the pincer/invasion, playing more severely in these circumstances
37 - a known technique, descending to connect either way (miai)
45 & 49 - two good peeps, carefully evaluated since KG dislikes peeping for the sake of peeping
(most of my mistakes come from 50-110)
113 - playing an angle point thanks to the backup of the hanging connection
(after that knock-out, the lead went from 40 to 80, with some minor hiccups)
In the light of the discussion elsewhere I'd indicate the following as either "dan moves" or moves I've probably acquired since evaluating my games with KataGo:
23 - playing away because of a slow opponent move (gote)
31 - the pincer/invasion, playing more severely in these circumstances
37 - a known technique, descending to connect either way (miai)
45 & 49 - two good peeps, carefully evaluated since KG dislikes peeping for the sake of peeping
(most of my mistakes come from 50-110)
113 - playing an angle point thanks to the backup of the hanging connection
(after that knock-out, the lead went from 40 to 80, with some minor hiccups)
-
John Fairbairn
- Oza
- Posts: 3724
- Joined: Wed Apr 21, 2010 3:09 am
- Has thanked: 20 times
- Been thanked: 4672 times
Re: Knotwilg's practice
To continue my baseball theme, here's something out of left field for you.31 - the pincer/invasion, playing more severely in these circumstances
This is an example of my language-based approach to things, as opposed to a numbers base: I can't run katago so I have no idea what katago will make of it. I could be way, way off, but even if I prove to be right my approach will probably seem alien to most L19ers. Still, in the interests of research...
31 seems fine to me. But I note you can't make your mind up what to call it. That indicates confusion, which is surely not a good mindset. However, your over-riding feeling is wanting to play 'more severely.' That seems good - but hold on to that thought for the moment.
Invasion or pincer or both? Well, it can't be a pincer. But it is an uchikomi. That is usually rendered as 'invasion.' Now I don't want to go so far as to say that is wrong, but I think it's misleading. There is a strong association of 'occupation' of the enemy's territory with 'invasion'. In other words, an invasion is a prelude to occupation from within. That is what happened in the game, and so if 'invasion' really was the intended idea, it has been achieved.
But to me an uchikomi, in its Japanese sense, which feels more natural to me here, is driving in a wedge. That creates a different set of associations - and therefore leads to different play.
So when White plays 32, the prime candidate that springs to mind for me is not your nobi 33, but the kosumi one point above. That continues the idea of driving in the wedge. In territory terms, it results in annexation as opposed to occupation.
That's half my point. Now severity. In games play we make a distinction between a play that is 'severe' and one that is 'sharp'. A severe move is one that results in bullying (or overconcentration or whatever) by one side against the other. A sharp move is one that results in complications for both sides. Or, to put it another way, that is severe for both sides. But such a way of play is not really severe for you - you are too busy defending yourself to get in a real hammer blow (i.e. an uchikomi). In my view, the nobi to 33 in the game is a sharp move: it results in complications for both sides. In contrast, my kosumi is truly is a 'severe' move (unilaterally, and correctly so, given Black's thickness to the left).
I'm exposing myself to possible ridicule once we know what katago will tell us. However, I am quietly confident because this particular kind of situation has been in my mind for decades. The nobi at both ends (33 and 37) is one of those tesujis we were introduced to when all we had available was Go Review and the like. It had bling! I imagine that was part of the Nihon Ki-in's idea: to make go appear to us as truly a beautiful game. I was certainly captivated and played this tesuji more than a few times. But it never really worked out for me the way I hoped. I also noticed that pros didn't seem to play it much - they preferred the kosumi (or even tenuki). But there were a few instances that were famous (e.g. Shuwa liked double nobis). Then it dawned on me that they were famous partly because they stood out, i.e. they were NOT standard - despite, as you say, us being taught it was a standard miai-connection idea. But all that glitters is not gold. Later on, it further dawned on me that the famous instances that did work worked because they were not the result of an uchikomi. They may have been the result of an invasion (English sense), but, as I say, that's really a different idea. They were more about sabaki, say, than trying to be severe.
It is now the pantomime season here, and so I feel as if I walk on stage as the wordy old dame ready to be assailed by custard pies from the numerous numbers villains who lurk on L19. But katago is supposed to be dispassionate. (Oh no it isn't! Oh yes it is!). What does it say about the kosumi for 33? What does this have to do with the price of fish?
-
Knotwilg
- Oza
- Posts: 2432
- Joined: Fri Jan 14, 2011 6:53 am
- Rank: KGS 2d OGS 1d Fox 4d
- GD Posts: 0
- KGS: Artevelde
- OGS: Knotwilg
- Online playing schedule: UTC 18:00 - 22:00
- Location: Ghent, Belgium
- Has thanked: 360 times
- Been thanked: 1021 times
- Contact:
Re: Knotwilg's practice
Believe it or not, I thought of including "wedge" in the list of options. I didn't think it was a pincer, because of it being "pre-counterpincered" nor an invasion because of the missing idea of "occupying enemy territory" which you stated. But I didn't think it was a wedge, because that has a notion of contact to me - and I mean, regardless of general use, intuitively a wedge is in touch with either side of the wedged.John Fairbairn wrote:But to me an uchikomi, in its Japanese sense, which feels more natural to me here, is driving in a wedge.31 - the pincer/invasion, playing more severely in these circumstances
"Separation" maybe?
Thanks for the elaboration. I agree with the notions of sharp and severe - i.e. they resonate with me.Now severity. In games play we make a distinction between a play that is 'severe' and one that is 'sharp'. A severe move is one that results in bullying (or overconcentration or whatever) by one side against the other. A sharp move is one that results in complications for both sides. Or, to put it another way, that is severe for both sides. But such a way of play is not really severe for you - you are too busy defending yourself to get in a real hammer blow (i.e. an uchikomi). In my view, the nobi to 33 in the game is a sharp move: it results in complications for both sides. In contrast, my kosumi is truly is a 'severe' move (unilaterally, and correctly so, given Black's thickness to the left).
Rest assured: KataGo suggests the kosumi but doesn't think the nobi is a mistake. The descent tesuji at 37 is obviously necessary to make 33 work.The nobi at both ends (33 and 37) is one of those tesujis we were introduced to when all we had available was Go Review and the like. It had bling! I imagine that was part of the Nihon Ki-in's idea: to make go appear to us as truly a beautiful game. I was certainly captivated and played this tesuji more than a few times. But it never really worked out for me the way I hoped. I also noticed that pros didn't seem to play it much - they preferred the kosumi (or even tenuki). But there were a few instances that were famous (e.g. Shuwa liked double nobis). Then it dawned on me that they were famous partly because they stood out, i.e. they were NOT standard - despite, as you say, us being taught it was a standard miai-connection idea. But all that glitters is not gold. Later on, it further dawned on me that the famous instances that did work worked because they were not the result of an uchikomi. They may have been the result of an invasion (English sense), but, as I say, that's really a different idea. They were more about sabaki, say, than trying to be severe.
You are right: the combo is a kind of "gluttony". I see it and I play it. I like the straight shape better than the diagonal shape too. But just like in the other thread, I realize we may have been fed with the particular, sketchy, rather than the generic, solid.