Life In 19x19

What are the different kinds of mistakes you can make and which are the once that hold us back?

There are undoubtedly different opinions on which mistakes are the worst in general and players' own idea of which mistakes hold themself back are probably even more variate. Still, I have the suspicion that there is a generality here that applies same to everyone. That is, the effect of a mistake may not be intrinsic of the player's rank but instead intrinsic of the mistake type.

I'm not sure if we can easily distinguish between perception and reality but it seems a typical sdk player you find on the Internet thinks the worst mistakes are in order:

1. Gote
2. Not playing the computer move
3. Not playing very active

It looks more unreasonable than it maybe is. I'll try to exaplain:
#1 Programs like KataGo only occasionally play gote moves and oftentimes, considering the unclear follow up, it feels right to give them sente.

#2 If you always play the best computer move you would easily be 12 dan professional.

#3 It's probably more fun to play active, so that makes sense even if not doing anything might often be a more effective approach.

My list is not necessarily as developed and logic proof as the list from the fictious but typical sdk on the Internet, still I offer it now:

1. Bad or backward logic
2. Blind spots
3. Applying incorrect lessons from experience
4. Missing chances to play big endgame
5. Not knowing something
6. Not seeing something from afar
7. Bleeding in small endgame
8. Point loss compared to top engine move

I can try (and fail) to explain:

#1 When you fail to respect the basic logic of the game it is inevitable that bad things happen. Most of the time it is possible to get a close enough game and stay within a reasonable margin by playing sure moves. It is the same when you fail to stay in the game, there is always a point when you could have recognized that you were going wrong and reverted to something that worked better. Alas, only if you had better understanding of how Go actually works.

#2 When you completely miss something it is pretty bad, you can miss a move in a variation or you can simply miss a move in the current position. Top pros sometimes miss something completely, it is less frequent but time and the depth of reading being attempted is a factor. Beginners miss something every move, and everyone else is somewhere between beginners and top pros. Often one misses the same kind of thing again and again, making it a real problem instead of a mere happenstance.

#3 We did something in a previous game or saw someone else do something and we drew conclusions. When we encounter a similar situation we go back to this memory, even if it is something horrendously stupid. This seems to both happen as a regular mistake and also when autopilot sets in.

#4 Missing the chance to play a big endgame is game over in close games. It can be disastrous to make a habit of it if it starts to affect our perspective of what is a close game. Still, the big endgame can be complex and it is possible to miss your chance to play a move to go straight to the "game over" screen.

#5 When we don't know some life and death or some follow ups it is pretty bad but we can learn. There are just so many things you should just know and your opponent will know, which is why this can't be much lower on the list.

#6 Sometimes the effect of some move is not clear until many moves later. What looked like a good move can fail when you realize much later that you missed something that only becomes clear when many moves have been played. I'd say it happens a lot, it is something akin to the horizon effect experienced by computer programs using alpha-beta pruning, at some point you just said "enough, I can figure it out when it is on the board" and later you figure that it is not great. This is different from a blind spot, which is involuntary omission of a move, as this is caused by deliberately not reading to the end.

#7 Mistakes in the small endgame lose us close games.

#8 If it isn't one of the above mistakes and you still can't pick one of the engines top choices I'd submit that your choice might still be reasonable even if not the best. The margin allowed here does decrease for stronger players but it also depends on the complexity of the position — probably even top pros sometimes have to be satisfied with whatever move the can come up with.


Possibly my list can be labeled and treated as ungraciously as I have the fictious SDK list Still curious what the rest of the L19ers think, maybe I miss something (i.e. CGT, East-West dichotomy,...) or have a completely wrong perspective.

Absolute Beginners

1. Not being able to visualise moves.

2. Total lack of knowledge.

DDKs

1. Not playing simple.

2. Not connecting important stones.

3. Not defending life of important stones.

SDKs and Low Dans

1. Making too many blunders.

2. Losing more by endgame-like moves than imagining in the worst nightmare.

3. Many gaps in knowledge and skill (which includes tactical reading especially of L+D).

Amateur High Dans

- Having some inferior knowledge and skill than stronger players. (Every player has different weaknesses or strengths but overall it is inferior for playing strength.)

- Having slower tactical reading speed than professionals.

- Having slower endgame calculation speed (at 1/2 point accuracy) than professionals.

Professionals (guess)

- Having some inferior knowledge and skill than stronger players.

As the OP said, different opinions will exist. And they do even among pros. The most dramatic list I ever saw was one by Kita Fumiko, and that was because it was so long. I have summarised it in my Go Wisdom appendices, but even in that form it has a depressing number of ways of cocking up a game. I also recall a series by an amateur in a national go journal (or was it Go Moon?), which was maybe not quite as long, but it spun the agony out over many months between issues!

However, I would imagine that the thinking behind any version of the list is to offer ways to secure an improvement in one's grade. For most of us, there is a quick way to do that, and it can lead to a massive improvement.

The trick is simply "come alive!" Or "think" or "concentrate" or phrase at as you will.

My old GoGoD marra T Mark Hall used to play on KGS at 4-dan and occasionally upwards. He kept his grade for years. Then he fell ill. Stuck at home played even more on KGS, but with his mind naturally on other things, his grade fell precipitously and reached 2-kyu at the end. His go knowledge did not worsen. It was just that his mind was not focused on the go board.

My experience is that this is fairly typical of amateur play, whether you are ill or not. You just let other things take away your focus and your grade tumbles. My experience, again of watching others, is that a four-grade fall is not unusual (especially among oldies, and some acknowledge that by entering tournaments at a lower grade - and get accused of sandbagging when they do concentrate!), but even the lack of attention implied in playing too quickly and thoughtlessly can (will?) knock at least two or three grades off your best one.

So, if this lack of attention applies to you (and it is very, very, very common - maybe even more so in online play?), you can expect to show a marked improvement at once. Many of the bad moves in the lists above, or in Kita's list, will disappear for sure.

I have seen the converse, where a player achieves tournament results above the grade expected of one of his go knowledge simply because he concentrates intently during tournaments. Apart from helping him avoid blunders, this often has the side-effect of exasperating the stronger opponent who then blunders!

It may seem that I am advocating inducing blunders, and also ignoring the ultimate goal, which is improvement beyond one's current best grade. Not really. Steeling your mind to concentrate and so play at one's real best grade not only provides immense personal satisfaction, but you can also convert that feedback into studying properly, i.e. by thinking hard. Recall that "effortful practice" is how the experts tell us to use our 10,000 hours.

Right, so what's on the telly now, and is there a beer in the fridge?

My only research is about my own games: https://senseis.xmp.net/?DieterVerhofst ... InMistakes

The table of contents of that page indicates there are 4 high levels of mistakes:

- a lack of awareness: not realizing what's going on, often the status of a big group
- a lack of or bad reading: realizing what's going on but simply failing to read correctly, or not making much effort at all
- bad strategy/wrong purpose: being aware that something is going on but not making the right decision
- bad technique: making the correct decision but carrying out poorly, due to poor move selection, insufficient pattern recognition, suji/haengma

Of these, the lack of awareness can be incidental. Not seeing an opportunity/danger to kill/live doesn't necessarily mean one is bad at life & death, or that one is playing too casually. But if it occurs often, either or both are the case.

Bad/no reading is probably the hardest bad habit to cure. One thing I recently found is that tsumego practice can be detrimental to reading. I find myself playing the most likely move, because I know there is a solution. That's a big clue you don't have in games.

Bad decision making can be more easily cured, I think, by studying and by game reviews.

Bad technique, likewise.

Knotwilg wrote:tsumego practice can be detrimental to reading. I find myself playing the most likely move, because I know there is a solution. That's a big clue you don't have in games.

Each tactical problem has a solution but not all solutions are simple. Maybe you only consider problem collections with simple solutions of their problems?

RobertJasiek wrote:

Knotwilg wrote:tsumego practice can be detrimental to reading. I find myself playing the most likely move, because I know there is a solution. That's a big clue you don't have in games.

Each tactical problem has a solution but not all solutions are simple. Maybe you only consider problem collections with simple solutions of their problems?

I think what happens is that my capacity for intuitive problem solving is growing faster at the intuitive level than at the reading level. This may be true for all of us. Especially when revisiting the same or similar sources, patterns will be recognized.

So yes, I should switch collections more often.

At my level (2k FFG), the big mistakes I see in my games or in my opponent's games are of two kinds:

1) Mis-judging the strengths and weaknesses of groups, and hence

• Not seeing that a group of the opponent is attackable.
• Failing to defend a group and then being under heavy pressure.
• Defending a group that doesn't need to be defended. So either the opponent gets a free move elsewhere, or he responds and repairs his own weaknesses at the same time.
• Attacking a group which is already strong. The opponent doesn't need to respond.
• Making a purely defensive move while double-purpose moves (attack+defense) are available.

2) Sloppy reading:

• Just playing the shape move without reading. The two groups look connected, but are they? The group seems to have two eyes but does it?
• Starting a ko without checking ko threats first.
• Playing a known sequence that requires a ladder to work but forgetting to check the ladder beforehand.
• Failing to consider all reasonable responses. There is a cut, I checked that the ladder doesn't work but forgot the net, or forgot to check if a peep would be annoying. Or I invaded and checked a few variations but forgot that the opponent can first lean on another group to make strength and then come back to kill the invading stone.
• Only considering "normal" moves. I thought about the jump but did I think about that weird empty triangle?

That said, being strong at go is not just about eliminating big mistakes. The game is long and the accumulation of small mistakes outweighs the big ones. And I'm talking about SDK mistakes (i.e. mistakes that an SDK can easily understand). A player who eliminates them might reach low dan level but would be light-years away from pro level. Pro games are a mystery for me, I sometimes replay them but I have the impression that they play a different game.

jlt wrote:Pro games are a mystery for me, I sometimes replay them but I have the impression that they play a different game.

Until 1995, pro games were easier to understand. Afterwards, they became more complex. I started to understand them again when realising that modern pro games run aspects of opening, middle game and endgame simultaneously from move 1 on. Later, AI since AlphaGo has done the same. I think that this flexibility has also spread to amateur games, although more so among stronger players.

Until 1995, pro games were easier to understand.

Why 1995?

Afterwards, they became more complex. I started to understand them again when realising that modern pro games run aspects of opening, middle game and endgame simultaneously from move 1 on.

I think we see this in old Chinese go. It's plainly evident in concepts such as zhao ying. It's implied in Huang Longshi's Five Lands theory which turns the go board into a map where populations can be seen to grow in response to changing conditions. It's clarified in Xu Xingyou's commentaries, which distil Huang's teachings.

I think it's clear, too, in old Edo go, too, but we lack contemporaneous commentaries of that period. (There's a case to be made that Edo players copied Chinese theory and didn't let on - we know they had the Chinese books.)

It's clearly evident in the debate over New Fuseki - and is Go Seigen really "easier to understand"?

We even see the simultaneity idea in amateur theory. Several amateurs have talked about it, but the neatest expression I ever saw was by a top British player, Paul Prescott. I don't think it's a mere coincidence that he was a topology expert (like AI bots?), but he put his theory in a nutshell like this: A move that does two things is better than a move that does one thing. A move that does three things is better than a move that does two things. And so ad infinitum. Is it also a coincidence that AI stands for artificial intelligence?

1995 because until then I understood pro games, then for a couple of years, when they became more complex, I did not.

Third language terms can be a mystery to me: what is zhao ying?

(You know perfectly well that AI bots do not understand topology with gender 1+. If AI shall not be called "artificial intelligence", we might need to discuss more fundamentally beginning from definitions of "information", such as input -> black box process -> output, and identifyable amidst physical objects.)

RobertJasiek wrote:AI bots do not understand topology with gender 1+.

You probably mean genus 1+.

jlt wrote:At my level (2k FFG), the big mistakes I see in my games or in my opponent's games are of two kinds:

1) Mis-judging the strengths and weaknesses of groups, and hence
(...)
2) Sloppy reading:
(...)

That largely corresponds with my overview.

jlt wrote: That said, being strong at go is not just about eliminating big mistakes. The game is long and the accumulation of small mistakes outweighs the big ones. And I'm talking about SDK mistakes (i.e. mistakes that an SDK can easily understand). A player who eliminates them might reach low dan level but would be light-years away from pro level. Pro games are a mystery for me, I sometimes replay them but I have the impression that they play a different game.

Kvasir made a similar argument when I related of my focus on eliminating big mistakes. I agree that it's not enough to reach pro level - at all! I stick with my viewpoint, for two reasons:
1) focusing on what eliminates big mistakes tends to improve my overall game play; I can't prove that, it was a feeling I got recently: increased awareness, more disciplined reading ... what is required to avoid big mistakes tends to have an effect on the smaller ones
2) more importantly, eliminating big mistakes simply increases probability of victory bigtime; the positive feedback by winning a game is strong; a big mistake will undo the aggregate of small improvements and the frustration will wipe out the reinforcement one should get from those

In my case:

Lack of reading

doesn't matter what is the cause

lazy
tired
slow
untrained
lack of focus

I’m interested in what would be the main categories though I probably mixed together in my list a high level breakdown and some specific problems that could possibly be put under a more generic umbrella.

I think many of the specific examples could be put under the generic heading of bad logic. Possibly that is a big umbrella, but it should include many things, including what may at first appear to be a problem of faulty reading, probably most problems of bad judgment are a combination of three things: 1. not understanding what is logical in Go, 2. Not knowing something about the shape or situation, 3. Not being able to read everything. I think the last one is a given, there are too many things to read, and you need to rely on #1 and #2 to make #3 manageable.

I would say things like: Defending the wrong weakness, Not looking to exploit opponents’ weakness, Not handling weak stones, Not playing simple, Bad technique and so on are often symptoms of a faulty approach in how to think and how to make decision in Go. I called that “bad logic”.



It was pointed out that lack of concentration and energy is often a problem, maybe it is a mistake to play when you don’t have energy (sometime happens to me) but it is not always possible to find a better time. It is likely that it is possible to manage once concentration and pace the effort to get better results.


I think some of the examples are what I’d call applying incorrect lessons from experience. For example, failing to take care of weak groups and failing to consider every move, sound like something that I have sometimes struggled with and can in part be explained by faulty experience. In case of not taking enough care of weak groups it is sometimes the result of all those endless games where we got away with this It feels good to play loosey-goosey when it works, it is almost like playing two moves in a row since the defensive move can be useless if the opponent would never take advantage. Same with not checking every move that you need to check, it feels much better to steamroll your opponent with moves that look like they work than it feels to agonize over a tesuji that refutes your idea

This reminds me of another type of mistake which is to see ghosts, which is one of the ways that studying can cause us to play worse (hopefully only temporarily while we find a good balance at a new skill level). It is both possible to play too carefully because of some ghosts and it is possible to strike at an imaginary weakness. I think this applies to most forms of study, the new thing learned is often over emphasized.

I think many of the specific examples could be put under the generic heading of bad logic. Possibly that is a big umbrella, but it should include many things, including what may at first appear to be a problem of faulty reading, probably most problems of bad judgment are a combination of three things: 1. not understanding what is logical in Go, 2. Not knowing something about the shape or situation, 3. Not being able to read everything. I think the last one is a given, there are too many things to read, and you need to rely on #1 and #2 to make #3 manageable.

There is a book somewhere in my labyrinthine library that is called something like "Mistakes amateurs make" and the sub-text is "... and pros don't". The overall difference is that amateurs have bad suji and pros have good suji.

Suji can sometimes be rendered as style, but in the sense you would say a pro dancer has style and an amateur doesn't, or a singer, or a writer. They look or sound good because what they do looks effortless and graceful. What I am getting at is that they are not putting on an act of being stylish, which is what go amateurs usually mean by having a style. The pro style is effective. The amateur style is affective.

Now the interesting thing about suji is that, even just within go, it can be used (and so translated) in lots of ways, but one of them is 'logic'. It's no surprise when you consider the etymology of the word. It refers to tendons, and from that derive references to long, stringy things, including veins and plots - and logic, of course.

When suji is used in go it refers mostly to an effective flow of moves. To put it another way, it refers to an effective flow of moves (and effectiveness is often associated with keeping an attacking flowing).

Pros, in their mother's womb or their cradle or wherever, have learned to make effective flows of moves instinctively. People who learn late in life - like us - tend to make one move at time, with no flow - no suji or bad suji. Our go is hiccuppy.

We can sometimes make a nice tesuji, where the te- (hand) adds a nuance of something being crafted. It thus becomes an identifiable sequence, something we can copy and so learn. That doesn't stop many people thing tesuji is just one brilliant move, thus omitting the suji element.

So how do pros acquire good suji, apart from starting young? Another book on my shelves is one by Yoda Norimoto in which he tries to explain this - unsuccessfully in my case, and I think in most people's cases because his theory never seemed to get any traction. He also spends a lot of time in the book metaphorically pulling his hair out because (Japanese) amateurs can't grasp what he is talking about. We amateurs seem to be hard wired differently, it seems.

What he posits is a concept (within a larger framework of tewari) called sujiba - the 'suji spot'. I would like to suggest that 'suji sweetspot' is a more idiomatic rendering. That would be nice if assume he believes each suji sequence has a spot within it that really makes the line work efficiently. It's a sort of packed gear train but with differential gearing and you have to know where to put the differential wheels to turn the corners. So you have to be a go engineer.

However (and this is why I think he fails), he says sujiba (a term of his own creation) refer to places where stones work inefficiently. As I see it, he is working backwards because he is a pro. He sees a spot in a sequence where the amateur characteristically makes a mistake. I would have thought the existing term sujichigai covers that adequately. However, it has to be said that sujichigai refers to the whole wonky line. Yoda is referring specifically to one 'spot'. He says the amateur badness. I see the pro goodness. In other words, I think it would be better to reserve sujiba for the 'good' place to play. It all comes to the same thing. It's just a glass half full versus a glass half empty. Therefore, in my own mind I think of it as a 'suji sweetspot.'

This still doesn't help with finding these spots. Yoda's approach is just to give examples. But I sense that for those willing to make an effort to think of suji sweetspots, using existing tesujis as templates where the goodness of a certain spot is highlighted, and often easy to grasp, Another approach is to follow the pro approach of understanding what each play characteristically does. For example, if you want to change the direction of play in a running battle, the 'good' ways to do, say pros, are the tactical manoeuvres of cuts, nobis, tsukes, caps, inducing moves, hazama, hasmaitsuke, sabaki and peeps. There are bad ways, too, but I'll leave those as an exercise for the reader (clue: if they are not in that list...).

In essence, I am agreeing with the quote above. What I am adding is the opinion (with explanation) that there is a existing mental apparatus, a go-based one, to help us cope, and it is one that has the backing of Japanese pros. So, try suji instead of sushi!.