I'm compiling, slowly, the 300 biggest mistakes from 100 games I played earlier this year.

https://senseis.xmp.net/?DieterVerhofst ... OfMistakes

It's clear that holding them all on one page will not be the most compelling presentation. Also, halfway I've changed the presentation mode from an ABC problem - choosing between my move, the correct move and a third move - to a corrective explanation. Another aspect is the emergence of patterns, such as the "slow connection".

I'm thinking of the following ways to present the lot

* chunking into say 10 pages of 30 diagrams
* in SL's problem format, so a problem page (ABC) and a solution page for each
* thematically, so that all the diagrams related to slow connections serve as examples

Any preferences, suggestions? (How) are you likely to engage with this kind of material?

My vote: in SL's problem format, so a problem page (ABC) and a solution page for each.

But in addition, make a page summarizing the classification of your mistakes, with 2-3 examples for each category.

For me the gold standard is Brady's Blunders but I gather that you are asking how to organize it on sensei's, not if you should produce videos of it.

Alas - I've never got round to produce videos that stand today's standards. One day maybe ....

An update: some more meat to the bones of the observed patterns.

https://senseis.xmp.net/?DieterVerhofst ... OfMistakes

It's a long shot but I intend to build chapters of examples along these lines.

Dieter - There are a number of books written by professionals along the lines of what are the key things that amateurs don't know, I think their topics chosen are subjective but based on their experience of teaching for say 20 or 30 years.

I think your analysis of the 3 top mistakes in 100 of your games is invaluable, it's based on actual analysis of a smaller sample but it's maybe more objective.It's also aimed at around 1 dan, rather than say 5-6 kyu or higher dan - where I suspect the mistake categories are different.

So I encourage you to proceed - looking at Sensei's Library I am not sure if you can cover up the answer, as it stands you present a problem with a, b or c and then give the answer next to the problem - I would rather the answer was hidden.

Stick to your categories - it would be nice to get input from a teaching professional on them.

I could provide details on some Japanese books which claim to cover what amateurs don't know. I am not sure how your categories agree/disagree - this is an interesting area!

One question in "Mistake 2" on SL you say "The right side is the big/urgent area to play." but move "c" is on the left side.

Take Care - John

The main page is messy as it is. I started out with the abc format but now I'm categorizing them.



Maybe if I get it into an accomplished state I can get a pro teacher enticed to look at it. Thanks for the suggestion.



Please do!

Thanks a lot for the encouragement.

The first book that looks at mistakes amateurs make that I suggest is:

"Aiming at Shodan - Three faults that hold you back" - there is a very thorough review at:

https://senseis.xmp.net/?AimingAtShodanThreeFaultsThatHoldYouBack

This was published in January 1996 - Gaku Yazuru - was a 5 dan professional, then aged 56 - so he would have played a lot of teaching games.

I think his opening words are key: " the greatest impediments to improvement in amateur Go" "are those bad playing habits which are unconscious." "lead to bad moves; but since we are unaware of them we take no corrective steps. So we continue to play less well than we could, with slow upward progress."

the book starts with 8 common mistakes and then concentrates on just 3.
- not distinguishing junk stones from key stones
- shirking a fight in an area where you hold the initiative
- not playing tenuki when you should

Interesting. I suspect that if you get stuck around 1-2 kyu for a long time then you need a good teacher - to show you your "bad playing habits" AI will not do paint the bigger picture, it can show you mistakes but not fit them into categories helpful for the kyu player.

Don't rush to try and find the book, there is enough in the SL review.

Take Care - John

Rather than recompiling the 2021 lot, I'm presenting middle game exercises from my KataGo reviesws in 2022

https://senseis.xmp.net/?DietersMistakesIn2022

Very interesting, thanks!
Can I check the answer somewhere after I have made a guess for myself?

I'll add solution pages later

I'm enjoying these, even though they're a bit above my level. I hope you continue adding to the collection.

One minor note: problem 3 doesn't indicate whether it's B or W to play

The series is now formatted as problems and solutions. The problems are often stated as a selection between a, b and c, but in some cases there's an open question, like "What can Black do now?"

https://senseis.xmp.net/?DietersMistakesIn2022

The work is done

- https://senseis.xmp.net/?DietersMistakesIn2021 holds 151 middle game problems, framed as a-b-c or a-b choices, taken from mistakes in my games, with separate solution pages
- https://senseis.xmp.net/?DietersMistakesIn2022 likewise holds 57 problems and counting.
- https://senseis.xmp.net/?DieterVerhofst ... InMistakes analyzes patterns in these mistakes, which may be "idiosynchratic" but could be representative of the significant improvements low dan players can make.

As mentioned on my practice page, three of the most frequent patterns are:

- take the opportunity to surround a group if it is in sente, i.e. it can be killed if ignored; if surrounding is not sente, then playing elsewhere may be better
- when cutting opponent stones or connecting your own stones, check if both groups are indeed vulnerable; otherwise this will result in a slow connection or a heavy cut
- when activity approaches a previously stable position with some aji, be aware that the opponent group or your group may be killable

https://senseis.xmp.net/?DieterVerhofst ... InMistakes

now gives an overview of the mistakes, organized per recognized pattern.

This concludes the editing work.

Now I'll go back to playing.

Thanks for sharing. I've saved it. I reviewed the squeezes and I found it helpful.

But I'm more interested in hearing about how this exercise has helped your game (or not) as time goes on. Let us know later.

I don't know if you care for comments like this but Raw Peep #49 doesn't discuss the Raw Peep #130 doesn't show on the board. It's probably not worth the effort to go back and change it now.

I am curious if you create problems from only your own mistakes or if you consider the opponents mistakes too? I think I am assuming it is only from your own mistakes because there is a "my" in the subject title.

Are you willing to share the SGFs of the games the problems are from? I tried creating some problems from my own games but I think I don't have this skill because I find some fault in every problem. Somehow you seem to have this all figured out. You were also claiming something about the distribution of mistakes between opening, middlegame and endgame.

Yes, my mistakes. The assumption was that I can improve most by "unlearning" my biggest mistakes, with KataGo's point loss as the driving heuristic. Next up is to reverse it and take good moves. The selection will fall on those moments in the game where I was making a deliberate choice in a difficult moment and made the right one.


Willing yes, but unfortunately I didn't keep track of which game in "my games" led to which numbered problem



I haven't figured this out as much as I'm an avid user of Sensei's Library diagram creating feature https://senseis.xmp.net/tools/sgf2diagram.php



Indeed. I counted more than 1 endgame situation. About 5/210. Opening is debatable but not more than 5 I'd say. So the vast majority of my big mistakes comes from my middle game. The heuristic used may make this a self fulfilling statement.

Thanks for the interest, as always.

Quantitatively, as I told John too, I feel a revived, more robust 2d now, in a landscape which may have deflated since AI came about. Qualitatively I can say there are plenty of occasions where I'm wary of the slow connection, the raw peep, surrounding in sente ... so I feel like better player too. The hardest part is getting rid of the non-conceptual, lazy, blitzy blunders. and infuse my game with consistent reading. For this I'm now playing some very slow paced games and also doing more tsumego to their full extent (i.e. don't stop at the solution, using them as a reading practice). This is 30 years of bad practice that needs to be destroyed and healing again.




You can always edit the page with such a comment. I will fix it. Thanks.