Participation in a study on Go in the field of mathematical psychology

Here is the text that a friend of mine, another go player and psychology student, asked me to share.
He needs participants of all levels and nationalities for a research for his Bachelor’s thesis.
I invite you to participate if you find it interesting :slight_smile:

I invite you to take 45 minutes to an hour of your time to participate in a study on the game of Go and mathematical psychology.
The work aims to build and validate a particular mathematical model, for the assessment of the procedural skills used in solving life and death problems.
This model can be extended and used to build automated teaching tools.
For its validation, the participation of players is required, you could help by answering 17 tsumego in this questionnaire
Thank you in advance

Of course, if you have any questions or would like to leave feedback feel free to write here or to the email you can find in the questionnaire


i started and left the survey halfway. some feedback:

no south sudan in the country list.

writing in coordinates is tedious. if it were playable tsumego with putting stones on the board, i probably would have completed the full survey.

if all problems are white to kill, but moves for both white and black must be correct, why should the correct move for black not be to tenuki immediately everytime? this is a feature of tsumego that black would play a strong resistance, but for me as a tsumego solver to try and guess what the creator of the problem thinks is the strongest resistance is not enioyable.

maybe a tsumego site like blacktoplay or tsumegohero can help make the survey be more like solving real tsumego and get more responses that way?


I also quit part way through. Frankly the problems seemed a bit too tough for me, and the interface of typing coordinates is quite tedious.

Actually, in the example tsumego provided in the instructions, was there supposed to be a bug in the example solution? I don’t see how it works and it seems to involve an illegal move?

With how difficult and tedious the problems were, I wonder if seeing at what point people give up, maybe even with the timing data gathered, is perhaps the real aim of the experiment?


Same here.

I wouldn’t mind putting these tsumego’s into a puzzle collection so that it would be easier for people to solve them.
But if you are halfway, you can not go back into the questionaire. So that is not possible.
Also getting back to the page with the email address is not possible otherwise I would have given this psychology student some feedback.


no south sudan in the country list

I’d be much more interested in a South Sudanese Go player than in this survey.


I completed the questionnaire and solved 11/17 problems after some effort. Several problems are similar to each other so if you got one of them wrong, then you probably got several others wrong. The email addresses are


First problem I was given had two solutions :thinking:

1 Like

I think several good answers are acceptable, the problem is considered as solved if you provide one of them.


Done. Got 100%. Pretty tough puzzles :joy:


Thank you all for your feedback :blush:!

At the moment, this is the only way we have to present the problems and collect the responses.

As some of you have guessed, most of the problems have many different alternatives responses. We have done our best to include all of them in the automated correction. However, if you feel we have missed something, please, you are more than welcome to contact me.

Feel free to contact me directly at

Ugh, that shape.

Carpenter’s Square

I knew I don’t know the answers. Only got a few correct. Didn’t consider a certain proverb, as I should have, looking at the solutions. I also got confused with sekis, kos and shapes to almost-fill. And I’m wondering, why it’s forbidden to submit shorter sequences. This and having the minimum length given as information, did influence my answers.


This was pretty enjoyable to do, despite having to input coordinates :slight_smile:

It felt a little like a “learn by doing” crash course on the shape. It gives me the idea to create some learning resources in a similar style for common shapes such as the L-group or J-group.


Maybe it’s better to say in advance that it helps the participants learn something and is not just a test (and I think the goal of this research study, ultimately, is to help people learn stuff like math?) so that a participant wouldn’t worry about failing to solve some questions.

Also the problems are too difficult for ddk. IMO participants below 5k can hardly benefit from this set of problems. Maybe you could indicate that at the beginning, too.


well here i am. though only a go player in south sudan rather than a south sudanese go player



I completed the survey. Or should I say, the tsumego exercise-book. (I feel a bit like I was tricked into taking a crash-course on the ******** ******.)

Thank you for presenting these great go problems. I have two questions:

  • Can I subscribe somewhere to be informed when the study is published? I like reading math papers, psychology papers, and more than anything papers about go, so a study about mathematical psychology in go problems should be very interesting to read.
  • Are the go problems available somewhere? I make flashcards with go problems so that I can easily study a few go problems everyday, so if you have the source with the problems in text-form I would love if you could share it.

I do have some criticism about the way the survey is presented.

My first criticism is about the need for spaces in the solutions. Why? Why do you need spaces? Do you know how much longer it takes me to type in “A1 A2 B1 B2” as opposed to just “a1a2b1b2”? Why are spaces so important if coordinates are always alternating letter/digit/letter/digit anyway? Please don’t make spaces mandatory.

My second criticism is about the need to guess correctly black’s response. For at least two of the problems, I got a wrong answer because my first move (white) was correct, but my second move (black’s response) was “wrong”. Black died in the end anyway. How am I supposed to guess which way you expected me to kill black?

As a side-note, the introduction of the survey was “The problem is always: ‘White to kill’. The solution of any problem can be unconditional life, seki, ko, unconditional death.” That doesn’t make a whole lot of sense for me. White to kill, results in unconditional life or seki? If that is possible, just say “White to play”, not “White to kill”. It’s not important but I’m pointing it out anyway.

Cheers, and thanks again for this fun problem collection.


you can see all the problems and the correct responses at the end of the questionnaire,

For further requests and clarifications, here are our emails


Thank you very much for your participation and your support.
Yes, we intend to publish an article on the results of this study, in case it will be my care to highlight it in this forum. Also, for further requests and clarifications you can contact me or Andrea at the addresses we provided in the survey


Just to mention, at least one of the problems I got incorrect because the correct solution involved a line I recognized quickly was doomed, and didn’t consider the line I wrote (which of course was also doomed, because the problem ended in death), all this to say, it’s hard to be sure if the input is correct for some problems. Be careful of the conclusions you draw from this exercise. Not to mention the similarity of problems will lead people to certain solutions based on the order they’re given.

Yes I’m ddk and yes I got all of them wrong, but dad-gum-it I deserved that one lol :stuck_out_tongue:

1 Like

Aaah, UniPD, la mia Alma Mater (più o meno) xD


I got 10 / 17. Carpenter’s square is so tricky :laughing: