top of page
Writer's pictureTakumigo

The Story Behind Creating Go Problems (Commemorating the 50th Problem Uploaded)


As of today (August 31, 2024), among the 161,230 problems in the 101 Go Problem Library, 52 of them are my original creations. These problems have been completed over 15,000 times, allowing thousands of Go enthusiasts to discover or learn new tesuji. Although it's just a small contribution, I hope it adds something to the Go community.


My attempt at creating Go problems was influenced by two main factors. Initially, I have always enjoyed sharing tesuji from real games. The reason is that whenever I see a "good move," whether from professional matches or AI recommendations, I feel a sense of joy and excitement. Wanting to share this joy with others, I used to share these Go tactics within small groups. Later, when I joined a larger community, I felt even more motivated to find interesting tactics. Although I eventually realized that what I find interesting may not resonate with everyone, the support from a few was enough to keep me going.


Another reason was my discovery of the charm in Go problems. Strangely enough, I've been playing Go for almost 13 years, but I hardly ever did problems during the first 12 years. It wasn’t until a year ago, after playing some rapid games and realizing my quick calculation ability was lacking, that I decided to practice with problems. The rich features of the 101 platform greatly helped me find motivation to work on problems, which led to my discovery of their intriguing nature. Gradually, I thought, "Why not try creating my own problems?"


There are a few things to keep in mind when converting real-game tactics into problems. First, whenever possible, the global Go position should be turned into a localized problem. Then, it’s best to make one of the tactical moves into the only correct move in the problem. In other words, if the correct move isn’t found, the game cannot be won, resulting in a black-and-white outcome. Moreover, in real games, there are often multiple sequences that lead to the best result, but in problems, it’s better to reduce the number of possible sequences, especially for the first move. Lastly, the main theme of the problem should be clear; you can’t just take any position and call it a problem.


On the 101 platform, problems don’t go straight into the library after being uploaded; they must be approved by a professional player. Therefore, not all uploaded problems make it into the library. A problem has to be "good enough" to be accepted, which pushes me to ensure the quality of the problems and improve them whenever possible. Over time, I feel that my Go skills have also improved through this process.



Among the 52 problems I've uploaded, the one above is my representative work, as it combines several key elements:


  1. The blind spot of the first move—this problem comes from a real game where both players, who were of at least fox 7 dan strength, overlooked the correct first move, even after some thought.


  2. The uniqueness of the corner—the third move is a classic problem that also represents a very realistic shape.


  3. The same shape but different moves—though the shapes may look identical, they require different moves, highlighting the subtlety of Go. Shapes cannot be simply memorized; they require constant adaptation.


In the upcoming articles, I will continue to introduce my top five favorite problems. Please stay tuned for more!

35 views3 comments

Recent Posts

See All

3 Comments


Jarosław Jusiak
Jarosław Jusiak
Sep 09

It's surprising to me that you were barely solving tsume go for most of your "go journey" and yet you are able to develop good reading ability. Is it a matter of natural talent (to aquired the skill just through playing go) or perhpas you did practise it in some other form?

Like
Takumigo
Takumigo
Sep 12
Replying to

I played a lot of games and I usually would try to spend most of my time to calculate in actual games. I hate to make mistakes, so I tried to avoid them as much as possible.


My talent is not bad, but I don't think it is far from average, In my level, I think I am the one who have more knowledge than reading abilities.

Like

Philippe Fanaro
Philippe Fanaro
Sep 01

Maybe put all your creations into an 101weiqi book?

Like
bottom of page