Discussion about Q16,R14,O16,P15,P16,S16,P13

[JosekiDictionary]

At position Q16,R14,O16,P15,P16,S16,P13, @aesalon started the conversation:

p13 is the most common local move. Stop demoting moves please.

( FYI @aesalon )

[mark5000]

Although it may be the most common move (when you shrink the search space enough), it’s more good than ideal as we defined them. For reference:

IDEAL: A correct move. This is a perfect move that cannot be shown as inferior to another choice. Of course, multiple Ideal choices often exist for each position. A series of Ideal moves forms a joseki.

GOOD: A move that is almost perfect, but can be shown to be slightly inferior to the Ideal choice. This move may be played if outside circumstances (like a big wall) dictate an unconventional approach. Or possibly it’s a way to avoid complications while only sacrificing a small advantage. A Good move cannot be part of a joseki.

Since “pro play” is the only source, I looked at the games to see if they commonly had an outside circumstance. I found that they did. For example:

• Zhang Yingting v. Gan Siyang (enclosure and extension in adjacent corner)
• Meng Tailing v. Li Qincheng (huge wall up to the 11th line)
• Wang Chenxing v. Gao Xing (huge wall and extension in adjacent corner)
• Cho Hanseung v. Kang Dongyun (weak group needs help)

(Same with more recent games not on Waltheri.)

Also since “pro play” is the only source, I verified the positions with KataGo. Indeed, it liked the move with favorable context but not without. Without context, the difference between this move and its first choice was 2.1 points. On balance, I feel there’s not enough to say this move will become joseki. At best it’s a prototype joseki used when there’s a significant area of influence to build from.

[aesalon]

You can turn this into quite the rabbit hole… and again: Play Go at online-go.com! | OGS is still ‘IDEAL’. Basically, you need a much higher threshold to demote to ‘GOOD’.

The basic rule is not to use the Good move type unless absolutely sure. There are few of these moves compared to the Ideal choice.

Is another guideline for reference.

[Eugene]

I’m keen about extended and insightful well researched debate. Let’s explore the rabbit holes and return with good answers.

This is one of the things that differentiates our Joseki Explorer from others: great, well researched and vetted content.

If I may note:

The basic rule is not to use the Good move type unless absolutely sure. There are few of these moves compared to the Ideal choice.

I’m not sure this in itself is a barrier to demotion. If a move is not Ideal (therefore is demoted), then it is either Good or it doesn’t belong at all. “Good or not at all” is a separate decision from “Not Joseki (Ideal)”.

I think the statement “there are few Good” also deserves consideration in the light of our ongoing experience. Since we have a high bar for Ideal, maybe we should be more accepting of Good… if we can’t find a source, can’t prove it is Ideal/Joseki … should that mean we don’t get to include it at all?

[aesalon]

The rabbit hole begins with performing the same objective review(using whatever AI weights you fancy in that minute) of these moves: https://online-go.com/joseki/531 (Most likely to be demoted to ‘GOOD’) and https://online-go.com/joseki/13083. Then comparing whatever metric you like in the moment (KataGo points, % points, professional play) to the review of the candidate move under discussion.

Anyway, I can guarantee that this isn’t a ‘GOOD’ move.

[aesalon]

You can see that dragging AI into it is a mess. Let the professionals analyze with AI and show us their conclusions.

Then we can gogo discussion and debate.

[mark5000]

@aesalon

I can guarantee that this isn’t a ‘GOOD’ move.

One thing they teach you in law school is to show, don’t tell, when trying to persuade others. I’m interested in @aesalon 's reasons for the move being ideal. Obviously aesalon feels strongly about it.

I showed the reasons supporting my evaluation of the move. The games where the move was played had special circumstances. I didn’t want to go with just my observations, since I might be wrong. So I wanted to check them with something else. When the AI answer was the same, I figured that with that smoke, there was probably a fire, and I demoted the move. Since a good source doesn’t exist, I felt that this kind of process was appropriate.

If I understand aesalon correctly, I should not have checked AI (unreliable and not authoritative). So all I have is my observations. If that’s the case, we should take a closer look at the data, since I do want to move closer to agreement on this issue. I provided some more recent pro game data below.

Newest Pro Games

Here’s a 12x12 square of professional games from the dataset that @aesalon gave me. Since the square is large, D is played 3 times. D isn’t the most common move when it doesn’t have context. And one of these games is the Zhang Yingting vs. Gan Siyang game I posted above, with the enclosure and extension in the adjacent corner.

This is a 10x10 square of the same dataset. Since the square is smaller, B is now the most common move. Pros play it more often when stones occupy the 11th and 12th lines from this corner.

These are those six games, excluding the game already posted above. I still think the takeaway from these games is that there’s usually some circumstance that makes the move attractice, whether it’s erasing the opponent’s area of influence or building a moyo with help from a favorable outpost.

[aesalon]

If it is Good, it is Good compared to what Ideal? Black blocking at the 3-3 is very rarely played.

If it is situational depending on Black wanting influence then Black already opted into that by backing off one space high.

Same 10x10 search with more games: https://snipboard.io/F9Vktm.jpg and still zero cases of Black 3-3. So is the pincer the only Ideal move? I think not.