It's fairly common knowledge among more experienced players that Kogo's is not really that authoritative. A couple of days ago I came across a post on Baidu Tieba that was a pretty good illustration.

The below is a sequence given by Kogo's as joseki:

Click Here To Show Diagram Code

[go]$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . 2 . . . . |
$$ | . . . . . . . . . . . . . 1 6 5 0 . . |
$$ | . . . . . . . . . , . . . . 8 X 3 . . |
$$ | . . . . . . . . . . . . . . . 7 4 . . |
$$ | . . . . . . . . . . . . . . . . 9 . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . , . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Further moves omitted because not that relevant (and also I don't quite understand why even that result would be considered joseki except under very very limited situations). Let's also not discuss how Kogo's also suggests playing 8 at 10 (giving the ponnuki???) for now XD

The post (at http://tieba.baidu.com/p/1812033521) suggests that instead of 10, black should play the descent as below and white's shape ends up completely broken:

Click Here To Show Diagram Code

[go]$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . 2 . . . . |
$$ | . . . . . . . . . . . . . 1 6 5 . . . |
$$ | . . . . . . . . . , . . . . 8 X 3 . . |
$$ | . . . . . . . . . . . . . . . 7 4 0 . |
$$ | . . . . . . . . . . . . . . . . 9 . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . , . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

That post made me wonder what other known errors are in the dictionary. I know that there's one instance in the below sequence where two identical positions (through transposition of two sequences) were given different assessments (one for which a previous move was called "questionable" while the other was called "possible") and different continuations (!). The actual sequences are too long to list and I'm too lazy to create an sgf file, but it starts off as below:

Click Here To Show Diagram Code

[go]$$Wc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . 0 . . 8 . |
$$ | . . . . . . . . . . . . . . 3 4 6 7 . |
$$ | . . . . . . . . . , . . . . X 5 1 9 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . 2 . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . , . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Does anyone know of any other obvious mistakes in Kogo's?

I once played that 2nd diagram noseki as white and black did descend. However, I found some move around r11 that was sente and then my corner was resurrected .

Another Kogo's brilliancy is:

Click Here To Show Diagram Code

[go]$$Bc It starts of with a 3-3 invasion of 4-4...
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . 7 6 . . . . |
$$ | . . . . . . . . . . . . . . 5 4 2 . . |
$$ | . . . . . . . . . , . . . . . 1 3 8 . |
$$ | . . . . . . . . . . . . . . . . 0 9 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]

Click Here To Show Diagram Code

[go]$$Bc11 ...and ends with "Black gets a thick position and White a living L+2 group."
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . X O 4 . . . |
$$ | . . . . . . . . . . . . . . X O O 2 . |
$$ | . . . . . . . . . , . . . 5 . X X O . |
$$ | . . . . . . . . . . . . . . . 1 O X . |
$$ | . . . . . . . . . . . . . . . . 3 . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]

I managed to find an improvement on this 'joseki' in http://www.online-go.com/games/board.php?boardID=144596

Wow, that's a pretty serious order mistake XD Thanks for sharing!

For a moment I just looked at the final result in Uber's diagram and was like "That's totally joseki".... Oh order of moves, you are actually important.

_________________
Tactics yes, Tact no...

Hmm. in the first diagram never appears in my game database (GoGoD, SmartGo, and others), even when adding the few thousand amateur games I have. Is that Joseki?

There's a variation in our beloved joseki below that Kogo's has no knowledge of. It's not that common, but still... Here goes.

Click Here To Show Diagram Code

[go]$$c Normal so far...
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 5 . . . . . . . . . . . . . . . |
$$ | . . . . 4 . . . . . . . . . . . . . . |
$$ | . . 1 , . . . 6 . , . . . . . , . . . |
$$ | . . . 7 2 . 3 . . . . . . . . . . . . |
$$ | . . . . 8 . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Most of you know the regular continuation...

Click Here To Show Diagram Code

[go]$$c Boring
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 9 . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . . . O . 4 . . . . . . . . . . . . |
$$ | . . X , . 8 3 O 2 6 . . . . . a . . . |
$$ | . . . X O . X 1 5 . 7 . . . . . . . . |
$$ | . . . . O . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Recently I've been studying the variation given by Takeo Kajiwara, who hates this result as Black and doesn't consider it joseki (especially with a Black hoshi at a) in his book "The Direction of Play". According to him, the only way to play is and below.

Click Here To Show Diagram Code

[go]$$c Shake things up
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . . . O . 2 3 . . . . . . . . . . . |
$$ | . . X , . . 1 O . . . . . . . , . . . |
$$ | . . . X O . X a . . . . . . . . . . . |
$$ | . . . . O . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Kajiwara then goes on to show how to deal with different responses, including if White plays a instead of . He considers every variation a success for Black if no mistakes are made. He also explains why Black can play regardless of the ladder.

Kogo's comment for is "Black [1] is a mistake", followed by the weakest variations Black could play and White punishing them. It doesn't even attempt to explore .

It's too bad to see such a great project fail in both research (as shown here) and common sense (as shown in the thread in general).

I'd like to see one of the statisticians from the 'monkey vs Yi Chang-ho' thread offer his opinion of the chances of constructing a joseki database without a mistake somewhere.

_________________
Help make L19 more organized. Make an index: https://lifein19x19.com/viewtopic.php?f=14&t=5207

It is zero

It's not zero (for the same reason why it's possible to reach perfect play through random selection of moves if nothing else, let's not even go to the trivial counter-example of a database consisting of one joseki).

Nonetheless, I find Kogo's an invaluable and convenient resource. I did not start this thread to complain, but rather to sate my curiosity (and also to poke fun at some variations).

Clearly you don't understand the mystical force that creates errors In joseki dictionaries.

No, it really is zero. Consider Muphry's law.

_________________
Occupy Babel!