Suggestions to Prevent Score Manipulation

[AdamR]

@knusbrick

the engine doesn’t handle the Japanese seki rules correctly at all

I am actually reasonably sure it does? David Wu - katago creator took huge pains to implement proper Japanese rules. And KataGo would be the one scoring these cases.

@_KoBa

if players disagree the score for, say, 3 times?

I think we should track two things:

1. The scoring was rejected, and resumed (here 3 sounds like a good start)

2. How many times the board state was updated (For players who just click away any proposed state) Here I would go much higher like 20 updates maybe?

[knusbrick]

@AdamR

I am actually reasonably sure it does? David Wu - katago creator took huge pains to implement proper Japanese rules.

Oh really? That’s actually really cool then, I was under the impression that that’s still a challenge!

[Myrsilos]

Even if it’s genuinely architecturally infeasible to “reverse” the Elo results of a finished game, you should be able to create hidden dummy games that yield the correct Elo point outcome.

[jonquil]

I don’t think it’s really possible and it’s genuinely up in the fantastic books, sadly