Incorrect bluff based on assymmetric information

[flackr]

Version (PM the bot with /version): v1.4.11 Convention settings: /setall 11 Steps to reproduce or replay link: https://hanab.live/replay/1199386#15 Additional information: robot2 gives 3s to muffincake which could work as a bluff to muffincake who knows that it can't be red 3 but should probably still be an invalid bluff since robot1 can't tell and keeps playing.

[pedroteosousa]

Shouldn't robot1 just assume a bluff when the blind-played card doesn't connect? I might be misunderstanding something, but to me if seems like the bluff is valid and the problem was it being interpreted as a layered finesse even tho layered finesses are "turned off" from bluff seat.

[flackr]

From Cathy's connecting principle,

look at whether the card that blind-played "connects" to the clue that was given. If the clue connects, then it is a Finesse. If the clue did not connect, then it is a Bluff.

Since the played card does connect to a 3 clue, and robot1 has no way to know that muffincake knows it can't be an r3, it must be a finesse from robot1's perspective.

From Mistaking a layered finesse for a bluff, it says:

• However, Bluffs take precedence over Layered Finesses from Bluff Seat.
  • In other words, if Alice gives a clue which causes Bob to blind-play an unrelated card immediately, then Bob should assume a Bluff when possible.
  • In still other words, Layered Finesses are generally "turned off" from Bluff Seat.
  • Of course, this only applies if the clue actually satisfies the criteria for a potential Bluff.
  • In particular, if Alice gives a clue to Bob but Bob does not blind-play immediately, then according to the Lie Principle, the clue should always be interpreted as a Layered Finesse as it cannot possibly be a legal Bluff.
  • In particular, if Alice gives a clue to Bob which is not a legal Bluff target, the clue should always be interpreted as a Layered Finesse.

Emphasis on generally "turned off", there are exceptions. From robot1's perspective I think there's two possible assumptions:

• muffincake doesn't play p3 because they know it can't be r3 because they can see both.
• muffincake doesn't play p3 because from their perspective it can't be a bluff as the played card connects but they see a possible layered finesse.

[pedroteosousa]

That makes sense. I though the "connection" needed to be rank and color, but that doesn't make sense because the connection is evaluated from Cathy's perspective. Thank you!