Bluff vs Layered finesse

[Rhydon MK2]

I think this was already discussed but I couldn't understand the previous post.

Field: R2-Y3-G1-B0-W3-M4-K5
A: X(R5) -X -X-X-X
B: X -X -X-X-X
C: X -R(R3) -X-X-X
D: X(W4)-X -X-X-X
E: X -X -X-X-X

B gave red to A to make C play red card (R3) and D's first card (W4).
What was discussed was whether D should interpret the clue as layered and dig for r4, or they should interpret it as bluff.

1. 'It's bluff': It is bluff because there was no gap of play before D responded. D's bluff seat changed to B.
It's implication: we cannot apply Layered finesse in following situation.

Field: R2-Y3-G1-B0-W3-M4-K5
A: X(R5) -X -X-X-X
B: X -X -X-X-X
C: X -R(R3) -X-X-X
D: X(W4)-X(R4) -X-X-X (or whatever sequence of playable cards -R4)
E: X -X -X-X-X

2. 'It's layered': It is layered because B cannot bluff D. Bluff seat does not change.
It's implication: we cannot bluff people not in your next position, so it may cost some tokens.

Field: R2-Y3-G1-B0-W3-M4-K5
A: X(R5) -X -X-X-X
B: X -X -X-X-X
C: X -R(R3) -X-X-X
D: X(W4)-X -X-X-Y5
E: X -X -X-X-X
B cannot give R to A because it will cause bomb (E should play leftmost as they know D will bomb next turn), so they may resort to less efficient clue, like just letting C make D play and making C play later.

I wonder what is more popular convention in BGA.

[Jellby]

I think this was already discussed but I couldn't understand the previous post.

Field: R2-Y3-G1-B0-W3-M4-K5
A: X(R5) -X -X-X-X
B: X -X -X-X-X
C: X -R(R3) -X-X-X
D: X(W4)-X -X-X-X
E: X -X -X-X-X

B gave red to A to make C play red card (R3) and D's first card (W4).
What was discussed was whether D should interpret the clue as layered and dig for r4, or they should interpret it as bluff.

C will believe they have R4, R3 and will play R3. If no one else responds, they'll play the first X next round.
A will believe they have R4. If no one else responds, they'll play R5 after R3.
D will not see R4 anywhere, they'll believe they have R4 and play W4
At this point, the line is broken, a bluff is revealed and no one has "passed" to suggest a layered move.

The risk is that C might clue D's W4 in some way instead of playing R3.

[Romain672]

Try to make A the one which give the clue or the first player.
I will by default assume you don't play with hgroup's layered finesse.

Let's take a simpler situation first:
Played: r2.
B x-x-x-x
C r3-x-x-x
D y1-x-x-x
E x-r5-x-x

B clue red to E.
Here C assume r3+r4.
C should have 100% r3.
But the chance for C to have exactly r4 too is really low. It's way more likely that D would try to play r4 as y1.
If you decide to ignore that, you will be able to do a rare 3for1, but will lose the ability to clue red here. It's like if you clue 5 on g5 turn 1 (with g clue impossible). You can assume you got like g1+g2+g4 (if there is a g3 in finesse position), but the probability for it to be the case is so low, that assuming something else, like a potential mistake will be far better.
So I will assume on this case you would just play r3 as C.

Now, if D's hand is y1-x-x-g5, the clue became riskier whatever you choose. But for me you should just trust B he would not do that clue with r3+r4, and still play r3.
Even if you got r3+r4, there is still a good chance D will clue something which reduce that risk.

And you can see, you got here a bluff not on the next player: D isn't promessed r4 because he played that card in bluff seat since C was 'forced' to play r3.

---

So I will vote 2 for your situation.
But the problem with that post is that most player which will discuss and share their opinion on the forum will be in average 'better' and will in average play with 'higher sensibility' (ie will play with higher 'trust', with more complicated clues), so you should get more 2 votes.

[Travis Hall]

I think this was already discussed but I couldn't understand the previous post.

Field: R2-Y3-G1-B0-W3-M4-K5
A: X(R5) -X -X-X-X
B: X -X -X-X-X
C: X -R(R3) -X-X-X
D: X(W4)-X -X-X-X
E: X -X -X-X-X

B gave red to A to make C play red card (R3) and D's first card (W4).
What was discussed was whether D should interpret the clue as layered and dig for r4, or they should interpret it as bluff.

C will believe they have R4, R3 and will play R3. If no one else responds, they'll play the first X next round.
A will believe they have R4. If no one else responds, they'll play R5 after R3.
D will not see R4 anywhere, they'll believe they have R4 and play W4
At this point, the line is broken, a bluff is revealed and no one has "passed" to suggest a layered move.

The risk is that C might clue D's W4 in some way instead of playing R3.

Jellby’s analysis is correct. The concept of bluff seat is a consequence of the principles on which the moves are predicated, not the other way around, and we shouldn’t let the shorthand - and that’s what bluff seat is - become the play.

However, on the risks of this clue… this is why responding to finesse-type clues should be regarded as a higher priority than giving new clues. It’s important without layered, and it’s more important again when playing with layered. That should be a safe clue, because everybody should be giving high enough priority to their responses.

There are other scenarios where this isn’t true. Be more careful when there is a player somewhere in the sequence of moves who isn’t obligated in any particular way.

[Blacktango]

I call such a move a "finesse + bluff" clue.

R5 is played as an answer to a R finesse, and W4 is bluffed immediately after.

[Tuut Tuut tis Luut]

I wonder what is more popular convention in BGA.

I've never really liked conventions. Partly cause I first learned to play when the term wasn't popular. And considered akin to cheating by most: Using agreed upon 'rules' instead of logical deduction.

Having been away for a few years it's usage seems to have shifted largely to shorthand for established concepts.
But I still think talking in terms of conventions is rarely productive. And this here is a great example.

D just needs to ask themself. Would B still (want to) give this clue if I didn't have r4 in my hand? Or did they have a better (or at least equivalent but less ambiguous) move available to them.

The answer here seems to be clearly yes. Two cards played and one more nearly playable 5 is now known with just one clue. That's a great opportunity and a waste to pass up.

That means you have no reason to suspect let alone assume there was another intention behind it than was already achieved.


As you concluded yourself, if D does have the 5, this logic would unfortunately prevent the even more efficient play the layered would give. But that's a very specific, very rare situation that likely only takes one clue to reclue.

But the alternative is missing out on the the simpler finesse bluff. Which is way more common (by such a large margin that it's barely worth comparing), as well as being more cumbersome to correct for.

By asking the wrong question and being concerned about the 'convention. You missed the actual question, what makes the most sense. Which you had already largely answered yourself. Conventions should follow that logical choice, not the other way around.