Fixing focus or prompt

[AlexT2067]

4 scenarios :

1°
Player A : Clue giver
Player C : g1/r5/r1/w2/w5 (g1 already played)

To avoid bad touch, player A clue red and player B clue 5 to change focus to the 1.

2°
Player A : Clue giver
Player C : X/r3/r1/r2/g1 (g1 already played)

Player A clue red and player B clue 3 to change focus to the 1. Then next turn player B clue the r2 and for player C it wasn't necessary because the 3 clue prompt the red2.

3°
Already played : r1/r2
Player A : Clue giver
Player C : y3/y5/r3/r5/X (y3 not in danger)

Player A clue red, then player B clue 3.
We were missing clues, and so player A needs to discard (while holding a bunch of playable cards) to save the y5.

I ask why player B didn't change focus with a 5 clue, for him it was a direct prompt for r4

4°
Player A : Clue giver
Player B : w5/w2/b1/g2/g5 (Clued b1 so he get a play)
Player C : w1/X/X/X/X

Player A clue white to reverse finesse the w1. Player B play his b1, player C his w1.
Then player A clue 5 to change focus and in addition save the g5. Both other players get it as a prompt to the w5 and player C bombs his finesse spot


For me, in scenarios 1/2/3, because the player A clues and player B is forced to fix, it cannot be a prompt.

In scenario 4, I understand it could be a prompt because player A gives both clues to player B

[_mg6maciej_]

All or most of these examples are what's commonly called extraction among master level gamers.

When a convention system says that play clue (either color or rank) is to make the leftmost card among newly marked play, extraction helps to make the rightmost card play usually with a combination of a color clue, which more concisely describes the original intent, followed by a rank clue.

In addition you should not be worried as much about bad touch as you seem to be.

When extraction happens, it's important clue the receiver really thinks twice why two clues were given and what alternatives other players had.
Touching card with high number tells that high number should eventually be played from the clues given.

Here are your examples:

1. Do not avoid bad touch. A clues red and B follows that with 1 to point to 1r. Bad touched 1g is globally known trash and will be discarded.

Why? Because alternatives:
If both ones are useful cards, either player A should clue 1 initially or should leave that clue to player B.
If you touch 5 instead of 1 as fix, this tells one of the other players should hold all missing cards: 2r, 3r and 4r and it's a long finesse.

2. This is not a series of clues I'd call extraction. Next round you don't need to clue 2. Just like how 5 clue in the first example tells someone should play all 2 to 4, 3 clue tells someone has to play 2. Because red touched all 1, 2 and 3, 3 clue is enough to make all 3 red cards play.

3. I assume you've made a mistake here and C's hand is instead y3/y5/r5/r3/X. If you have not make a mistake, red clue is enough to make 3r play.

The correct way to extract 3r is A clueing red followed by B clueing 3.
Alternatives:
a) A clues 3, telling leftmost 3 is to be played, so B tries to find missing 1 and 2y.
b) B clues 5 instead of fixing with 3, telling this 5 should be played (or revealed unplayabled with a bluff), so after C plays 3r, A or someone else must attempt to find 4r.

4. There is no need to use extraction in this example.
If 2g is a useful card, A should simply clue 2 as reverse. B then plays his previously marked 1b and C plays leftmost, expecting 1w. Next round B can play leftmost marked 2, knowing it must be white.
If 2g is trash, this 2 clue can still be given and it works the same (requires fixing in the future; again, don't be too afraid of bad touch, 2 clue toching trash 2g might be the best option under some circumstances like 3w on player's C chop), but depending on the global game situation (how many clues are left, if 1b is marked with color or is unknown playabled marked with 1, etc.) it might be better to look at one of the alternatives:
a) A skips and clues white or 1 to C, then next round after both 1 blue and white are played, clues 5 to B to save 5g, waits for 2g to be discarded and in an undetermined future clues white or 2 to make 2w play
b) A discards and expects B to bluff 1w, then two or more rounds later repeats the same dance with B: clues 5 to save, waits for 2g discard and clues white or 2 to make 2w play.

I see from your history you are playing mostly simple 6 colors variant. A lot more interesting (and to many master level gamers confusing) extractions happen when you play with true rainbow and black cards.

Here is a few simpler ones (m is multi, marked by any of 5 colors)

5. C has X 2m 2w X X
1w is already played. A clues white to C, B fixes with 2, C plays rightmost 2w
But why? Because alternatives:
a) If A clued 2 instead (or a different color), this tells B that 2m is the focus of the clue and attempts to play leftmost as 1m.
b) C knows to play rightmost 2w among all 2w cards she can see in her hand, because if leftmost 2w was 2w, 2 extraction / fix clue would not be needed.

6. C has 4m 2y 2m 2w X
1w is already played. A clues white to C, B fixes with 2, C plays rightmost 2w
Even tho this is slightly more confusing, because these two clues don't tell exactly what player C should play, the popular convention is that C should play the rightmost card that both clues touched.

Which leads to an impossible extraction:
7. C has 4m 2y 2w 2m X
If A and B collaborated and tried to clue white + 2, C would bomb 2m as 2w. Why accept 6 as valid and 7 as not? It's arbitrary, but a good argument would be that 2m is the easiest card to bluff on (A clues red/green/blue, not touching rightmost card X, B plays leftmost) and after that simple white clue to make 2w play works.
You might think that white followed by 4 would work here, but if you accept that touching high number card is saying this card should be eventually playable, then C will think she has 4w, 2w, 3w and A will think she must have the missing 1m, 3m.

[AlexT2067]

3°:
X/r5/r3/y5/y3 (wrong order in the first post)

[_mg6maciej_]

3°:
X/r5/r3/y5/y3 (wrong order in the first post)

This does not require extraction either. 3 clue works to make leftmost newly marked card play if no 3 is discarded yet and B does not have 3 on chop (in which case 3 clue should be interpreted as double discard save over a play clue).
If some 3 is discarded and you want to keep 3y in C's hand, red + 3 is ok.
If you see 3y elsewhere (maybe fully marked with color and rank on yourself) and don't care about this copy, just wait for 3y to discard and then clue 3 to make 3r play.

[springiest]

1. B should assume they have r2/r3/r4 (in that order) on finesse position. A can first clue 5 (saving w5), which would allow B to clue red. but A chose to clue red first, so it must be a play clue on r5.
2. i dont think C can know that their slot4 is r2. for some reason (probably a trash 1 somewhere, or k1 in slot1), A's red clue needed a fix. therefore its not necessary that r2 is on the board at all.
3. this is a play clue on C's slot1. if any 3 other than r3 is playable (say g2 is already played), then C should note that their slot1 is g3. if not, this is a Self-Finesse (C has a hand like `k3 k5 [r3] [r]5 k4`) and C should blind-play slot2.
4. firstly, A should clue 2 instead of white - it picks up g2 in B's hand, now the team doesnt have to worry about losing the first g2 before g1 shows up. but having already clued white, i dont think anyone on the team is necessarily promised w3/w4 because without this fix B would bomb w5 as w2.