## Fafmark

We have a few critical, base convention **special agreements**. First, our common aggreements are:

- [Rank Trash Promise](/VdgZXm4nQXWSjI-Avj1mZQ#Rank-Trash-Promise)
- [Negative Fill-Ins](/VdgZXm4nQXWSjI-Avj1mZQ#Negative-Fill-Ins)

Next, the changes to base convention interpretation:

#### Elim Rider Deduction / Safe Action Principle (Shared?)

If a *possible referential discard clue* reveals a play for any reason, then it *can't* also give a discard signal, except through good touch or 0cs agreement. A *possible referential discard clue* is an unloaded clue with rank that touches at least 1 new card.

#### Information Over Truth

My understanding of our blindplay agreements is that we typically prefer greater freedom over clues that provide information to the hand rather than devote conventional space to convenient connectors. In other words, I think I would require a discussion of a particular circumstance in order to take extra action signals from a play clue (beyond those given by good touch), even when it may be a "true finesse".

The only circumstance I am confident would get more than one play (i.e., we have discussed) is also the most common: the unloaded finesse before prompt of a 2-away-from-playable. So with g1 on the board and two touched green cards in hand, cluing 4 to previously touched g4 would cause a blindplay and if the blindplay was g2, then the other touched green card would play as g3.

#### ~~Slot 1 Interpretation Preference~~ Quasiloaded Ptd

We play with quasiloaded ptd.

#### Loaded Fill-Ins with Delayed or LPC Component

If Fafrd gives a loaded fill-in through his own hand while occupied, then it calls for a blindplay. If Fafrd gives a loaded fill-in that also touches a previously untouched card, then it calls for a blindplay. If hallmark gives either of these clues, the only action signal is on the filled-in playable, and it means nothing more.

#### 8 clue discard clues touching slot 5

If a rank clue is split and touches slot 5 on 8 clues (not in a "first turn" scenario), it holds its meaning as a discard clue and not a stall. This is useful because slot 1 can be made to bomb, but other cards are harder to kill, so the discard clue should not simply mean stall and discard slot 1.

#### Zero Clue State > Discard Modulation

Zero clue state conventions supersede discard modulation. Choosing to discard rather than play into partner is a chop move and will require partner to stall for a turn if locked. This applies even at low pace (players should just be aware if it demonstrably interferes with optimal endgame strategy).

And finally, some **variant-specific agreements** are:

#### New Suit PTD Revokes

In Sudoku or Up or Down, a discard clue can revoke ptd instead of being a loaded play if the clue giver just played and committed to a new suit direction/starting point that they didn't know about. (I think we agreed to this? Not 100%)

#### Rainbow 1s Color Promise

Color to get rainbow 1s should match one of the 1s or be the greatest available lie.

#### Pink Discard Targeting

When the oldest unclued card, which would normally be touched for a pink discard clue, has permission to discard, the expected pink discard clue touches the next oldest unclued card instead.

#### 5 Truth Value is 2

The truth value of 5s in any variant where there exists at least one pinkish 5 and the 5 rank clue is unavailable (e.g., Muddy-Rainbow-Fives & Omni) is the rank 2.

#### No 1s Assumption: Only First turn vs first clue

In **pinkish** variants, 1 clues are not treated any differently than clues of any other color. The only exception is first turn/clue conventions:

Currently, game start conventions apply to the first clue to each player for pinkish and muddy-ones variants