Delete the Weak Prompt about hanabi.github.io

[alercah]

The levels are basically meaningless at the moment. It might be nice to reorganize the doc to allow them to be something like "let's play level 3 only", but that would require it to be fairly straightforward to remember which level conventions are. At the moment, there are too many inconsistencies. A few examples: 5CE is "rarely used" based on being in level 5, and layered finesses are level 2 but layered GDs are level 4 for some reason. That should probably be a separate conversation, though.

[sjdrodge]

I vote against this. Is it worth trying to avoid weak prompt situations? Yes. Do they sometimes come up anyway? Also yes. Is it true that they "never work?" Definitely not. I have successfully navigated weak prompt scenarios on both sides of the clue before.

[Zamiell]

Q: What is the benefit of the Weak Prompt?
A: It allows us to easily give save cards that are currently playable and about to be discarded. Without Weak Prompts, the team would have to sometimes let playable cards die. Letting playable cards die is unfortunate, but not necessarily a big problem, unless the other copy happens to be on the bottom of the deck. Thus, the cash value of Weak Prompts is it prevents some percentage of bottom-deck losses.

Q: Every convention has a benefit and a cost. What is the cost of the Weak Prompt?
A: The cost is three-fold:

1. Loss of tempo

• With Weak Prompts turned on, cards will not necessarily play right away from a Prompt.
• Not only does the team lose tempo in situations where a real Weak Prompt has occurred, but they also lose tempo if a perceived Weak Prompt has occurred.

2. High cognitive load

• Since any Prompt has the potential to be a Weak Prompt, players have to always keep this convention in the forefront of their minds. We refer to this as a Headspace Convention. (In fact, the Weak Prompt is the thing that I chose to type up as the primary example of a Headspace Convention when I first invented these concepts back in 2018.)

3. High potential for mistakes

• Since Weak Prompts are so rare (more on that later), it is common for even experienced players to forget that they exist, assuming a normal Prompt and bombing a clued card.
• Separately, another thing to consider is when a normal Prompt happens that looks like a Weak Prompt - this has a huge potential for confusion.
  • For example, imagine that Bob plays red 3 into Cathy's red 4, and then Cathy discards. From Cathy's perspective, she is entertaining a Weak Prompt.
  • However, from other players on the team who don't see that intricacies of the situation, Cathy's discard looks like a Scream Discard. Now, the entire team becomes desynchronized and the cascading failure can lead to a loss.
• Yet another failure mode of the convention is when players do not account for the "last resort" component of a Weak Prompt:
  • When players are Occupied, it transforms a Weak Prompt into a normal Prompt.
  • Similarly, when the clue receiver is Loaded, it transforms a Weak Prompt into a normal Prompt.
  • Both of these contingencies are easy to miss. For example, after getting her chop clued and Prompting Bob to play a red 3, Cathy (luckily) remembers about the rare convention called the Weak Prompt. She mindfully does not play any of her cards and discards instead. But Cathy missed the fact that the clue was not a last resort, so it was a normal Prompt after all.

Q: How rare are Weak Prompts? In what percentage of games does the "payout" from the cost-investment occur?
A: Weak Prompts are rare - they only happen one in every 100-200 games.

Q: That is quite rare. Overall, is the Weak Prompt worth the cost?
A: Headspace Conventions that also happen to be rare are a fairly big red flag. We don't want to introduce a high level of cognitive load unless the convention has a lot of bang for the buck. In the rare cases where Weak Prompts provide value to the group, do they provide enough value in order to justify the complexity cost for all the other times? It is unclear, but the more and more games that I play, the answer seems to be trending further towards no. Even in the rare situation where a Weak Prompt can finally shine, the success rate of it working is extremely low (even with experienced players), because players forget that the move even exists.

Q: Shouldn't players just get good and remember that it exists then? Players making mistakes should not be a reason for deleting a convention.
A: This question literally gave me cancer.

Q: Why was the Weak Prompt invented?
A: When Hyphen-ated invented the Weak Prompt, the group was exclusively playing No Variant and Rainbow. As you know, these variants are fairly easy, and our win rate was extremely high, somewhere around 90%. The goal of the group was to push the win rate as high as humanly possible. By incorporating Weak Prompts, it helped us increase the win rate a few percentage points by preventing some bottom-deck losses.

Q: How is the situation now different than when the Weak Prompt was invented?
A: In a few different ways:

1. The variants

• Nowadays, we play a slew of much more difficult variants than just No Variant and Rainbow. Subsequently, the goal of the convention framework has also shifted. We no longer need to optimize the convention framework for a few-more-percentage-points of win-rate in trivial variants.

2. The framework complexity

• Nowadays, we have about five times the total amount of moves that we did back then. Weak Prompts are not so bad if all you play with is basic Finesses and Bluffs. In our current environment with Priority Finesses, Scream Discards, and 5 Color Ejections, it is a whole different story.

3. The skill level of the group

• Nowadays, the group hosts a wide range of skill levels. We have plenty of fresh beginners and most of the players are intermediate. Expert play is uncommon / rare.
• The framework should adjust accordingly. If we are on the fence about deleting a convention or simplifying a way of doing things, then we need to tend towards the side of simplifying. The exception is when a convention is common and provides a lot of value. For example, 5 Stalls have been proposed as something to consider deleting, but 1) they happen in almost every single game, 2) they get a huge amount of observable value. Notice that these two points do not apply to the Weak Prompt.

[Zamiell]

Q: What does play look like with Weak Prompts turned off / deleted? Will this leave us with no options to get playable cards in certain situations?
A: No, it will not leave us with no options. We have several things:

1. If the card is a playable 4, players can simply let it die. This is something that we do already in hard variants (since Weak Prompts are already turned off in hard variants). 4's suck and it is not worth bending over backwards for this shit.
2. If it can be seen that the red 3 is trapped behind an unplayable card and the red 4 will be discarded soon, players can plan ahead and give a tempo clue to the red 3 early.
3. Players can utilize one of our many Chop Moves to save the playable card indirectly. Note that the total amount of Chop Moves that exist in the framework nowadays is much more numerous than when the Weak Prompt was initially invented.
4. In some situations, other players on the team can help with a post-Prompt fix. For example:

• Alice uses the last clue to clue Bob's red 4 on chop.
• Bob discards.
• Cathy sees that this will Wrong Prompt Donald, so Cathy gives a red Fix Clue.

5. Notice that in some other situations, even Bob can do a post-Prompt fix. For example:

• Alice uses the 2nd to last clue to clue Bob's red 4 on chop.
• Bob knows that he has the red 4. (It is either explicitly a true red card or he knows from context, it does not matter.)
• Bob sees that this will Wrong Prompt Donald, so Bob gives a red Fix Clue.

[sjdrodge]

You've convinced me. However, as a general point, if the only concerns were about cognitive load and group skill level, a strong alternative solution would be to simply recategorize the convention to a higher "level" in the docs.

The argument(s) that this rarely-used convention does not offer significant enough benefit in the most commonly played variants these days is convincing.

[Zamiell]

The levels are basically meaningless at the moment.

Sounds like a hyperbole to me. Just because you think some particular sections are mismatched doesn't mean that all of the levels are "meaningless". I've painstakingly separated each section in what I believe to be the correct level. If you think that a section needs to be moved, then you can submit an issue.

"let's play level 3 only"

This is something we already do every time we play with a beginner, agreeing to only play with conventions that are in the beginner doc, or agreeing to only play with level 1 + level 2 conventions, etc.

[Zamiell]

/accept

[conventions-bot]

• Some time has passed since this issue was opened and the group appears to have reached a consensus.
• ✔️ This change will be integrated into the official reference document.
• This issue will now be closed; feel free to re-open it if you wish to continue the discussion.

(For more information on how consensus is determined, please read the Convention Changes document.)