Color clue on chop

Forum rules
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/#!bugs
Post Reply
being_kinder
Posts: 2
Joined: 27 November 2022, 06:27

Color clue on chop

Post by being_kinder »

The situation: When I make a color clue on smbd's chop, people tend to play it. Or read as another playable combo: finesse, bluff, etc.
Per my understanding any clue on chop can be either play or save. If I can exclude save(for instance no critical cards of this color and I know it's not a five), only then it can be playable.

What's the reasoning of playing color clues no matter what?
Malo77
Posts: 43
Joined: 24 July 2022, 15:05

Re: Color clue on chop

Post by Malo77 »

This is the most common convention on BGA.

When ou give a color clue, this is a play clue.

So when you mark only the card on chop with a color, this is a play clue.
When you mark several cards with a color, you play the leftmost one.

Except in a bluff situation, where the colored card should be held for later.

There are some other conventions set where you can save with color, like H-group.

Do you use color save oustide BGA ?
User avatar
Jellby
Posts: 1343
Joined: 31 December 2013, 12:22

Re: Color clue on chop

Post by Jellby »

I would say that, from "first principles", number for save would be more common, if only because you can save several 5s with one clue, but saving several cards with a single color clue would be much more unlikely. Hence the "number on chop is a save" convention, which leaves color for play, because if you wanted to save you could use a number.
User avatar
Romain672
Posts: 1012
Joined: 05 April 2016, 13:53

Re: Color clue on chop

Post by Romain672 »

1) Colors for plays:
And when you give a play clue, by default you give a color because then the card is known. If stacks are r1, b2, y2, and you give a red clue, that should be r2. If now you give a 3 clue, that can be b3 or y3.
And since you can delay your plays easily (depending of mainly others chops), having the best information about your hand is better.

2) Clue 5 for 5-saves:
Now for saves, like Jelby said, 5 clues are commun and usually best when you save a 5. If you clue a color like red on r5, maybe the card is exactly r5 by context, but if it is not, you got the whole game (until r4 is played) to clue that 5 for free. That can be done by direct clues, negative clues or by 5 drawn by others players.

And from those two points, you can decide your rules.

My personaly opinion on it is that allowing saves with color or number are both really good.
Hgroup allow both of them (but force 5-saves as 5, and 2-saves as 2) let you directly in color save know it's a 3, a 4, or is playable, which is already pretty limited (and can be known with specific clues). That let you touch less trash, and sometimes give a direct identity of a saved card.
Bga allow only number for saves, that let you give play clue/finesses even in chop.

.

In recent but really advanced set of conventions (reactor, or referential sieve)(wwith clue focus on adjacent cards of clued cards), you often choose one of the way (color or number) to give play clue, and the other one, to ask for a specfic discard. Atm, it's done by saving all the cards to the left or the right of a clue.
being_kinder
Posts: 2
Joined: 27 November 2022, 06:27

Re: Color clue on chop

Post by being_kinder »

Thank you all for the replies.
The good idea was to check the conventions on BGA. I was learning the game by this tutorial https://hanabi.github.io/docs/beginner/.
But turns out BGA clearly states the convention for color and number clues.

Maybe my biggest resistance against it is because it is just a convention. Basic things like finesse and bluffs are built on a pure logic. So a clever enough player could understand(or reinvent) the move without smbd explain it to him. But rules like color only play and number only save are based on a convention, not strict logic. The game lacks purity in some way.

But after all there is a community and if it is happy about these rules, I'm not gonna change it.
User avatar
Wreckage
Posts: 287
Joined: 18 January 2017, 02:10

Re: Color clue on chop

Post by Wreckage »

There are a few players that play "logical leftism" convention style. Their arguments are much the same as yours. Basically they save with color or number, and possibly anywhere in a player's hand except left-most.

The logic is; if it was playable then you would have given the clue as soon as possible when the card was on the left. If you wait for the card to move to the second slot, then it must be a save. However they also take into account if there was a reason the card just became playable this turn, and they look at the card just discarded to consider if there is any reason you waited until this moment to mark this card.

I have only seen logical leftism in 2-player games. It is a very effective convention by top players. I think it would be very confusing in multi-player games though. I have almost no experience with this convention. Maybe it's very good for multi player as well, but I haven't seen it played that way.
User avatar
Romain672
Posts: 1012
Joined: 05 April 2016, 13:53

Re: Color clue on chop

Post by Romain672 »

being_kinder wrote: 09 June 2023, 21:27Maybe my biggest resistance against it is because it is just a convention. Basic things like finesse and bluffs are built on a pure logic. So a clever enough player could understand(or reinvent) the move without smbd explain it to him. But rules like color only play and number only save are based on a convention, not strict logic. The game lacks purity in some way.
It's off-topic, but we could talk a lot about that point.
Everything you assume will be a convention.

Even if you play with new players, if another player turn 1 give red: x-r-x-x to his next player, and that player discard, you learn that player usually discard on that specific case. It seem it's not a convention, since that answer was deduced from scratch.
If now you play a second game, and turn 1 your previous player clued you red: x-r-x-x to you his next player, you learned how to react to that clue, knowing the cards of others players. So this clue has some similarity with the previous situation. From that, you can either react to that clue the same way you did before, but then it became a convention. You are not deducing that action from logic, but you are just either copying what happen in the previous game, or playing what others seemed to expect you to do.
Or you can check other hands, assume those hands is the reason why you should assume something different than the previous game, and do your own move to it.
Or, if what happen in the previous game seemed bad, you can do something diffferent too.

But even if you do something else, you just move the problem. Next time this situation will happen, you will get two previous ways to react instead of one.
After a certain point (we are humans :p), in most moves, you should start following the same interpretation. And you got a convention.

.

Here is a real example: "3-bluffs" from hgroup (level 13). You are enough experienced to see that you lose most of your games because you lost early one useful non critical card, and search a way to save them more.
Furtheremore, finesses are 'logicals' (you clue red on r2, next player plays r1 from the clue. r2 plays after)
Then, bluffs are 'logicals' (you clue red on r2 and next player plays y1 from the clue. You then know you got r2).
Then, you learn about finesse+bluff / double half bluff (you clue red on r3, next player plays r1, second next player plays y1, r3 is known).
You then arrive on a situation where someone clue red on r3, next player plays his finesse position and it's y1. Then another player plays in between. By default with the previous rules, the r3 is assumed to be r2. So you need to assume/do something to correct that.
There is lots of ways:
- you can ignore the problem until r1 is played
- you can assume you got two r2/y2
- you can clue r2/y2 (if available)
- you can clue 2/3 to that player and make the clue do something else (if available)
- you can clue 2/3 to that player to fix that problem and assume the clue do nothing else (which look bad)
- you can play your finesse position to create a situation with two bluffs, which should resolve the problem
- you can assume you got r2 in finesse position and save it for later (implied bluff)
- you can do another interpretation

If you want to play 'logically', it usually means that you will assume one of the most obvious interpretations which are listed there.
But that reasonning will be worst for 4s and 5s.
When you give a red clue on a 4, it will require 3 diferents cards.
When you give a red clue on a 5, it will require 4 differents cards.

What is interresting about '3-bluffs', is that it's very powerful, but it's not listed here.
'3-bluffs' are pretty simple, they just assume that 3s (which don't connect, and red don't connect with 3s, even with self finesse) ask for only a single bluff.
That let you make '4s-double bluff' which ask for 2 bluffs. (5s are different, look at ejections (slot 2 play) level 16).

.

Bga has a very big community, changing any rule, even very simple is extremely difficult and take some years. Just see how difficult is it to make experts and masters plays their 1s from right to left (it's not the subject, but compare hands x-g1-g1-b1 and x-g1-b1-b1 with the possibility to wait someone to discard once). People are so used to play from left without thinking, than even a logical play from right will not be done.
On the other hand, hgroup got a small community to build their own conventions, and was open for discussion of changing any conventions. They were just taking in account the complexity-cost (ie not adding an hard convention to follow if it's too complicated) and historical cost (ie not changing a convention for something slightly better if that would cause too many fails in the short run)(there were some big changes, but for big upsides, like ejections).

.

But to try to conclude my long post, you can't play without conventions, you will just follow someone else logic. You can play with random players every time, but you will ask more questions about "What if that player make a mistake?", "What's the probability of X?", "What mean the clue X? Was there more than one interpretation"?
And it's very interresting too, but it's very different. Like Wreckage said, the most popular is 'logical leftism'. It start with the assumption that you give play clue as early as possible (unless you got a reason not to). And for me, it's a convention. It depend of how much you follow the rule, but again, since there can be difference between people, you got more ambiguity in conventions.
GeraldineMerida
Posts: 333
Joined: 15 December 2020, 07:27

Re: Color clue on chop

Post by GeraldineMerida »

I mostly play in a regular group of 2 masters and an expert, and we do NOT use this convention. Nor would I expect it to be used in games with random BGA players, unless the table host has specified the convention in the game set-up.

I don't like the rigidity and artificiality of 'When I say x it always means y'. My group tends to look at the context of clues, for example to determine whether it is a save or play clue.
Post Reply

Return to “Hanabi”