• Basics - level 3

    Version française ici.

    Previous: level 2

    #14 Chop clues

    When you receive a clue that includes your chop card, be very careful about it. Had the clue not been given, you would have discarded your chop card soon, so it is very likely that the clue is only telling you "save it".

    Some 4s are playable and you get a 4 clue on chop and the discard contains unplayed, unplayable 4s. You would take a big chance playing your clued 4. If you're wrong, your team would lose 1 life and 2 points.

    You can give colour chop clues if colour would be clearer than value.

    The discard contains Y4 B4 G4. Your LH1 has had Y4 in hand for a long time and Y4 is now on chop. 'Yellow' will tell them more than '4'. The latter clue could mean Y4, B4 or G4, while 'yellow' would disambiguate this into Y4.

    #15 Save 2s

    This one took me a good hundred games to figure out, so you’re welcome.

    If a 2 is the last (or nearly last) card of the deck and its twin gets discarded during the game, you are pretty much doomed.

    This is why a good thing to do is to save 2s even when the 1 is not in play. So when you receive a chop clue that involves one or more 2s, you should suppose it is just a precautionary save, not a play clue.

    Likewise, saving 3s (especially when the matching 1 is in play) can be a pretty good idea.

    Saving 4s can also be a thing when the target player’s hand is not too tricky to manage.

    #16 Colour or value?

    Your clues should be as informative as possible. Most of the time, colour clues are more efficient than value clues. The reason is that, except for save clues (hitting chop), single-card colour clues tell the target player both the value and the colour. Therefore, they can react accordingly. Given an equal choice, choose colour.

    3-player game, turn 1.
    Player B has 4 1 1 4 5. Player C has 1 2 1 1 3.

    If player A clues 1 on B, B would only know these 1s are playable, but not which 1s they are.
    If player A clues 'red' on R1, player B knows he has R1. Knowing this, he can give player C a pure anticipated clue on R2 or on the three 1s (see Level 4, anticipated clues).

    Value clues should be used only when a colour clue wouldn't be pure enough or in the very rare cases where they give more information than colour.

    #17 You are in charge of LH1

    Want to give a clue to the player 3-4 seats from you? Ok, but you know, all the players in-between could do it, don't panic. In general, you are responsible for and should give clues to LH1, i.e. the first player to your Left Hand (playing in clockwise order) because you are the last person in the turn who can give them clues. If you give a clue to LH2 or LH3 or LH4, you must have a good reason for it.

    Good reasons:
    - Tell a player in-between they can discard safely. If LH1 has nothing to play and has a safe discard and you think they will give a clue on their turn, you can give this clue for them. This way, you tell LH1 "look, I gave this clue for you, you have nothing good to do anymore, so just discard and trust me".
    - Every player in-between has something to do – they have a card to play or can discard safely their chop card (see first point).
    - You need a follow-up clue. Say LH2 has two vital chop cards, which need two clues to be saved. You need to clue one and LH1 will clue the other one.

    #18 Forced play

    Some call this the semi-finesse. It's about forcing a player to play a partially known card from their hand to prevent a strike.

    Bob has partial information about a card (e.g. a 2). Alice gives a clear play clue (e.g. red) to Cheryl who plays after him, about a card (R3) that can't be played unless another card fills the gap. Bob thinks: "We're up to R1, Cheryl is going to play R3 and we'll get a strike. What the...?! Oh wait, I have a 2 here. What if Alice is forcing me to risk-play that 2 that might fill the gap?".

    There you go.

    #19 Clue instead of playing

    As we saw previously, it is important to get cards played as soon as they can be clued in a clear manner. So if you know you have a card to play and you see that LH1 should be given a play or save clue, give the clue first (unless the clue might be redundant, see level 2 – “No duplicates”), you can play your card on the next turn.

    #20 End-game timing

    Once a player draws the last card of the deck, every player gets to play only one more turn. So the last man drawing will get to play twice. So if there is one card left in the deck and Laura has two cards to play and the others have only one or none, pace by spending clues until it is turn to play so that she plays at least twice.

    There are more complex situations than this, but the general rule is – in game end, don't rush to draw the last card. Count turns and anticipate on every player's moves to make sure the most cards will be played.

     

    Next: level 4.

    « Basics - level 2No conventional leftism »

  • Commentaires

    1
    fr
    Mardi 26 Mai 2015 à 16:31

    that's pure gold!

    Interesting consideration about colour and value clue... especially because it changes when playing multi variant.

    2
    berijeux
    Mardi 26 Mai 2015 à 16:50

    All my articles, for the moment, are regard the base game. I will certainly make some articles about the rainbow variant, which changes a lot of things, especially the interest of value clues ;)

    3
    el payo
    Vendredi 13 Mai 2016 à 14:21

    What you tell us in the "forced play / semi-finesse" chapter is really brillant !

    I've never thought about it, I love it. Not sure my friends would understand this kind of clue (cause sadly I don't play Hanabi as much I'd like to) but I'm pretty sure to use it in my next 3/4 players games. 

    What pleasure it should be to play with players that master all these tricks !

    4
    Alex
    Jeudi 28 Janvier 2021 à 14:08

    And... congratulations, you've invented a set of conventions, despite of what you were saying at http://hanabilogic.eklablog.com/conventions-pervert-minds-a117598652

      • Jeudi 28 Janvier 2021 à 14:16

        Hi Alex, thanks for your comment.

        Actually I’ve discovered all these tricks without anyone teaching them to me, just through logic. No agreeing about them beforehand.

        And I’ve seen other players reach the same conclusions without being taught.

        Another way to see it is that not using these tricks will get you lower scores. So it is totally logical to use them.

    5
    Alex
    Lundi 1er Février 2021 à 03:05

    Of course, I agree with this.

    And it's fine to discover tricks on your own. In my group I also try not to agree beforehand what hint means what exactly, but instead provide some logical explanation if I do anything not immediately obvious.

    I just mean that such set of rules you listed here, can be considered a convention. Well, it is again a question of definitions. And it's possible for someone to read this page, get the list of rules from it, and start applying them as rules without fully understanding their logic. Probably in that case this list will become more conventional by your definition of "convention"

    6
    Lundi 1er Février 2021 à 22:47

    If someone starts applying these tricks without fully understanding them, bad for them. They should read the header in "level 1" again ;-)

    Rather than "such set of rules can be considered a convention", I could agree with "such set of tricks can be considered a playstyle", or "such set of tricks are the result of a particular ‘Hanabi itinerary’", although I’m pretty sure a set of four NASA engineers – which I believe to be highly logical people – would reach the same conclusions.

    And yes it is again a matter of definitions :)

      • Romain672
        Vendredi 30 Septembre 2022 à 01:30

        And this is where I will disagree :D

        Two examples: save of 4s, and clueing asap (as soon as possible).
        This is the kind of moves which could totally be interpreted differently.
        "Why he is saving a non valuable 4? There must be something else going on.". The problem of saves is that it's a cursor you can move all the way to always assume 4s are a save, to the other way to always assume (non critical) 4s can't be saved. And knowing where to put that cursor is really hard and can depend of people. You decided to put it low, but not at the minimum, which has a cost.
        And if you instead decide that saving 4s is good in rare cases, you could instead decide to never do those which will make you be far better in all others cases.

        Then at one point, you can think of discard newest, which look from bots to be better than discard the chop. So what's is your goal? Is it to play the more instinctive possible, or to win games (define win games :D)?

        And I think it's the main reason of why you run into a wall if you try to play without conventions.

        You could imagine a tree: https://www.zupimages.net/up/22/39/swod.jpg , where you get all the differents paths you could decide for your conventions. I put a commun base but I don't think it's commun (hello hat guessing), and at many steps it just depend of feeling, or about winrate which can be immensively hard to analyse to know which of two decision is better (like savings 3s and 4s).
        As a last example, you can play with probability. So deciding for a certain probability for each hand you can have, and act in consequence. But the complexity is really hard, but can lead for you to find exceptions to every rule. Exception that maybe not everyone will agree on since it's so hard.

        Thanks for reading :)

         

    Suivre le flux RSS des commentaires


    Ajouter un commentaire

    Nom / Pseudo :

    E-mail (facultatif) :

    Site Web (facultatif) :

    Commentaire :