first clue in 2 player

[Stroom]

Ah, the old 2s before 1s argument. The only "policy" I stick to is giving 2 opportunities for my partner to deal with my hand before I discard.

1s first is usually best if they have multiple 1s.
But if their hand is 4 3 2 5 1 2, Then you would have to clue 2s or 5s first or you wont be able to save the 2 on chop.

To me, 5 clue first means that the rest of my hand is useless or the player especially wants me to discard the chop before giving the next clue.
This isn't the case with 2 as 2s are very important to save.
And 1 first is basically shifting the same choice to the other player - Either give me a 5, 2 or 1 or play your own card to tell me if I can start discarding or not.

[Qualitaetsgarant]

Interesting discussion for sure.

I strongly disagree with Stroom letting 3s go. Why would I do that? If the 3 i am discarding is the bottom card of the deck, my max. score i can reach now is 28. Definetly not what i want to achieve - playing for 30 should be your main goal of the game if you play advanced Hanabi.

The main problem at BGA is that there are only few conventions that are universally agreed by the community. For getting deeper into Hanabi i suggest finding a main partner to practice with instead of playing with random partners once in a while. You wont be able to get insight into Hanabi if your partner doesn't understand more difficult situations.

Example: The listed hand 4 3 2 5 1 2 .

I would start with 1 clue here and continue with 3. Why would i give an early 3 clue save here? Early save = all the cards closer to the chop must be important. So my partner takes 3 silent saves and continues discarding the 4. I am able to save the 2s, 5s and the 3 with just 1 clue. This method is far superior than the listed methods so far.

[Stroom]

Interesting discussion for sure.

I strongly disagree with Stroom letting 3s go. Why would I do that? If the 3 i am discarding is the bottom card of the deck, my max. score i can reach now is 28. Definetly not what i want to achieve - playing for 30 should be your main goal of the game if you play advanced Hanabi.

The main problem at BGA is that there are only few conventions that are universally agreed by the community. For getting deeper into Hanabi i suggest finding a main partner to practice with instead of playing with random partners once in a while. You wont be able to get insight into Hanabi if your partner doesn't understand more difficult situations.

Example: The listed hand 4 3 2 5 1 2 .

I would start with 1 clue here and continue with 3. Why would i give an early 3 clue save here? Early save = all the cards closer to the chop must be important. So my partner takes 3 silent saves and continues discarding the 4. I am able to save the 2s, 5s and the 3 with just 1 clue. This method is far superior than the listed methods so far.

1. I have written notes about playing with you before. You did not try to prevent double discard situations. This would probably lose points more often.

2. What is the probability of the other same card being in the end of the deck? 1/59 = 1.69%. What is the probability of a players hand being filled (almost) fully with marked cards? Almost in every game. If you propose saving every 3, you have no room left in your hand for 5s, 2s or even 4s that have a duplicate already discarded. You are constantly wasting clues to make non-critical cards discardable. And in the end you will lose even more points than the potential 2 point loss from dicarding a 3 early.

3. If you add many 3s into the mix that are not necessarily important or the other player might even mark them in YOUR hand as well... It's just a lot of confusion. Especially if you start marking the 3s in early game.

Your example: What you are proposing is that both players waste 3-4 clues to play around your play. You want to immediately tell the other player that only the leftmost card can be discarded. I do not really agree with this as it does not say ANYTHING about the cards right to the 3. And as they are 2s which can be discarded as needed, the other player doesn't know so he might save them even if you play them from your hand. What do you think happens if the newest card is a 5 or must be saved otherwise? Random discards from the right. You should rarely give a clue that forces the player to shift chop. If you shift the chop 3 spaces in the beginning... What you would do it create critical situations straight after the initial game.

Also: Your logic here is flawed. You use 1 clue and I will play it. But if you give a 3 clue next, I might think that my chop card is a duplicate that might need to be discarded in order for you to clue me the next clue Or maybe I have a multi 4 or something on chop and a playable 2 to the right of the 3 that is the same color as the 3 so you want me to discard the chop and then give a clue that does not clog my hand too much. Stalling to save 3s straight out of the gate makes very little sense. Make the game speed up with better clues... The best play here would be 2-1-5. And if the 3 is immediately playable, use 3. Otherwise, after using 3 clues, maybe 1-2 clues usually also used by the partner, you have to start discarding anyway. Discarding 3s happens. Discarding 4s happen. Yes, you might want to prevent that but it has to be case-by-case and you can't prevent it every time. And as I said, one cards duplicate being in the end of the deck is less than 2%. Wasting clues causes bad scores more often.

More statistics:
As 1-2-5 clues are marked in the beginning, it is quite high probability that the chop cards left are 3s or 4s (5% that the next card is the same as you discard). If you mark 3s too, then it might leave only 4s (11% that the 4s are the same). This greatly increases the chance of being bogged down with preventing double discard situations. And double discard situations happen much more often than the same 3/4 being in the end of the deck.

Overall you want to play to 30 but you have to do it in an educated way. Sometimes you have to sacrifice something for a chance at maximum score.

[Qualitaetsgarant]

Sacrificing 3s early is not an educated, but a suboptimal, limited way to reach maybe 30 or maybe not. It is conceding to the fact that the team is not able to clue efficiently enough to save 3s.

You should not think that you have a duplicate on chop if i give an early 3 save. If your chop is a duplicate, i give the 3 safe clue AFTER the discard of the duplicate.

You see, Hanabi is a lot about timing. Giving protection clues too early because you cant wait for it might lead to misunderstandings.

As you refered to notes about each other: I noted about you that you lack the understanding of timing and therefore I might have double discarded as I stopped caring about the game which would be the last game together anyways.

[Stroom]

Sacrificing 3s early is not an educated, but a suboptimal, limited way to reach maybe 30 or maybe not. It is conceding to the fact that the team is not able to clue efficiently enough to save 3s.

You should not think that you have a duplicate on chop if i give an early 3 save. If your chop is a duplicate, i give the 3 safe clue AFTER the discard of the duplicate.

You see, Hanabi is a lot about timing. Giving protection clues too early because you cant wait for it might lead to misunderstandings.

As you refered to notes about each other: I noted about you that you lack the understanding of timing and therefore I might have double discarded as I stopped caring about the game which would be the last game together anyways.

I guess you just don't get it. Maybe not enough experience with 2-player or something. Let me explain again - how many cards can you hold between 2 players? 10/12. If you save 5s and 2s, how much space does it take? 10-12. If you start adding 3s that do not match any previously marked 1/2 colors, they just clog up the hand. The timing here is quite important but not how you explain it. Better to let both players play 1s and 2s. Maybe you or the partner has the cards needed to play everything up to the 3 already. THEN give the 3 clue for playing. Much more efficient than marking in the beginning.

Every situation is different and we are probably arguing over different things. I am saying that a 3 save should be done when you are confident that you do not have one in your hand and that the hands will not be clogged up. The very beginning of the game is NOT the time to start saving 3s that are unrelated to any 1s and 2s in your hands.

I have played about 20k games and I can see from a mile away when someone plays inefficiently. That also changes my playstyle sometimes. Sometimes the other players logic is just so bonkers that I just play based on the usual logic and that sometimes isn't what the other player wanted.

What you propose, causes problems down the line - unmarked cards that might be 5s, might be 2s - if the rest of my hand is filled, I don't know what is safe to discard. Maybe some of these unmarked cards are already played at some point but I don't know as I don't have ANY clues on them. This will need 2+ clues to fix. Meanwhile I would also need to give stalling clues so it just doubles for every turn you use a clue. There are just much better ways to play. The probability <2% of the same 3 being in the end is much better risk than playing around a clogged up hand. Just look at the maths. Or play more and see how often you have to deal with low amount of clues or filled hand. Better playstyles will emerge more by experience than theorizing. And you start seeing the probability and statistics will also support this playstyle.

Hanabi 2p is more about managing your and your opponents hand in order to prevent clogged up hands and playing as efficiently as possible at the same time. Timing is a part of it, sure. But there are also exceptions why you stall at times... I usually try to conserve clues when we are at 1 clue so I discard and let the other player discard or clue my cards for a while - even if I could have used a play clue. Being at 0 clues often means neither player can not clue what is critical to the game so you end up losing more than you win by giving a play clue for some card. Sacrificing a playable card might be better in some cases. Double discard situations very often happen exactly because one player used the last clue, then discarded because there was nothing to do and then the partner also discards as he has no info anyway.

The way you double discarded was basic - I discarded a 4 and you discarded right after me - oh, the same 4. Basic double discard situation that is usually preventable.

[Qualitaetsgarant]

Every situation is different and we are probably arguing over different things. I am saying that a 3 save should be done when you are confident that you do not have one in your hand and that the hands will not be clogged up. The very beginning of the game is NOT the time to start saving 3s that are unrelated to any 1s and 2s in your hands.

In the situation given you have plenty of time. The player will probably give you a playing clue aswell, then play his 1, then chop a 4 - until then maybe one of his 2 will be playable. I can reclue the playable 2 to create another silent save if the card he is drawing is worth keeping aswell.

If the situation gets very dense and the 3 cant be kept, i can reclue the 3 to indicate the discard of the 3 and keep the rest of the hand. Giving up the 3 is the last resort, not the thing to start with.

On a side note: Never refer to how many games you played when trying to prove an argument. It is more the opposite: You played so often you got used to your mistakes without realizing them.

[Qualitaetsgarant]

The probability <2% of the same 3 being in the end is much better risk than playing around a clogged up hand. Just look at the maths. Or play more and see how often you have to deal with low amount of clues or filled hand. Better playstyles will emerge more by experience than theorizing. And you start seeing the probability and statistics will also support this playstyle.

Quite obvious that you are stuck with low clues and a filled hand because you chose to clue inefficiently earlier.

[Stroom]

In the situation given you have plenty of time. The player will probably give you a playing clue aswell, then play his 1, then chop a 4 - until then maybe one of his 2 will be playable. I can reclue the playable 2 to create another silent save if the card he is drawing is worth keeping aswell.

You are playing on probabilities here again. Hoping that the other player clues you. There are plenty of situation where the player might NOT want to clue you right away. And your confusing clues are quite opposite to any kind of normal logic and could mean anything from discard to save cards up to that card etc.

Also, you should see that you are setting yourself up to later give a 2 clue that actually means nothing to the other player. That is inefficient. Can't know what to play or to discard... And the 5 is there too. So it's overall all kinds of dangerous. Your logic does not really match with anything out of the ordinary.

If the situation gets very dense and the 3 cant be kept, i can reclue the 3 to indicate the discard of the 3 and keep the rest of the hand. Giving up the 3 is the last resort, not the thing to start with.

Usually a number clue that marks nothing but the same card that was marked before means that you should play it. Not discard.

Overall, I already showed that you are playing about something that would happen in 1.69% of cases. At the same time all other things happen much more often than that but you neglect them. It's just statistically better. And I have gather this stats from a high amount of plays. I have changed the playstyle quite a lot and tried things you mention. And they just will not work as well. Playing by math and probabilities is better.

[Stroom]

Quite obvious that you are stuck with low clues and a filled hand because you chose to clue inefficiently earlier.

LOL. I am usually the one who is more efficient with the clues. I'd say I can handle the hands pretty well if the opponent doesn't play too weirdly. Running out of clues happens usually when the deck just throws playable cards in a bad order and you are forced to clue one by one. If you fill someone's hand with useless clues too early, it will cause you to waste more clues later.

[Qualitaetsgarant]

Quite obvious that you are stuck with low clues and a filled hand because you chose to clue inefficiently earlier.

LOL. I am usually the one who is more efficient with the clues. I'd say I can handle the hands pretty well if the opponent doesn't play too weirdly.

What you are describing is the main issue. You can only play as efficient as your partner will understand you. As you described i could not use the early save with you to create silent saves because you won't follow this clue and discard as if nothing happened. I have to adapt and in this process i am losing efficiency.

The key in creating silents is maximizing efficiency to be able to survive a locked hand later. By clueing less efficient as described by you i surely have to let 3s go because i cannot afford having a locked hand later.