Flamboyants are not a good part of the game.

[KP1024]

It seems like everyone is playing with flamboyants these days. And I understand the appeal - they give more opportunities to win, and make up for the extreme difficulty of black powder. They can also compensate for an unlucky start and keep the hope of a perfect score alive longer.

However, in practice, they add more luck than they remove, and significant parts of the game end up revolving around them, without any additional strategy.
The most powerful ones are the ones that pull cards out of the deck. Not only can they bring lost cards back from the dead, they make turns more efficient and extend the length of the game. However, if they come up at the wrong time, they're almost useless. The instaplay might come up with no opportunity to play anything, and if the pull a card out of the discard gets drawn during the endgame, it has no benefit.
Then there are the immediate clue flamboyants, which are rarely useful. They're cool when they are useful, but because flamboyants are unpredictable, you probably won't plan your strategy around them.
The combination of these means that while using flamboyants does make the game a little easier, it doesn't make the game better.

Here I propose, against the official game rules, an alternative flamboyant set of rules:

When a stack is completed, you get a flamboyant, and instead of getting a random one, you get to choose which one you get. But once you take it, the flamboyant is discarded and cannot be taken again.
The flamboyants are as follows:
1. Get an extra clue token.
2. Take away one misfire and get an extra clue token.
3. Choose a player to take a bonus turn, on which the player may only play a card from their hand. You may choose yourself. If the player misfires, you do not gain a misfire token.
4. Give a clue immediately to another player. It can be a number or color clue. This does not cost a clue token.
5. Choose a player and take any number of cards out of their hand and shuffle them into the deck, without revealing them to the player. You may choose yourself. The player then redraws cards to a full hand size.
For games with multicolor and/or black powder, flamboyant 1 is duplicated additional time(s).

The idea is, by choosing which benefit you get, you can make more intelligent decisions about how to use them, but by making them slightly less powerful, it reduces the extreme luck element that flamboyants currently have. 3 can be used near the end when one player has a ton of cards that need to be played, and not enough turns to do so. 4 still allows for some cool clue opportunities, but by choosing when you get it, opens up those possibilities more. 5 allows you to rebalance cards in the hand and potentially free up a locked hand. None of these allow cards to be pulled back from the dead, but they do give more opportunities to prevent lost cards in the first place.

[Blacktango]

I enjoy playing with flamboyants on BGA, but not necessarily to increase my chances of winning.
I like it because it mitigates mistakes caused by inattention.

When you play online with strangers, everyone has their own life and circumstances behind the screen.
Maybe someone is waiting for a pizza delivery or has a sleeping baby who "should't" wake up just yet. Or maybe your neighbour will suddenly start drilling into their wall and make a lot of noise.

We can't plan or even be aware of all these events like when we are all around the same table in the same place.

Sometimes an inattention mistake has no consequence (it might even make us play a random playable card! ), but sometimes it can ruin the game, which is quite annoying when it's your last game of the night, for example.
IMO, the Flamboyants help us to overcome these situations.

[Wreckage]

I have always thought flamboyants don't really fit with this game. It felt like they added dice rolling to the game. There are a number of better ways they could have added a challenge imo, without adding this total luck factor.

[Travis Hall]

Here I propose, against the official game rules,

Separate from any other comments I may make, I want to point out that as soon as you say this, you are talking about something that will not and can not ever be implemented here on BGA. BGA has a fairly strict policy to only implement rules variants that are approved by the original publisher of a game.

You can do what you like with your physical set at home, but you won’t be able to here.

[Travis Hall]

However, in practice, they add more luck than they remove, and significant parts of the game end up revolving around them, without any additional strategy.

The assertion that Flamboyants provide no additional strategy is simply not true. The random effects that they have do not prevent us from using strategies that help us make the most of Flamboyant effects, especially by mitigating the effects of that randomness.

The big strategic question that Flamboyants introduce is, when should we play our 5s (and 1k), and when should be instead hold them for later use? (Your suggestion of effects that are chosen instead of random almost completely negates this consideration, because the players can simply choose not to use the effect that they wish to save until later.)

Much of the time, good players will hold back their flamboyants if and only if they have a good non-discard move to make (until late in the game when it becomes important to reduce the plays held in your hand). We don’t hold them if that would require discarding, because it makes our discards unpredictable and prevents our partners from effectively managing our hand. Holding the flamboyants allows us to use the retrievals to make up for errors or extreme situations that force the loss of a card, or even to prevent getting bottom-decked after losing a non-unique non-trash card. If we do lose the last copy of a card, though, things become even more complicated, as we work to arrange the highest probability of a successful retrieval. I will sometimes change how I manage another player’s hand, giving them options to hold a flamboyant for longer, while we wait for the cards needed to allow the play-from-discard to retrieve the lost card.

By contrast, bad players will often try to hold flamboyants regardless of the situation. I have seen this snowball, with panic over the loss of a card and trying to hold flamboyants to effect a retrieval later without regard for the immediate risks leading to the loss of several more cards.

We can clearly see from this that flamboyants as written do provide additional strategic options. They also make the game a little easier for highly-skilled players, and when your group is highly skilled, you want both strategic options to allow you to flex that skill and high difficulty so that the game remains challenging.

[ozymandias_]

Thanks for your post! On the substance, you're clearly right: flamboyants are counter to the philosophy of hanabi (only randomness is shuffle), and have contributed to dumbing down of bga hanabi

@tango agree the game remains interesting by avoiding lost games, but solve for the equilibrium! "1/2 mistakes a game np" -> lower quality throughout. If flams coexisted I'd agree with you, but they've eaten the website

Personally avoid the flam tables nowadays. Would be nice to see more w/o flams!

[Travis Hall]

"1/2 mistakes a game np" -> lower quality throughout.

If your team’s attitude is that they can afford to make 1 or 2 mistakes each game because Flamboyants make up for it, you’re not playing with a good team to begin with. Flamboyants can retrieve up to two lost cards in a game, but they don’t do so perfectly reliably. Also, it usually isn’t possible to get through the game without getting bottom-decked becoming a risk, so I would much rather retain all unique cards and use the flamboyants to mitigate the risk of being bottom-decked.

I play a lot with flamboyants because I host a lot of games for experts, to help them train up to become masters. I know that many of them will make errors (they don’t really need to be playing in the games I set up if they won’t) and flamboyants give me leverage to make up for the errors with good play. I’d be more interested in playing without flamboyants when playing with masters only… although even then, there are enough masters whom I expect to make errors that 5-player masters games still becoming unnecessarily risky.

If I was playing with clones of myself, I think I’d probably play Avalanche, Black Powder, no flamboyants a fair bit. But I’m not.

[KP1024]

Here I propose, against the official game rules,

Separate from any other comments I may make, I want to point out that as soon as you say this, you are talking about something that will not and can not ever be implemented here on BGA. BGA has a fairly strict policy to only implement rules variants that are approved by the original publisher of a game.

You can do what you like with your physical set at home, but you won’t be able to here.

I do know that, I'm just demonstrating that I'm not opposed to the idea of flamboyants completely, and that a set of flamboyant rules can exist that are better than what the official rules have.

[_mg6maciej_]

If you are like me and prefer the game to be more challenging (which no doubt a flamboyantless game is), try these two variations:

1) Clue Starved: only every second 5 play or card discard generates a clue token back; this is a lot harder than your standard hanabi game requiring on average a clue to make ~1.5 plays instead of 1
2) Same Color/Number: instead of telling others what color or number pointed at cards are, just tell them if the cards you are pointing at have the same color or the same number; this seems quite hard at first, but is only slightly harder than normal hanabi