Mechanic Help Needed

Now that's a..., not a good idea.
How?

Red, picks Red/Yellow
Yellow, picks Blue/Green
Blue, picks Yellow/Blue
Green, picks Green/Red
results in..
Red choses Yellow
Yellow choses Blue
Blue choses Yellow
Green choses Red

***

I suppose you only want each player to know, for who they are a double agent.
I also suppose you only want 1 double agent for each player.

Putting a card back means the card of that player is still in.
Putting a card back as third player means that, that player is going to have the fourth player as double agent.

I see your problems in this mechanic.
(Unless I am horible wrong again)

***

Keep the card mechanic. However, adjust it. If one player has his/her own colour, than at least one other player has this as well. They can't simply trade, because every one will know.
I suggest, EVERYone gives their card back, and will pick a new card.
This untill everyone has a different colour. Although it is not a perfect solution.

Draw 1 card at a time. If any player draws their own card, they put it aside and draw another card, then put their card back (so the next player could have a chance of drawing it).

The only time this wouldn't work is if either the last or the second to the last players drew their own cards. In either of these cases you'd need to toss all the results and do all the draws again.

But if my math is correct the chance of either of these happening in a 4 player game is almost 50%, so it would end up causing alot of do-overs. The chance would go down the more players there are (and would go up with a 3 player game), but this would probably feel a little messy even with 6 players (which still has about a 30% chance of requiring a do-over).

Live with the redraw. But to make it quick and painless have somebody shuffle and deal the fours cards and everybody looks at their card at the same time. If one or more people have their own card they immediately say "Redraw" and the next person immediately gathers in, shuffles and deals. Rinse and repeat until you don't get a redraw.

If it helped you could make this question of whether redraws happen or not have an actual effect on the game - A player who declares a redraw is punished or rewarded in some way. At least this way something is actually happening in the game, it's not just a case of everybody sitting round waiting for the game to be able to start.

Yeah, I don't agree with Let-Off. A little bit of randomness during the startup phase may appeal to you or not, based on the tone of the rest of the game. It might unbalance the game or it might just add a little spice at the beginning.

The players need to be double agents for another player so there is no point in anybody trying to be a double agent for their self in order to pick up one small bonus because they won't be able to complete the game properly. In fact, as soon as somebody says "redraw" everybody should turn their card over, see how many players picked their own card (assign rewards, or bonuses, or not - as you please) and carry on.

Perhaps.

Aye, Let-Off, you might well be right about the bonus/penalty issue. I'd have to know more about the game to really form an opinion. If it's a game that incorporates a good deal of dice rolling or card shuffling or some other randomness generator, then I wouldn't worry about a little randomness at the start. Each time we roll dice or shuffle cards we are opening ourselves up to being punished or rewarded by an event we are not able to influence. Some games are chance heavy and some aren't.

Hey, you actually got an idea that might work in one go, without redrawing over and over. With 36 possibilities they might remember eventually.

Use at least 3D6 different coloured dice. This means 216 possible double agents. Also, make sure that you have 4 different "lists" (for 4 players). With that I mean that you have to divide 12 possibilities amongst those 216 randomly:
RYBG
RYGB
RBYG
RBGY
RGYB
RGBY
2x RY-BG (now Red and Yellow can be each others double agents, and so do Blue and Green)
2x RB-YG
2x RG-YB
This list has to be implemented 18 times. Then each possibility is equal.

Let's say that the dice roll 1,1,1 [1]
Then Red has a chart that says Yellow.
Yellow has a chart that says Blue.
Blue has a chart that says Green.
And Green has a chart that says Red.

Now, when you roll 2,1,1, you use [2]
Red has a chart that says Yellow.
Yellow has a chart that says Green.
Blue has a chart that says Red.
And Green has a chart that says Blue.

If you find 216 to much, I find 36 to less. But that is my opinion.
A lot of work. It looks complicated. The end result is an fast, correct way for choosing a double agent.

Hey all,

Interesting ideas (I was wondering when Excel would come in - such a useful program!), and I had a really random one to throw out for your consideration Arthur: what about each card simply having an arrow on it, facing either left, right, or up (which signifies "across"). This would point to who they are a double agent for. The game would have a deck of, say, 20 cards, and the players would each draw one card.

What this does is 1) it avoids the problem of the redo (as players wouldn't "draw themselves" in this setup), and 2) it allows for the game to be standardized for up to 20 players. It also means it doesn't matter who goes first/last - there's no lack of being able to draw a particular person, no problem of being "left with the last one," and for the vast majority of games you wouldn't use very many of the cards.

Now it does mean that if you have more than four players people could figure out who "is not So-and-So's agent," but that also leaves room for alliances - you know who you're "safe" choosing as an ally. And hey - still leaves three choices of who they might be secretly working for.

Just a few thoughts - sounds like a cool game!

I know you don't really want to make an app, but I'm pretty sure that no scheme other than potentially unlimited redraws is ever going to work. Any of the schemes that involve dice rolling will break the game if someone remembers, for instance that we rolled 2-4-1 last month, when I was red, so I know red is double agent for blue. My son has that sort of memory -- he couldn't forget if he wanted to.

The app should accept a number of players, then it chooses a legal combination. Each of the colors are shown, and each player privately presses his color and it shows whom he is double agent for. It should ding when someone presses a color, so you can hear if anyone presses more than once. You pass the phone around once, and everyone knows their status. No communications needed, so nothing tricky to write.

A minimum of 4 was already required for preventing knowing who got you. But other than that, this is briljant!!

Allow me to upgrade this mechanic.

As addition. You could add one dud envelope with no name. And one no name card inside. If you have the dud in your envelope, you have to take the dud envelope.
It is still impossible to get your own name this way. This because all the name cards inside, all go in one direction.
But there will be the feeling of more randomness and A could pick B with B getting A. (I don't know how big the chance is for this happening. Can someone test it out?)

This mechanic works though.

Clever solution, SLiV.

X3M, I don't think that your modification will work. Remember that for 4 players, if A and B have each other, then C and D have to have each other, too. There's no way that your dud approach can make this happen.

So, going back to the simplest solution where you have a marker with each color, everyone draws one and checks it, and you throw them all back and redraw if anyone has his own color. If you wanted confirmation that it's a bad idea, here it is. (I actually started this analysis expecting to say, "Look, it's not so bad.")

With 4 players, of the 24 ways the markers might be distributed, only 9 of them "work" -- that is, in only 9/24 combinations everyone gets a color not his own. With more players, it gets worse.

So, 62.5% of the time, you'll need at least one redraw, etc. Here's how it breaks down. This shows how many times it will take exactly that many draws, and then the sum, which is the chance that it will be that many draws or fewer.

Unfortunately, this says that nearly 10% of the time it will take six or more draws, which is pretty unacceptable if there needs to be shuffling for each draw. I know that I have trouble shuffling 4 cards without tracking the one that I know (which I do without even thinking), so either I have to shuffle enough that I can distract myself by talking, or I have to shuffle and then hand to another person to shuffle more.

If you had a system where there are beads drawn from a hat or something, that might be faster from a shuffling point of view, but you lose the ability for all the players to grab and check their cards simultaneously.

So I agree that this approach is a non-starter. I still say that a simple app is your best bet.

Apologies if you have already answered this ...

I understand that if player 1 is assigned to player 2 then player 2 is not allowed to know. However is it ok for player 3 or 4 to know?

I'm intrigued by this problem :-)