Looking for fun imperfect information games, tips please!

[skydyr]

I'm surprised no one has mentioned Diplomacy. There are a few weird corner cases and rulings that give it slightly less elegant rules than, say, Go, but on the whole it's the best imperfect information abstract game I know.

Of course, it's possible what you like about poker is the randomness, rather than the imperfect information. Diplomacy does not have any random variables (although the outcome of your own strategy is unpredictable because it hinges on the outcome of the other players' strategies.)

Also, like poker, it's a game where if the other players are playing incompetently, it completely changes the game for everyone; unlike poker, losing incompetently is not punished any more harshly than losing competently.

I'm not sure that Diplomacy qualifies as an imperfect information game. Everything is on the board available to everyone when you plan your turn. You may not know what everyone is planning, but how is that different from go, when you can't read your opponent's mind?

Of course, the interpersonal dynamics in diplomacy and necessity of alliances make it completely different from go as well.

[Boidhre]

I agree with skydyr, Diplomacy isn't an imperfect information game in the usual sense of the term.

[drmwc]

Hungarian Tarok, and Illustrated Hungarian Tarok, are both awesome games with similarities to bridge. You need a Tarot pack and 3 other palyers (although it can be played with 5 in total).

http://www.pagat.com/tarot/xx-hivas.html
http://www.pagat.com/tarot/illusztr.html

[TheBigH]

I personally dislike games of chance. I would be interested if there are any games of imperfect information, but in which luck does not have any influence whatsoever on the result.

[Boidhre]

I personally dislike games of chance. I would be interested if there are any games of imperfect information, but in which luck does not have any influence whatsoever on the result.

Not being a game of chance and luck not having an influence are not the same thing.

[TheBigH]

I personally dislike games of chance. I would be interested if there are any games of imperfect information, but in which luck does not have any influence whatsoever on the result.

Not being a game of chance and luck not having an influence are not the same thing.

In practice they almost always are. I can imagine it's possible to devise a game where the initial conditions have an element of randomness, but in such a way that the randomness can't confer advantages or disadvantages. But I don't know of any examples of that. Do you know any?

I prefer games where the skill of the players is the only factor determining the result of the game.

[Boidhre]

I prefer games where the skill of the players is the only factor determining the result of the game.

Then you want a complete information game. In an incomplete information game you have to do some level of educated guesswork to fill in the gaps for what you do not know and this introduces an element of luck (not randomness). Actually not even complete information games will give you what you want, games of go for instance are based on more than just the degree of skill of the players.

[jts]

I'm surprised no one has mentioned Diplomacy. There are a few weird corner cases and rulings that give it slightly less elegant rules than, say, Go, but on the whole it's the best imperfect information abstract game I know.

Of course, it's possible what you like about poker is the randomness, rather than the imperfect information. Diplomacy does not have any random variables (although the outcome of your own strategy is unpredictable because it hinges on the outcome of the other players' strategies.)

Also, like poker, it's a game where if the other players are playing incompetently, it completely changes the game for everyone; unlike poker, losing incompetently is not punished any more harshly than losing competently.

I'm not sure that Diplomacy qualifies as an imperfect information game. Everything is on the board available to everyone when you plan your turn. You may not know what everyone is planning, but how is that different from go, when you can't read your opponent's mind?

Of course, the interpersonal dynamics in diplomacy and necessity of alliances make it completely different from go as well.

I agree with skydyr, Diplomacy isn't an imperfect information game in the usual sense of the term.

Erm. I believe that not knowing what move your opponent is choosing makes it an imperfect information game, whereas the lack of random elements makes it a complete information game. I believe (perhaps I've mangled the terminology) that what makes poker a game of imperfect information is the fact that each player's pocket is hidden from all the other players. Information asymmetry --> imperfect information, randomness --> incomplete information.

Skydr, in Go, after any specific sequence W1, B2, W3, both W and B know what the game state would be; and therefore W always knows that game state will result from whichever W3 he chooses. In Diplomacy the game state produced by {W1, B1, ...} depends on the value of all the vectors, so in choosing between W1' and W1'', W is not choosing between two game states.

[Boidhre]

Erm. I believe that not knowing what move your opponent is choosing makes it an imperfect information game, whereas the lack of random elements makes it a complete information game. I believe (perhaps I've mangled the terminology) that what makes poker a game of imperfect information is the fact that each player's pocket is hidden from all the other players. Information asymmetry --> imperfect information, randomness --> incomplete information.

Skydr, in Go, after any specific sequence W1, B2, W3, both W and B know what the game state would be; and therefore W always knows that game state will result from whichever W3 he chooses. In Diplomacy the game state produced by {W1, B1, ...} depends on the value of all the vectors, so in choosing between W1' and W1'', W is not choosing between two game states.

You're correct. I've come across two uses of the term imperfect information in Economics, where it's just an application of game theory, where in one definition it is as you say and hidden choices = imperfect information and the second is where players in the game have different information (they can also have shared information). So for instance, Diplomacy and Poker would be viewed as the same under the first usage, they would be viewed differently under the second. Hidden choice + shared information is not the same as hidden choice + hidden information.

Where it comes up in economics is that there's a marked difference in hidden choices and information asymmetries when it comes to people's behaviour. It's been years since I read about it but, hidden choices can actually aid markets reflecting true values of things when the markets are *very* small, e.g. selling off licences for mobile phone operators for instance. If companies have to put in sealed bids once rather than bid in an open auction it results in much better prices for the State. On the other hand information asymmetry tends to causes prices to deviate from what would be true.

[TheBigH]

Then you want a complete information game. In an incomplete information game you have to do some level of educated guesswork to fill in the gaps for what you do not know and this introduces an element of luck (not randomness). Actually not even complete information games will give you what you want, games of go for instance are based on more than just the degree of skill of the players.

I think we're using two different meanings of the word "luck". I consider it to be entirely the workings of chance. I do not consider the hidden choices of my opponent to necessarily be equivalent to chance.

If I'm playing a multiplayer FPS, I don't know where my opponents are or what they're up to (ie. imperfect information). If I get sniped from some dark corner, that's not luck, that's the skill of my opponent. But if I trigger a teleport trap that warps me randomly and I end up falling into lava, that's bad luck because my talent at the game does not influence whether I end up in the lava.

Similarly, in go the outcome of a game may depend on whether I know a certain tesuji or not. But that's just equivalent to whether or not my skill is sufficient to win. Perhaps if I'm tired or hung over, I play worse, but that's just that my skill is diminished on that day. You might argue that the set of tesujis I've mastered or my time-variable powers of concentration contain an element of randomness, but I don't care about that. That's meta-luck, so to speak. Once my opponent and I have sat down at the board, the better player wins.

[jts]

I think we're using two different meanings of the word "luck". I consider it to be entirely the workings of chance. I do not consider the hidden choices of my opponent to necessarily be equivalent to chance.

There's another way of looking at it, though.

Consider the difference between rock-paper-scissors and 21. Rock-paper-scissors does not have an optimal equilibrium strategy. That is to say, whether your strategy is winning or losing depends on your opponent's strategy, and if you both knew the other's strategy at least one of you would want to change. (The technical way to say this, iirc, is that a single round of RPS has no Nash equilibrium, and over repeated plays RPS has no unmixed subgame perfect equilibrium.)

On the other hand, if we look at 21, whether your decision to take another card or stay is winning or losing depends entirely on a random variable. But given the cards you can see on the table, there is one choice that is statistically the correct strategy. If you follow the winning strategy and lose anyway, that's

Can you see why someone might think that rock-paper-scissors is the game that is "no strategy, just luck," whereas 21 is the game where there is a mathematically precise winning strategy, even if the winner of each hand is determined by luck, rather than by strategy?

And can you also see how each of these (no Nash equilibrium in RPS; an optimal strategy that doesn't guarantee a win in 21) is different from Go? In Go what we mean when we say "I couldn't guess my opponent's strategy" is "I was too incompetent to realize how good a strategy S would be for him" or "I didn't realize that he might play strategy S at all." In RPS, what we mean is "If I was sure he would choose S, I know I should have played R; and if I was sure he would play R, I should have played P."

[Zombie]

Some notes on Magic:

To beat a good deck, you need a good deck unless the opponent is a drooling moron, and occasionally even then. Good decks are usually expensive in the traditional formats. This is not always the case, though, where a cheap deck might be in good position to attack the current metagame. If people are, say, playing a lot of grindy board control decks, a fast combo deck that doesn't really operate on the board is a good choice.

A Warmachine player recently wrote a bunch of blog posts that describe the feel of Magic pretty well:
http://rayofpaintstudios.blogspot.fi/20 ... lance.html

Back to Magic, and the expense thereof.

There exists a format called Pauper. The basic idea is that it is cutthroat tournament Magic, but only Common cards are allowed. It's not an official format in paper, but is the second-most popular format played online where it is an official format. The result of the all-commons nature of the format is that deck prices are slashed to something like a tenth or even less of what they usually are, and that the games usually lack really big splashy things. Pauper matches are usually about tempo, momentum and eking out little advantages here and there - something surely familiar to a Go player.

Online, there are ready filters for two different Pauper formats, and official automated tournaments and player-run events run all week. Good communities to check out are pdcmagic.com and gatherling.com

If you want to start a playgroup irl, here's a conversation I had with another person on that:

One word: Pauper.

In more words: An all-commons format. You can be completely cutthroat, and there's like a grand total of two truly expensive cards in the format if you're playing on paper. On MTGO it's the second most popular Constructed format, so the staples are more expensive, but the prices are still something like a tenth of other formats. In paper, about 10-20 dollars a deck barring Sinkholes and Lotus Petals, online, 10-50 depending on deck.

Apart from being cheap, Pauper is surprisingly high-powered and varied as a format, and consequently quite fun too.
So cheap, fun, varied. MtG.

Paper pauper isn't standardized, but here are the more common definitions:

Whatever Gatherer says is common is common. MTGO Masters Edition is not used.
Banlist alternatives:

1. Nothing (surprisingly superpowered madness fun times)

2. Cranial Plating, Frantic Search (a couple of the most nasty things are banned, a lot of power and some uninteractive things are still possible)

3. Cranial Plating, Frantic Search, Grapeshot (the most insane combo decks and one very uninteractive combo deck that is ok power level wise are banned, fast combo is playable but there's a lot of fast hate available in every color).

4. Cranial Plating, Frantic Search, Grapeshot, Empty the Warrens, Invigorate (Most anything insane is banned, playing fast combo is basically impossible)

Pros:
* Clear easily verifiable source
* means Mishra's Factory isn't playable which is kind of sad...
* But also avoids STRIP MINE, which is not sad.

Cons:
* Some lists used online can't be used because key cards like Death Spark and Holy Light are commons only in online only Masters Editions.
* Mishra's Factory is not playable

Whatever Gatherer says is common is common. MTGO Masters Edition rarities are included.
Banlist alternatives:

1. Nothing (surprisingly superpowered madness fun times)

2. Cranial Plating, Frantic Search (a couple of the most nasty things are banned, a lot of power and some uninteractive things are still possible)

3. Cranial Plating, Frantic Search, Grapeshot (the most insane combo decks and one very uninteractive combo deck that is ok power level wise are banned, fast combo is playable but there's a lot of fast hate available in every color).

4. Cranial Plating, Frantic Search, Grapeshot, Empty the Warrens, Invigorate (Most anything insane is banned, playing fast combo is basically impossible)

Pros:
* You can get decklists directly from online sources, which helps define the metagame and aids people who just want to play and not brew. Nearly all Pauper play is on MTGO, anyway.
* Clear easily verifiable source, means Mishra's Factory isn't playable which is kind of sad...
* But also avoids STRIP MINE, which is not sad.

Cons:
* Some people may be annoyed that stuff isn't actually ever been printed as a common in paper form.
* Mishra's Factory is not playable

Whatever was actually printed as Common in paper is allowed
Banlist alternatives:
1. Nothing (Pure insanity)

2. Some of the above. (You're allowing Strip Mine)

3. Some of the above, plus Strip Mine (See ratings from previous definitions for what you chose)

4. Some of the above, plus Strip Mine and Mishra's Factory. (why?)

Pros:
* True vintage, everything allowed feel.
* Mishra's Factory is allowed, which means you can play a manland in Pauper. This is not possible otherwise.

Cons:
* Gatherer says some stuff printed on Common sheets way back in the past is Uncommon, which can cause confusion. That is, checking rarities is hard and you may have to consult obscure print run details from Wizards' website.
* If you allow Strip Mine, it will be hell. Strip Mine makes playing the game miserable.
* If you ban both Mishra's Factory and Strip Mine, there is no real reason to use this definition.

Haha nice to see that there is a cheaper way for people to play the game. Although coming from fighting games, I still find the idea of "formats" a little alien. Mtg is one game, but then there's really a ton of metagames that people are playing with different rules. I guess that's not too different from poker or something though.

In fighting games everyone is playing the same exact game, and tournaments almost always follow the same standardized rules.

20 dollars for a deck seems reasonable, but how often would you have to buy a new deck?

Also I didn't understand anything that you quotes as I've never played MtG so it's all just strange jargon to me haha

It's just suggestions if you want to start a local irl playgroup for Pauper, because there't no official rules. Good to have a format that is defined so there's no confusion as to what is allowed and what is not.

The formats are a bit like fighting game tournaments. SF and Marvel share a lot of similarities in controls and basic mechanics, but there's a ton of difference in metagame and power level. Same basic idea with Magic formats. Different speed, power level, and so on. People just find what they like.

How often you have to buy a new deck depends entirely on the format. Some Magic formats are rotating, that is stuff comes in, stuff goes out every autumn. Others don't rotate but instead build up an ever bigger card pool, and thus change more slowly. They typically have a bunch of staple decks that cycle in the metagame. Think if SF had Yun, Zangief and Dhalsim as cornerstones of the metagame. When Sim is big, people pick up Yun. Yun becomes big, people pick up Gief, which begets more Sim, and so on.

Limited formats: "You get cards onsite as part of the tournament. Build your deck from those, then play the tournament with them. Minimum deck size 40 cards instead of 60." Example formats: Sealed, Draft.

Constructed: "You build your deck from cards you own, none are provided as part of the event. These sets are allowed, anything else is illegal. Of the allowed cards, these cards are banned."
Example formats:
Standard, where the allowed sets are printed within the last two years. Every autumn, the old block (three sets) and it's associated Core set fall from the format while the first set of the new block comes in. This process is called rotation, and thus Standard is a rotating format.

Modern, where the allowed sets are any block or Core set printed from Mirrodin/8th edition onwards. Thus products like Planechase are illegal, but Modern changes more slowly because stuff doesn't fall off. Technically a Constructed format, but in practice close to an eternal one. Famous for Wizards banning stuff with a hair trigger.

Eternal: "Anything ever printed with white or black borders is legal, unless it is on the banned list (silver and gold borders are reserved for silly/promotional cards, respectively, and are not allowed in tournaments). Thus stuff like Planechase is allowed."

Example formats are Vintage and Legacy. Eternal formats don't rotate, but change slowly as new cards get printed. In Magic Online, there is an Eternal format called Classic, which is roughly analogous to paper Vintage's "you can play nearly every card ever printed, even the super insane ones" idea.

How Pauper is usually played:
In paper: "Vintage Pauper". What I suggested in the quote boxes in the first post. Anything ever printed as a common is allowed, apart from a banned list. So choose what's considered common, choose a banned list, play. Since it is an Eternal format, it rotates slowly. Varied, surprisingly powerful, very stable. If you choose to allow cards printed as common in online-only sets, you can port lists directly from online tournaments.

In paper/online: "Standard Pauper". All the commons in Standard. This is the ultimate super cheapo format, because you can usually play with other people's (or your own) draft castoffs that just aren't playable anywhere and that no one wants. Rotates once a year, so doesn't get stale.

Online: "Classic Pauper". Has it's own filter on Magic Online. It's the format which all the official Pauper tournaments get held in. Basic idea is the same as Vintage Pauper: Whatever is printed as a common online is allowed, apart from the banned list. The second most popular format on MTGO. The middle quote box allows you to take Classic Pauper lists and play them in Vintage Pauper. (Some cards are commons only online). As an Eternal format, changes very slowly.


So, lots of text. Did this help at all?

[Zombie]

@jts, afaik, 33% of everything totally at random in RPS should be the "optimal" strategy. A non-exploitable one, at any rate (nothing can win against that, but it doesn't really win either). If the opponent plays imperfectly, you can then exploit his range and actually win at the game.

One of the big things about those kinds of hidden choice games is that while they may feel like they're dumb luck, humans are atrocious random number generators and amazing pattern recognition engines, so that kind of flawed play is the norm, and thus skill (reading the opponent) matters. Of course most people need more than pure RPS to entertain them, but playing double blind games against humans is never truly about blind luck.

[cyclops]

In bridge a deal is a partition of the deck of 52 cards over 4 players. Each player gets 13 cards.
Usually such deal is played at several tables. Each table using the same partition. At each table 2 pairs of players compete for a better score at the cost of the opponent pair. Comparing the score on some deal among different tables measures the performance of a pair relative to another holding the same hands. In bridge you don't depend on a good deal but you depend on getting the most out of a deal in cooperation with your partner. You might compare it with two neighbouring farmers farming the same type of land. If one gets a good result and the other a bad one luck is not the probable explanation especially not if it happens on the average over many years. In this sense bridge is not a game of luck or chance. Most of the luck factor is canceled out by comparing between tables playing the same deal and the rest is levelled out by the number of deals played in a match. In the end you can have about just as much faith in the result of a bridge tournament as in the result of a go tournament.

[skydyr]

In bridge a deal is a partition of the deck of 52 cards over 4 players. Each player gets 13 cards.
Usually such deal is played at several tables. Each table using the same partition. At each table 2 pairs of players compete for a better score at the cost of the opponent pair. Comparing the score on some deal among different tables measures the performance of a pair relative to another holding the same hands. In bridge you don't depend on a good deal but you depend on getting the most out of a deal in cooperation with your partner. You might compare it with two neighbouring farmers farming the same type of land. If one gets a good result and the other a bad one luck is not the probable explanation especially not if it happens on the average over many years. In this sense bridge is not a game of luck or chance. Most of the luck factor is canceled out by comparing between tables playing the same deal and the rest is levelled out by the number of deals played in a match. In the end you can have about just as much faith in the result of a bridge tournament as in the result of a go tournament.

Duplicate bridge for tournaments is a bit of a different animal from bridge as you'd play it socially, as players will often try to push a hand to the very limit to score better than others playing the same hand instead of taking what they are fairly certain to make independently.