Thursday, February 16, 2006

Simultaneous Action Selection = Randomness

There are many games with a mechanic such that all players choose an action in secret at the same time, then reveal them and resolve. This is better known as "Simultaneous Action Selection" or SAS. It is used in games as diverse as 6 Nimmt!, Caribbean, Fairy Tale, A Game of Thrones, Himalaya, Hoity Toity, Niagara, Nobody But Us Chickens, Piranha Pedro, Pirate's Cove, RoboRally, and Wallenstein.

It is my contention that SAS is simply another form of Randomness, not too dissimilar from rolling dice or drawing cards.

When players make choices, they are making intelligent choices, not random ones.

True. But given a finite set of choices, there are usually a small number that are directly beneficial. Given that, the choices are somewhat predictable. This means that you should not always be predictable. A certain amount of reverse psychology and reverse-reverse psychology ensues.

The best play for me is A, but everyone knows that, so I'll take B. But everyone knows I'll probably take B, so I'll really take A. What if they think that as well? Maybe I should take C just to through everyone off.

With multiple choices and multiple players, the final choices basically become unpredictable. Although they are being made intelligently, they are statistically random.

Is this a bad thing? Not necessarily. Some games mitigate SAS very well. Wallenstein, for example, handles it by giving each player the same set of chocies. You put them in a certain order. Each player will always do (or have the option to do) all the choices. So it's less about what actions you ar choosing, and more about when they occur and how much money wil be available to carry them out.

In most games, however, SAS significantly lowers how seriously I take a game. Most of my list above shows the typical SAS game to be relatively light. This is where I think SAS belongs for the most part.

What heavy games have SAS? How is the "randomness" of SAS mitigated? Are there any really clever or novelle SAS mechanics?


At 4:12 PM, Blogger JMV said...

Do you consider blind-bidding random as well?

At 4:16 PM, Blogger ekted said...

Good point. I suppose that qualifies as SAS. The mitigating factors there are potentially: limited resources up for auction and limited money. The only blind auction games I've played are Modern Art (some auctions) and Ys (semi-blind).

At 4:24 PM, Anonymous Anonymous said...

Simultaneous action selection is the essence of game theory. Using a mixed strategy (randomly selecting from your choices according to a probability distribution) is optimal for many intresting games.

It is possible to create two-player turn-based games that effectively introduce simultaneous action selection, assuming the players have similar computational power. (If they don't, it might not be much of a game.) This basically entails both players encrypting their choices and then alternating revealing bits of the choices so that their intent cannot be guessed until all but the last few bits are revealed. Players can't change their intent in the last few bits because they can't find a move with a similar prefix ... finding any move at all will be harder than guessing the ending bits, and the move probably won't be a good one. This is just the situation where you can tell that your opponent has a plan but you can't figure out what it is.

With more than two players, another player might be given an choice where each option is equally valid but one favours you and the other doesn't.

That said, simultaneously selecting the same choice as another player, repeatedly and with disasterous consequences, is annoying.


At 4:27 PM, Blogger ekted said...

And a good encryption system makes the plaintext virtually 100% random bits, cribs notwithstanding.

At 7:51 PM, Blogger Rick said...

SAS and blind-bidding are elements that a lot of gamers equate with "psychology" or "bluffing". The kingdaddy of them all, Adel Verpflichtet, is SAS in almost pure form. I'm not much of a fan, because so much goes on based on a dearth of information. This creates a huge amount of chaos particularly with a lot of players.

Now, one of the interesting ones is Kramer & Kielsing's Maharaja. It's SAS, no doubt, but the spark in the game is getting things done in spite of the chaos via long-range contingency planning. I enjoy Maharaja because the SAS can be used tactically and strategically at the same time, while most of the other games in this category are purely tactical.

At 9:16 PM, Blogger Gerald McD said...

SAS is sometimes used in wargames, and very effectively. Examples: Ace of Aces, Wooden Ships & Iron Men, Diplomacy, and Richtofen's War (if I recall correctly). These, especially the last three, are not necessarily "light" games. However, your point about randomness does apply, since both players (or multiple players in some instances) usually have many options to choose from. If you know your opponent and how he plays the game, you can make intelligent guesses about his likely choice of move, but, again, the reverse-psychology approach plays a role. A player may make what would normally be a bad choice, because his opponent knows it is a bad choice, and that move might give the first player a real advantage several moves later -- or not.

If you limit the question to Euro-games, I can add For Sale as an example, and it fits with your premise of "light" games, as a blind-bidding game. I do not typically do well in blind-bidding games, but used to be able to hold my own in the above wargames.

At 4:39 AM, Blogger Pawnstar said...

Many games using SAS will start off random, but evolve into something more predictable. Adel Verpflichhtet for example; if you know the red player's thieves are in jail when you visit the castle then playing detective is out for you.

I think it is this evolution from total randomness to deductive, cognitive and logical elements that makes some of these games stand out.

At 8:22 PM, Blogger Doug Orleans said...

What do you consider heavy? How about Amun-Re, Game of Thrones, or Diplomacy? Though I'd say Wallenstein is at least as heavy as any of these. I'd also say Lord of the Rings: The Confrontation and Dungeon Twister count as medium-weight two-player games.

At 2:10 PM, Blogger Seth Jaffee said...

I think I agree with your premise here, that SAS is similar to randomness. I don't know if I agree that it means you'd do just as well choosing randomly as making some intelligent choice.

When I first played Wallenstein (recently on SpielByWeb), I thought it was somewhat silly compared to the 'great' game I had heard about, as not only do you not know what order all the actions will come up in, but you also don't know the turn order each round, nor which effect from the event cards will occur.

The only thing that did not bother me about the game was SAS... though that just adds to the chaos.

After playing Wally a bit more (like 3 or 4 games now), I've begun to appreciate it a bit more, that is to say I enjoy it. However I don't think it's any less chaotic or any heavier than my first assessment. I think I just realized that the game is inherantly a lighter game. It's not Caylus, it's Wallenstein.

I might like to try my hand at designing a game with some of the elements I like from Wallenstein - in particular the balance between all the territories: High Gold or High Grain or lots of cities...

At 5:16 PM, Blogger Jason Little said...

Interesting... I don't agree with all of it, but interesting and well written. I think you can still endeavor to optimize your own actions without full knowledge of what your opponents may do -- in fact, all games have this to a degree.

In these SAS sorts of games, it just happens to be more intertwined than in a regular back and forth game.

Is it guessing? Intuition? Probability management? Regardless, it's a decision point influenced by just as many factors as a game like, say War of the Ring or Blood Bowl, which lack the element of "surprise" in action resolution, but still can rely on a lot of intstinctive play, opponent reading, guesswork and more than a little luck.

At 1:07 PM, Anonymous Anonymous said...

Hey. Neat thread. I responded in a geeklist:


At 7:42 PM, Blogger hibikir said...

I don't think SAS has to bring randomness: it's just that most euros that use SAS want to be light and random.

Let's take, for example, one of my least favorite games, Pirate's Cove. The game feels random because the inherent value of each selection is dwarfed by te importance of choosing something different from your opponents. A move can be the best possible or the worst possible, you only know after seeing what everyone else picked. The designer obviously added SAS to create this 'guessing game'. Unfortunately, guessing wrong the first 3 or 4 times takes you out of contention quickly.

If the value of the actions by themselves is more important than predicting what your opponents are doing, SAS can serve other purposes. For example, it can speed up the game, just like in Wallenstein.

SAS can also be used to 'seed' a game situation. Just think of the different value of each of your pieces in 'survive', or the initial bidding in Royal Turf. It'd be trivial to turn those decisions into SAS. After this seeding, normal moves are public, but you've added a guessing element to the game that doesn't feel half as random as a Pirate's Cove island selection. Picking a tile to put in your parrotfish in Reef Encounter causes the same thing.

I think that SAS could be a great tool in heavy games. Unfortunately, most people that enjoy heavy games are big fans of perfect information and analysis. This doesn't make SAS very appealing to designers that design heavy games: Why risk losing part of your target market by leaving the beaten path? If the heavy games market doesn't become stronger, I'm afraid SAS will be relegated to light games.

At 10:37 PM, Blogger ekted said...


You have made the best point yet. Games in which all choices have equal value for all players are "more random" in my sense of the concept. In games with valuable choices, less valuable choices, and no-value choices, at least there's something to base your decisions on. Still though, since an opponent can choose anything, even if it is a stupid move, you have no basis to predict. Maybe there's mathematical (game theory) methods to find the best choice, but people don't play that way.

At 9:44 AM, Blogger Steve Janecek said...

I;m floored by an earlier comment made by Doug that "Wallenstein is at least as heavy as diplomacy" You have got to be kidding...

Although diplomacy utilizes a SAS system, it is certainly not random. With Eks infinite wisdom he did not include it on the original list, which was a smart idea.

In general I agree, for some game SAS is used to create randomness and silliness allowing for moms and children to play alike, while for others, it is necessary for the strategy of the game.

At 9:56 AM, Blogger ekted said...

I omitted Diplomacy because I believe the primary mechanic is negotiation, not SAS. But I still think that the SAS in Diplomacy equates to randomness insofar as the unpredictability of the opponents is concerned. You can make a list of the most likely outcomes, the slightly less likely outcomes, etc., and still fail to do well through no fault of your own. This is no different from shuffling a deck of cards.

At 3:00 PM, Anonymous Anonymous said...

Using the Minimax theorem can give you an edge in the long run. Short run, SAS makes the game very random, as you correctly point out. If there are many successive SAS in the game and not too much hidden information, Minimax theorem analysis should be used.


At 12:44 AM, Blogger Steve Janecek said...

I dont understand how SAS is random in the case of diplomacy. To me, given ample time to negotiate, you should have a good understanding of everyone's moves, and those moves you incorrectly deduced is not likely due to randomness, but lacking the deduction required to determine those moves.

If you think Diplomacy is random for that reason than puerto rico is just as random in such that you prepare to do something based on what "you believe" your neighbors best move to be. If he doesn;t do it, is that randomness or lack of foresight into that player's mind? To me thats even more random than diplomacy because in diplomacy you are given the opportunity to ask those questions needed to determine that player's move, in puerto, all you can do is hope he isnt a numpty.

The greatest randomness factor in any game I have played is the person you are playing with. Either they are an idiot or a good player, or even worse an idiot who thinks they are a good player :P. Thats where your diplomacy randomness comes from, not the mechanic.

Anyway, you did not put diplomacy on your list, and for good reason, I just took homage to your recent comment ;P

At 12:50 AM, Blogger ekted said...

Steve, now you are crossing into the realm of what I call chaos. SAS is when you have to choose before knowing what others do. Chaos is when you have to choose not knowing what they are going to do (I use the term only in extreme cases where too much of the game situation can change before you get to go again). In both cases, the possibilities are known. But with SAS, you have to make your choices based onwhat think they will do. In Chaos, you make your choices partly based on what they have done. It is a fine point.


Post a Comment

<< Home