This one really stretches the definition of "role-playing game." If games were sports, and D&D was football, most other games would be things like hockey, basketball, baseball -- all recognizably sports. Universalis would be ping-pong or synchronized swimming, something on the very fringes of the definition and arguably over it into something else. It is a game without a setting, without a game master or equivalent, and without set characters. While it is a role-playing game in the sense that people take on roles, it is in virtually all other ways different from a traditional role-playing game.
That said, this is a great role-playing game. It challenges virtually every sacred cow of gaming, killing them all with aplomb and then serving them up finely grilled and coated with a yummy barbecue sauce.
How Universalis Gets Played
One of the many, many ways that Universalis breaks "what an RPG should be" is that previous preparation is not only unnecessary but actually counterproductive. Played by the book, as me and my group did, everyone shows up and together creates the setting. No one player has undue influence over anything in the setting, and while they can have elements they want in it, so does everyone else. I strongly suggest approaching Universalis with an open, flexible mind.
Each player starts out with some tokens, called coins. Twenty-five is the default amount. The game starts by selecting someone to go first. When building the setting, each player may spend one coin to create one "fact" about the game world. For instance, when we played tonight I went first (basically because I was being loud and aggressive and had read the book through) and said, "The game will involve superspies." Thus, the game involved superspies. Then the player to my left went, spent a coin, and she said, "The game will have a decadent empire." Thus was born a decadent empire in the game.
Remember, we came into the game with no knowledge of what the game would be about. The facts of the setting we determined were:
* Has superspies.
* Decadent empire.
* Large pack animals -- larger than elephants.
* Alternate late Victorian era.
* Lighter than air craft.
* We're spies for the Catholic Church.
* Fighting the Illuminati.
* There is an alternate indigenous religion.
* Pliocene animals did not die out.
* The Illuminati controls the Nazis.
* Nazis are the occult and intelligence arm of the Illuminati.
* Indigenous people have a technological superiority.
* Race to find a lost city.
Then we all passed and proceded to more normal play. Since there is no-one in charge, the way play starts is this: people blind bid, from their coins, and whoever bids highest gets to frame the scene. They get to decide when, where and the initial event to start play. Each time a player adds an element to the scene, or world, they spend a coin to add that element. That element is then a fact about the game world and acquires importance -- making it harder to be rid of. The more traits that get added to an element the more important it is to the story, thus the harder it is to remove from the story.
So, for instance, when starting (I also framed the scene) I said, "Francois du Maurier, a Catholic superspy, stumbles into the Vatican offices saying, 'I saw the plans the Nazis have . . . .' Then he fell unconscious." I paid from my stash five coins: one to create Francois du Maurier, another to make him a superspy, a third for him to do something (stumble into the Vatican offices), a fourth to create the Vatican offices into which he could stumble, and a fifth for him to fall unconscious. Then I said my turn was over and I passed to the next player, who created a secretary who then proceded to do things . . . and so forth and so on.
For a coin, you can also interrupt someone -- you remove control of the narration from that player, but they still control any story element they introduced. For instance, when I passed control to the next player, I still "controlled" Francois. The narrating player could still do things involving Francois, but I could engage in dialogue with Francois (which can shape the narrative, obviously) and if a conflict occurred I'd decide where Francois's traits went.
For a coin, you can also take control of a story element. Let's say that someone wanted to do something with Francois; they'd spend a coin and, voila, now they'd be in control of Francois.
Pretty much anyone can spend a coin to add a trait to an object in the game at pretty much any time (or, for that matter, subtract a trait). The number of traits that an object in the game possess is collectively referred to as its importance. If people expend a lot of resources in a game object, its clearly vital to the story. An object's importance determines how hard it is to destroy -- traits can also include damage. For instance, during the course of the game, Francois was in a glider shot down by the new generation of Nazi aircraft, the Sturmgeist VIIs, and had his left hand blown off and went into shock. So I wrote down for Francois "no left hand" and "in shock" as his traits; thus his importance went up by 2.
To totally remove something from the game, a person has to spend as many coins as that object's importance, as well as have a narrative justification for it. Ironically, wounding something (thus increasing the number of traits it has) makes it harder to ultimately destroy. This had, for me, a fairly strange but quite pleasant feeling. While the simulationist in me naturally rebelled ("Injured people are EASIER, not HARDER, to kill!"), it was an interesting experience knowing that in this game you could really do the whole business with horrifically injuring a character and have them struggle valiantly on.
How Do You Get Coins Back?
You get coins back, during play, in two ways. First, scenes end. When scenes end, all the players get a fixed number of coins back (the default is five, which worked well enough). The second is to, basically, start trouble -- to create complications.
Here, the game most resembles an RPG. In a conflict, the person initiating the conflict buys or controls the objects of the conflict, describes what the conflict is against a target (which by definition must be a game element they do not control). Then everyone creates a dice pool, made out of d10s; they get a die for each relevent trait towards the conflict, and anything "named" gets a die. So when the Sturmgeist VII rocket fighters attacked the gliders the Catholic superspies were in, they got a die for their name (c'mon, clearly a Sturmgeist VII is cool!), being "rocket powered", "accurate missiles", "highly trained pilots" and "hell guns," for a total of 5d10. The opposite side did basically the same thing. For instance, the person controlling Fritz, the Swiss Guard character in the scene, got a die for his name, "high powered rifle", "exceptional glider pilot" and "flak jacket" -- to project him from the attack, after all. Meanwhile, Reginald Smythe got a die for his name and his "sixth sense" to predict the attack. Isabel Delacourt got a die for her name. They, then, rolled 7d10. A "success" on each die is a roll between 1 and 5 -- the higher the better, but 6 through 10 are failures.
So, lets say I rolled for the Nazis 1, 3, 4, 5, 5, 9 and Fritz's controlling player rolled 2, 3, 6, 8, Reginald's character rolled 2, 10 and Isabel's player rolls 9, well, the Nazis would win the conflict.
The losing side regains the number of coins equal to the dice they rolled -- so Fritz's player would get 4 coins, Reginald's palyer 2 and Isabel's player would get 1. The winning side gets the sum of the numbers on the dice that are successes -- so in this case I would get 1 + 3 + 4 + 5 + 5 coins, or 19 coins. Clearly, in most situations, winning is better. Then the winner narrates how they win, and then the losing side narrates how they loose, both sides expending resources as normal. As I said under importance, actually destroying something requires expending a number of coins equal to its importance -- so important plot elements are very hard to destroy. Even if someone wants to do so. When I had Francois attack Colonel von Metrick, I had absolutely no desire to get rid of the character even though (on account I won the challenge) I had plenty of resources to do so -- indeed, by wounding von Metrick, I increased his importance to the story, and thus made it harder for him to kill.
As you can see, while it is traditional in the sense that dice are rolled, the actual mechanic for what happens is very non-traditional. There is none of this, "You get shot for such-and-such points of damage unless you dodge" business. Everything is handled through narration alone.
What If Someone Does Something You Don't Like?
Clearly, in a game such as this it is highly possible for people to do things you don't want them to do -- such as kill your favorite character, or introduce a silly plot point. The game comes with a formal challenge system. Basically, if a person does something you don't like, you talk with them about it and try to work things out. If both sides are intractable, a bidding war starts and other people can join in on either side. The highest bidder wins. Very simple, but useful in resolving disputes that might arise. (None did in our session of the game; I'm guessing the mechanic is rarely used but vital when needed.)
I'm still getting my head fully wrapped around this one. I know that I want to play it, again. I might even want to play in a regular campaign using Universalis. Furthermore, the game is ideal to certain situations that other games are wholly incapable of handling -- such as cooperative world design. I've heard tell that people use it for figuring out metaplot issues, too, at which it would be excellent.
For me, what is looking to be its biggest strength is that, well, frequently a person wants to play in a given setting and ends up running a game set in that setting, instead. I think I could sell my game group, pretty easily, on playing a given setting using these rules . . . meaning I'd get to play in the setting! I'm actually thinking of trying that.
The game is also very fast. Since the detailed combat mechanics simply do not exist in Universalis, and what happens is handled via narration and a resource expenditure, I don't think hours long combat is something that happens very often in Universalis unless the players really want challenge after challenge in hard progression. In one short game session there were four challenges that were fights -- an attack on a hospital room, a brief combat between Francois and Colonel von Metrick and two bits of gliders against Sturmgeists. In a traditional RPG, getting done with even TWO fights in a single game session would basically absorb all the game time, but in Universalis it was a breeze.
I should moderate this rosy picture with some issues from the other side. Its requires a lot of quick thinking. With nothing at all set, it takes a lot of focus and energy to come up with scenes. Because a story could basically shoot off in any direction at any time, I think it would be hard to keep a tight plotline going. Players that aren't verbally aggressive could easily be drown out by players who are -- particularly if those players work the challenge mechanic and thus accumulate many coins. But, to me, these "flaws" (which would only be flaws with certain groups) are fairly small compared to the virtue of the game.
In short, this game is great. You should go buy it right now. It is easily the single most innovative RPG I've seen.