Simulation versus Stagecraft

An excerpt from an interview and an occasion to save old comments:

Yoru: Right, it’s uh… the buzzword was.. (taps pen) Systemic game design.

Raph: Right, you can poke and prod it and players can have an impact and see it. Statically created content, this is why the debate in the industry is simulation versus stagecraft.

Yoru: Disposable content.

Raph: Stagecraft, yeah, it’s consumable content.

This stuff rubs me the right way. But I also think that systemic game design isn’t the same of procedural or generated content. Yes, they may work together, but they aren’t the same thing.

I say this because I’m ALL for systemic game design. It has been one main concept in all the things I write about and my design style always pivots areound a systemic point of view.

I like control in game design. PvP is essentially systemic game design. It isn’t procedural or generated, and for sure it isn’t disposable. PvP is repetable and it works ONLY when what happens was already well thought and calibrated by a game designer. Or there’s *absolute control* or there’s just an exploit that you have to patch. No control = bad.

This doesn’t mean that PvP is fixed, static or predictable. You provide the players the elements that work together, that affect each other. Is “Chess” fixed, static or predictable? I think not. But you don’t invent rules while you play. You just have control on fewer elements that are pre-designed to work together. Then you can create tactical situations and infinite variations.

Reactive content, but, again, not procedural and not generated.

(btw, why Raph knows always the one word that summarizes pages and pages of concepts? It’s incredible)

Some comments (on Brokentoys, here and my notes file):


In a systemic model:
– The players are brought together. The model is represented as a circumference, where the players/dots create groups or “cells” and move within while bouncing one against the other (creating alliances, conflicts, politics etc..). The space belongs to them (known) and is “managed”.

In a linear model:
– The players are spread apart. The model is represented as a vector, where the players are pointed toward an obligatory direction and have a set position that “qualifies” them toward the other players. The space is external, alien (unknown) and only conquered and progressively consumed.

In a systemic model every element has a precise function and is then linked with other elements in a complex relationship. This means that the function is always preserved. In a linear model, instead, the idea of progress means that you leave things behind. You use up. The function of an element is just about leading you to the next (consumable content).


My personal choice would be about players’ interaction and “structures”. So PvP systems and rules more than artificial worlds responding to mathematical, sophisticated algorithms.

The point against ecologic systems is that they are cool ideas™. But they serve very little purpose if not as an academic experiement. What’s their role in the game? It looks like design experimentation just for the sake of it and not because it answers to a real need of a game.

In the past I also imagined quite complex herbalism systems. But just to “fake” an ecology with those patterns that I think are fun to play and help the immersion, not to build an autonomous system following obscure routines hidden to the player.

(on the choice/selection of elements)


Herbalism system, simulate the growth of plants, but without animals coming and eating them. Because that’s not anymore the player playing. It’s the system playing, while the player is a spectator.

(see Ubiq’s “system design that is more interesting to watch than to play”)


It’s the system that plays behind that I see as a sophistication. Where the player isn’t a player but just a spectator. Like watching a game of “Life”.

Your example about PvP would work if you started to design each tactical element carefully. So it would be more about enabling those patterns and faking them more than “automating” them.

It’s like “chess”, you can have some sort of emergent situations coming out of the game, but the elements into play must be known and strictly defined. So in a PvP game you need a strict control on what comes into play, then you can let the players move the pieces and create situations.

The “control” is always in the hands of the designer, sometimes passed to the players, but never directly to the system itself.

The idea I criticize in the ecologic system is that it wants to be closed and autonomous. Whether the players are there or not, it continues to move on following its rules. This specific pattern, in a game, can work only if the players are the manipulators (god game) and not the manipulated. It’s a matter of “awareness”.

At the end in a simulation (and I’m pro simulations since they help the immersion) you are forced to choose the elements that are useful and that you are going to add, and those that are superfluous. I see this argument from this perspective.

In my herbalism system I imagined growth cycles, seasons and climates, but I didn’t think about critters coming and eating the plants to start a food chain. The reason is that the few elements I added were used by the players and contributed with some concrete complexity, while a food chain would be not anymore about the player playing, but about the system playing itself. With the player as a spectator.

It reminds me that article about the simulated world in Second Life. That’s a place you visit. It’s not a place where you stay and play.

Which you also defined as “cool™” right in the title.

Concluding, what I criticize is not the strive for simulations, realism and immersion. But considering that approach as a better way to generate worlds passively.

And in PvP you need immobility. You can have unexpected situations coming out of a proper use of the pieces, but you cannot have new pieces showing up and coming into play.

So let’s revert the design approach: let’s imagine what are the patterns that are interesting and then think how implement them, instead of implementing an automated system and then think which interesting patterns it could provide.

Leave a Reply