On Stats

On Twitter, Mattie Brice (@xGalatea) made some comments about ‘stats’ in video games, the necessity thereof, and the relationship between the presence or absence of stats and games being designed for mass appeal. Essentially Mattie suggested that eliminating stats from games was often just good design, rather than designing for mass appeal, and that generally they weren’t necessary in video games. This set off a whole lot of conversation about stats and games and mass appeal. A lot of that conversation touched on things that I have an interest in, and I was mentioned in connection with the discussion, but all that happened while I was asleep, because I live in Australia, and am currently (this week) maintaining a normal sleeping pattern. So I couldn’t respond at the time, but since I had several things I wanted to say, I thought I’d use this much-neglected blog to do so.


Firstly, Mattie wasn’t too explicit about what she meant by ‘stats’.¬†She did eventually clarify that she was referring to things like D&D’s basic attributes of STR, STA, WIS, DEX, CHA, CON. Two things on this: 1) The discussion didn’t distinguish much between stats that are visible to or alterable by the player and those that are not, and 2) the vast majority of video games have stats, even if the player never sees any of them.

What I mean is that almost all video games assign numerical values to things to represent their properties, their characteristics, just as D&D’s attributes represent those of a D&D character. These ‘stats’ could be as simple as the dimensions of the character’s sprite on screen, or how much damage a weapon does, or factors in complex systems that derive other values from these basic stats. And I don’t see any particular reason to distinguish between any of these, as they’re all numerical values representing properties of things within the game. They’re all equally ‘stats’.*

The distinctions that are interesting are whether the numerical values are visible to the player, and whether the player can alter them. And there are levels of visibility, in terms of how the numerical value is represented to the player, at what level of abstraction, etc. Does the player get raw numbers like D&D stats? Do they get a large explosion from a weapon with a high damage value? Do they see a weapon that fires quickly doing so? Do they get enemy health numbers, or percentages, or just a green/yellow/red colour coding? Do damage numbers float above an enemy’s head, or does the player just learn that it takes three or four arrows from a bow to kill a certain creature? Different ‘stats’ have different levels of visibility in different games. The same goes for how the player can alter the ‘stat’.

Okay, so most games need ‘stats’ of some kind, but do games need those visible, D&D-style character stats? Well, that depends on what kind of game it is, and how it’s played. Different games do different things differently, for different purposes, for different players and approaches. This is where my tabletop RPG background comes in, because I think the language of tabletop RPG theory can be useful here. In particular, things like Ron Edwards et al.’s GNS theory, especially its more expanded version as part of the ‘Big Model‘, can be useful for articulating and discussing different modes of play, and how different games can serve them. It’s the language that interests me with this tabletop RPG theory stuff, because I feel it is useful for discussing things that video game criticism and design language does not seem to have well-developed vocabulary for, even though video games often venture into very similar territory.

That’s where I’d point people to to explore questions about whether games ‘need’ X, Y or Z. Suffice it to say that I think, yes, visible D&D-style stats are useful, even needed, in certain types of games, and suited to certain approaches to those games. I think it’s important to recognise that not every game needs to be suited to every approach, and I’d actually argue that the more diverse approaches a game tries to be suited for, the less well it serves any one approach. So whether games ‘need’ stats might depend on whether they simply serve an approach that you might not have considered.

As to mass appeal, I personally think that’s a term that’s too loaded with negative connotations in video game discourse. I’d prefer ‘accessibility’, though even that has its problems. I do think that trying to serve too many masters generally does not serve any of them particularly well, but I also think that far too often the critical and player discourse around games is more about whether the game serves the particular player or critic’s tastes, or preferred approach. A greater recognition of the point I made in the previous paragraph, that some games are simply better suited to certain approaches over others, might be useful in that regard too.

*Some games really don’t have ‘stats’ in this respect. For example, many visual novels (not all) or other choose-your-own-adventure style games (not all) don’t have numerical representations of anything, they just record different states or navigate a tree of choices. But we’re not talking about those games. Because they don’t need stats.