Friday 26 April 2013


Raph Koster recently sparked off a discussion about the role of formalism in game design and criticism by responding to critic and games writer Leigh Alexander in a letter. I recommend you read his response to Leigh's questions:

"when people say games need objectives in order to be ‘games’, i wonder why ‘better understanding another human’ isn’t a valid ‘objective’
games need ‘challenges’ and ‘rules’, isn’t ‘empathy’ a challenge, aren’t preconceptions of normativity a ‘rule’" -  Leigh Alexander writing on Twitter

Raph poses a number of questions himself in reply, but frames the overall mode of a game as being a conversation between designer and player; and argues that the types of art games which Leigh is defending the conversation is a monologue because the designer of these games takes away player choice. He draws such a strong response from the community he criticises because he's defending the conventional position against a more radical interpretation, but he also picks the wrong point to make and the wrong analogy to make it with. Andrew Vanden Bossche has already written about the fallacy of choice in games: in short, games are interesting because they do not give the player unlimited choice, but a selection of choices from a restricted palette. Raph's analogy of games as conversation fails for the most part because it is so rare to see a game redesigned after the wide release - although this may hold somewhat true for games with long public development cycles (Minecraft) or in limited alpha and beta releases.

I've been meaning to write about one of the more remarkable achievements in game playing I've seen for some time but haven't had the opportunity until now to do so. ggoDeye, a regular participant (and 8 time winner) in the Brogue weekend contest, and organiser until recently of the mid-week contests, recently achieved what could be described as the pinnacle of achievement of Brogue 1.7.2: a 'mastery' victory in a competition. A mastery victory is where you leave the dungeon at level 40 with all 25 lumenstones it is possible to get in the game. Short of conducts, there is no other way of beating the game more conclusively in the current version (in earlier versions it was theoretically possible to get to level 100 although no one had).

In the tradition of Brogue a maximum lumenstone victory usually means that the game has been broken in some important way, and a game balance fix will almost always be forthcoming in a future version of the game. Indeed, ggoDeye writes in the victory post of the jelly master build he used requiring almost superhuman levels of concentration to get this victory:
By D19 I had 30+ jellies and the ability to isolate and destroy dar parties. [...] I was so exhausted and frustrated that I took all of Sunday off from playing so that I wouldn’t do something stupid and blow my first real shot at a max lumenstone run in a contest. [...] Throughout this run I gained valuable insights into what I think of as jelly “herd physics.”
- Low HP allies can essentially teleport to the other side of the herd by shifting the herd’s inertia through shifting its center of mass (you). I once noticed a dar priestess move 20+ squares in a single turn as the direction of the herd reversed and higher HP jellies continued to displace her. I used this phenomenon to keep crucial allies alive when they were low on HP.
-High HP, fast moving allies can stick to the center of mass (@) better than anything else.

[...] While I have always enjoyed jelly master builds, I had not come anywhere close to fully grasping their true power and utility before last week. 1.7 has introduced bloodworts and superior ally AI and has really opened the doors on making this a reliable and predictable build option for securing max lumenstones. I’m fairly confident that I can turn just about any seed with a wand of domination and either a staff of tunneling or a charm of shattering (all before D17?) into a max lumen run now. However, I am certain that I will never be willing to do it again with a staff of tunneling. This run took me over 16 hours to pull off and I would not recommend it to anyone who wants to continue enjoying this game for the foreseeable future. This ended up being the most mind-numbing and eye-straining gaming experience of my entire life.
A quick couple of notes on the above commentary. A jelly is a monster in Brogue which doesn't inflict much damage but splits when you hit it into two jellies each with half the hit points. These new jellies do not regenerate naturally, but can recover with a source of healing such as the gas released by a pod from a bloodwort plant, or if they learn the healing spell from a dar (dark elf) priestess by consuming the corpse after gaining sufficient experience to do so. The wand of domination allows you to take control of a jelly and its children, and a jelly master build is a character who relies on having a swarm of jellies (up to 100 on a level) fighting on their behalf, and staffs of tunnelling and charms of shattering allow you to open up the dungeon by destroying walls so your allies can surround enemy monsters.

ggoDeye specifically uses the term herd physics because in Brogue there is no mechanism with which you can directly control your allies: they move simply based on the direction you move in in combination with the game AI, and knowledge of the surrounding terrain and enemy positions. Several items give you more granular control by allowing you for instance to build walls to block movement, or entrance a monster to have it move in the direction opposite to yourself, but none of these have the scalability necessary to benefit a jelly master. So in effect you are controlling tens of monsters at a time with a single 8 way movement each turn.

(Intriguingly, the 'teleportation' mechanic ggoDeye describes is known as enexto() in Nethack after the function that displaces monsters into adjacent grids as they push past each other, and is used in tool assisted speedruns of that game.)

The process ggoDeye goes through in this game is nothing like a conversation: it is more like mountain climbing, or an ultra marathon. We can appreciate ggoDeye's performance because it is on one level a pinnacle of achievement which is unlikely to be duplicated in the same way we can appreciate athletic success. But trying to find a sport analogy to fit games to is ultimately circular because sports are often games, and games can be enjoyed like sports - which means any comparison we make between the two will be clouded.

There is another performance comparison that is more appropriate than a sporting one - that of playing a piece of music. People have compared music and games before: both have groups of 'indies' and in some ways this comparison is more appropriate than games as films, but I'm suggesting here that game players rather than game designers are the actual performers in this analogy.

If we extend this idea of player as performer, what role does the game designer have? Song writer is the most obvious - but I'm going to suggest that Raph's formalist argument is about game designer as the maker of a ludic instrument; and while art (and many AAA) games take the position that a game designer is a song writer, they are in fact constructing instruments that only support one song.

(Or more facetiously, I tweeted this as "A game designer is an instrument maker who thinks they're a song writer and everyone else mistakes for a musician.")

The counter argument of course, is that the instrument may only play one song, but it is a song worth learning to sing.

And while this is mostly a metaphor, you can watch Chris Novello literally play Mario in a way that you have never seen before.

As for ggoDeye, he's chosen to play an instrument that many agree is a finely tuned work of art, and he's bent it it ways its creator only mostly intended. But as hinted in the title of this post, this piece was also a swansong, and he's moved on from playing Brogue to starting a business with his wife, leaving behind a recording of his melody and the notes he's made while playing. I wish him all the best.