🔒 Locked

Emergent patterns

Started by Mephs Jan 25, 2007 at 5:22 AM 5 replies 2.1k views

Original Post

Jan 25, 2007 at 5:22 AM

I recently read the article on gamasutra that was an extract from A theory of fun for game design. In the article part of the point made is that players find pattern recognition fun, and learning new patterns increases the fun factor, but once a player is repeatedly faced with the same patterns, the fun factor diminishes. I would agree with this in most cases. This may be an assumption in part, but assuming that this is correct, it would explain why collectible card games are fun. Players learn a simple set of core rules and must utilise a large number of cards which adhere to these rules. The large number of varying cards, combined with the simple core rules combine to form a very large number of patterns. Introduce one new card to a deck and we don't simply add one new pattern to a game, the card has to be played bearing in mind the current situation and how it will interact with other cards both at the present time and in the future. From the addition of a single card the game designer gets a lot of bang for his buck (though it would introduce more work in balancing the game). The player will be entertained for a longer time because of all the introduced patterns. Because the core rules are relatively easy to understand players can anticipate the new patterns and their effects, but because of the huge number of patterns available the player must discover the patterns first, which in my opinion only adds to the fun, there is a lot of room for the player to optimize their strategy. So from this, I would deduce that a good way for an Indie developer to create a game with a lot of replay value and a high fun factor with minimal effort is to develop a game that features such emergent patterns. One way to achieve such emergent patterns whilst keeping a core simplicity is to rely on multiple sets of simple factors. If we take a crazy golf game for example, we could develop 64 obstacles to provide 64 different challenges. Alternatively we could develop 8 types of ball. Each ball would have a simple physical property that on it's own is easily understood, it could be magnetic, spiky, light, heavy, invisible, fast and so on. Now we need a second set of factors, which in this case could be obstacles. Each of this second set of factors should react differently in combination with the first set of factors. Taking the example of obstacles, we could have magnets, fans, bumpers, speed ramps and so on. If we have 8 factors in each group but each factor reacts differently with the factors of the other group, the player now has 64 different combinations of these factors, and as such has many more patterns to learn, whilst only having to base their decisions on an understanding of a total of 16 factors. That's less work than having to design 64 obstacles, but provides more patterns, and probably more fun for less effort. Obviously I would imagine this could be overdone, we could introduce a 3rd or 4th group of factors, but I imagine there would come a point where we get diminishing returns in the fun factor as the game becomes too complex and thus offputting. Do you know of any other games utilising this kind of priniciple, or do you have any ideas of your own that use combining factors? Most games do to an extent, but some more so than others. Chess for example offers a small number of basic moves, but there are many combinations of moves as there are other factors such as the position of the piece to be moved, the choice of which piece to move, and the position of the opponents pieces. I'd be interested to hear peoples views on the subject :) Cheers, Steve

FadedPhoenix

Jan 25, 2007 at 9:56 AM

Great book. I am about to be finished with it.

Tigernado

Jan 25, 2007 at 11:18 AM

I own this book too. I read it twice. The hard part will be applying emergent patterns to many diffrent types of games.

Wush

Jan 25, 2007 at 12:50 PM

Ah Emergence is something really good just hard to handle because emergence in essence means unpredictability, that why you need good intution, experience and have to do much testing and balancing.
You know I am a fan of artificial life, so in this regard I occupied mysely with cellular automata which are somewhat the epitome of simply rules with emergent behavior and I thought how great it would be to base an artifical life game on cellular automata,our to use them for creating a basic biochemistry our hidden genome which can evolve in unexpected way so you can have a true evolution of orgasnisms, but hell its just so hard to predict anything aboout those thinks thats you have to be a genius or just have damn luck to find the right automata rules for that purpose.

Well but I was always for games having more unique game physics so emergence can arise, so that just on my taste. Unfortunately many game designers seem to be afraid of the difficulty I mentioned can be involved with this.
Well but if you start with simple stuff and later maybe develop some way to optimize balance and behaviour by using some kind of evolutionary algorithm you can probably get to pretty advanced heights of gameplay with it.

SantaClaws

Jan 25, 2007 at 1:37 PM

I wonder the same result of an increased number of patterns can be achieved by instead limiting the time a player has to learn patterns. An example would be a multiplayer team game where the round goes through stages and the tactics (but not mechanics) change significantly for each stage.

If only 20% of the total time of the round is spent playing in stage 2 of the gameplay, the player only has a limited time per round to learn and adapt strategies that stage. Even if they "master" the patterns in stage 1 of the round, they may continue having fun because they're constantly working towards the still new experience of stage 2.

makeshiftwings

Jan 25, 2007 at 4:01 PM

I agree that emergence is a great thing for many games. The chess example points out another thing that's good about emergent patterns: they make the game easy to learn, but difficult to master, which is almost always a good thing in my book. You should strive to keep the individual rules minimal and simple, but let them combine to form a more complex game.

Wush has a good point though, designing around emergent patterns can be very hard. Because the game is also "difficult to master" for you as well. That's why a lot of games that rely on this sort of thing (Collectible card games, RTS's, etc.) often end up getting patched or revised, when the players begin to find patterns that the developers didn't see that destroy game balance. I don't really have any advice for this though; I've often wondered myself how the designers who do get the balance right ever managed to do it. Extensive beta testing, I guess.

Tohwaku

Jan 27, 2007 at 2:36 PM

Hi there people ! This is my first post here and I just can't help talking about Emergence, attracted to this topic as a drowning fly in a sink's whirlpool. I'm fascinated by this unpredictable, ever-changing god-like creative force.

First I'd like to say I'm theorically supporting the idea of independent devs relying on emergence whenever possible (and relevantly useful). In practice, however, I would myself mention the difficulties to control its phenomenal powers (for by essence Emergence cannot be predicted, though I do believe good preliminary thinking can anticipate many flaws).

I'm totally ok with Mephs' idea about card games, and I would add something about a subject I know better (Pokémon for instance). Pokémon's success lies, in my mind, not only on some players craving a complete collection of whatever item/monster/x-out-of-y thingy, but rather on the ever changing nature of the gameplay. Yes, each Pokémon has a limited set of stats (but in my mind the most important is clearly the "nature" of the Pokémon — ghost, vegetal, fire, and so on).

This is just because any combination is nuanced by a simple variable. But having only one Pokémon wouldn't cause emergence to appear. It is only because you have a whole bunch of them at the same time (and like 5 in the same team), and because they are slightly different that you will have a virtually boundless fun combining them. The critical complexity needed for Emergence to sprout new pattern is achieved not when you just get a Pokémon (even if it can have 10 different shapes or powers, which in reality is much more than an actual Pokémon can show), but when any change in a team or on a Pokémon affects the way the whole team has to be used. Thus, a mere change in a trainer's team forces him to think again his strategy in order to optimise it, and new patterns pop-in again and again. I might not be so clear, so please forgive me :)

In my mind, card game designers need not to have such a hard (and long) time beta testing any combination — because it quickly becomes clear that it is impossible to test and balance everything manually. One player will always find the uber card combination, cross-using several extensions' cards (if not within the same edition). I personally think that the one and only rule when designing emergence is to always add very very VERY minor changes to the basic rules, proportionally to their proximity to the core of the original game design, and only if absolutely necessary.

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.