Decay, Resets, and Entropy
Thursday, March 29, 2012 at 12:12PM Decay
I've talked about decay many times over the years from stale-move negation in Smash Brothers Brawl, to the decay cycles in Neo*RPG, in my post on gameplay dynamics, and in my series on interesting choices. Decay dynamics are powerful yet oddly underplayed in discussion of game design and gameplay. It's an important concept in understanding how many gameplay advantages are created and maintained by players. Considering how decay operates on this level will allow us to better understand level design.
We'll start with outlining what a decay gameplay dynamic is. A decay dynamic is when the incremental increase or decrease of virtual resources or elements affects other gameplay elements, features, or challenges. I intentionally put "increase or decrease" of a resource because many examples can be thought of as either depending on your perspective. It's the classic "glass half-full half-empty" idea. Typically we think of decreasing hit points, magic points, or ammunition as resources that can decay but so too can Chess/Checker pieces, allies, and enemy units. Like the other organic gameplay dynamics, decay is an organic dynamic that is closely related to the dynamics of space and time.
Decay dynamics are really just the dynamic of changing values or what we all know as "counting." Decay dynamics also include the abstractions of counting. The range includes mathematical concepts like addition, subtraction, multiplication, percentages, and rates. Decay dynamics are why the tactic of focus firing so easily emerges in nearly all games with multiple attacking units (see example here). When you use your maximum damage output to reduce your opponent's damage output, you gain an increasing advantage over the opponent. It takes a bit of math to explain how this works, and Sirlin has already done it here so I won't continue with the explanation. The point is, decay dynamics can easily introduce high level mathematical operations into gameplay. Yet, they are so commonplace in our everday lives it can be difficult to see how special they are.
As I've explained in my article Dynamic Clarification, discussion gameplay dynamics is all about outlining a relationship between multiple gameplay elements so that we better understand how changing one element affects the others. The dynamics of space and time mainly involve defining the position of elements in relation to each other or to a timeline. But decay dynamics are different. Decay dyanmics start with a numerical value that defines how much there is of a gameplay element. But the part we are most interested in is how the change in this value affects or potentially affects the gameplay of other elements. For decay gameplay dynamics, we try to understand how potential player actions are affected by the change in the virtual resource.
Just like with any gameplay dynamic, the implementation of decay gameplay dynamics can range from very impactful to functionally useless. As I explained in my series on interesting choices, some RPG systems have decaying resources that hardly affect the gameplay in any significant way. In some games, when you run out of magic points, for example, all you need to do is use one of your many potions outside of battle or visit a save station to restore your abilities. And in some of the most simple cases, the decay of magic points only really matters when you're practically down to the last magic spell you can cast. In these ways, decay dynamics can have a delayed and minor affect on a game's interplay.
Reloading in FPSs is a good example. Often times, reloading is not a problem. Just shoot your clip when you encounter a target and reload when there are no targets around. So you may ask yourself, besides keeping the realism, why keep the reloading feature in shooter design? The answer is, reloading does make a difference. On a basic level reloading puts small breaks in the action forcing the attacker into a defensive position while giving the defender a chance to make a move. Normally reload times are very short so this break in the action isn't very obvious. In this way reloading is often a drawback that develops into soft counters, or elements of contrary motion that only have a small influence on its target. Normally, it takes much skill and metagame development to take the gameplay to a level where any soft counter affects the flow of combat in any large way. In the same way that water finds a crack and widens it, skill and metagame pressure will eventually find these small influences and disadvantages and increase their importance.
The influence a decay dynamic has on gameplay is a matter of how hard or soft the counters are and how dynamic the system really is. For now, all we need to consider is that some decay is hardly a consideration to the player seeking to win, while other types of decay are very serious. The seriousness of losing even one unit of a resource depends on how the gameplay is balanced. And for games designed with significant slippery slopes, like Chess, losing one unit can be very costly so that the consequences ripple forward in time never to be undone. Which brings us to a new term and concept.
Resets
Reset is a term commonly used in fighting games communities to describe when a player gives up a guaranteed advantage or guaranteed control (e.g. a combo) in attempt to gain an even greater advantage. Resets are strategies that emerge in games with a basic risk-reward relationships to the player controlled advantages. Resets are generally attempted by using some type of mixup.
In Street Fighter 4, the longer one keeps an opponent in a combo more of their total damage given scales down. By stopping and hopefully starting a fresh combo, the damage scaling can be "reset" back to 100% effectiveness. Resetting the invisible and somewhat abstract damage scaling in Street Fighter is a great example. But upon taking a closer look, it's not the damage scaling that allows resets to emerge within play strategies. The damage scaling only makes the reset more obvoius. You don't even need combos in a game for resets to emerge. All you really need is for players to exert control over each other temporarily. When they have control over how long their control lasts, they have the option of forfeiting control to try and regain control again. Watch this video for a great example and explanation.
For analyzing games in general, I defined a reset as when the balance of advantages and disadvantages between two or more opposing forces is restored. You can't restore this balance if the disadvantaged player continues to suffer disadvantages because of the slippery slope. In fact resets are the design features that stop slippery slope disadvantages from spiraling out of control. David Sirlin calls such cases "limited slippery slopes." Basically, instead of losing and losing and losing advantages like with normal slippery slope design, a reset allows the situation to balance out so that all players have an equal chance of securing a victory in terms of their viable options. In this way, resets can be a great reliever of tension for players who make mistakes.
While decay dynamics create slippery slopes and resets stop the slipping, both of these concepts help us understand entropy.
Entropy
Entropy is a concept that's closely related to counterpoint and decay. I've decided to change the definition of the term found in the Critical-Glossary to this new definition. Entropy is the potential for emergent gameplay to not work for the player. In this case, work is defined by gameplay challenges and the interplay and goals that drive gameplay. The more tuned a gameplay challenge, the less entropy in the system. Basically, emergence is a reality that is more powerful than we can comprehend. Just look at this article I wrote on John Conway's Game of Life for a fantastic example. Emergence can easily surprise us by showing us new results and allowing us to do new things that we never consciously designed or intended. The more complexities we design into our games, the greater potential there is for entropy; e.g. for code to go awry or for there to be emergent possibilities to exist that make the game less polished, less balanced (in terms of interesting choices), or less focused in terms of content and message.
The more linear a game, the easier it is to design, program, and tune. Having only one limited way to win just makes less scenarios and conditions for designers to test. With less to test, more bugs can be fixed and gameplay can be precisely tuned knowing that players are expected to play in a particular way. Just compare a game like Rhythm Heaven to Skyrim. One is a linear rhythm-action game. The other is an open world RPG. Rhythm Heaven may only have one bug, from my assessment. While Skyrim has bugs by the bucket load. Though I will be writing soon about how linearity is ultimately what works best for gameplay, the allure of emergent options and possibilities makes "open" games very appealing (at least in the west). Remember, linear games have less entropy than very emergent games.
What makes Super Mario Brothers such a great game is how it keeps the entropy low despite a very emergent, layered level design (counterpoint). Open world games like Skyrim or Grand Theft Auto are very emergent in terms of scale, options, and the ability to move around without too many limitations. Put simply, you can do what you want and go where you want, and the game figures out how to make it work. In contrast, the design of Super Mario Brothers is emergent in a layered way.
Four years ago, when writing my Mario Melodies series, I sought to describe and explain the gameplay of Super Mario Bros. By breaking down the level design into 3 main groups of elements (Mario, level structure, enemies) and 2 optional layers (coins and secrets) I attempted to explain the richness of Mario's emergent gameplay. Because of SMB's layered design, JUMPing for coins, searching for secrets, manipulating the level, or attacking enemies all has a tendency to simultaneously influence some other interactive layer of the game. So doing one thing often puts you in a situation where you're forced to confront another. More importantly, how you interact with one layer often sets up uniquely tuned challenges with the other layers. By designing levels around these accidental, emergent, rippling effects, designers were able to create levels with neat emergent secrets and scenarios that are activated by player actions and guided by player curiosity and learning. The design of SMB represents high emergence, low entropy, and high scaffolding. This is the highest level of level design I know and explaining why will be the focus of much of my writing this year. In an Iwata Asks interview, Miyamoto's description of the design of the intro area of World 1-1 in Super Mario Brothers puts it clearly. Do read it for yourself here.
Reader Comments (6)
last link to Iwata Ask interview doesn't work.
I have always found more enjoyment in more emergent, open-world games (though mostly because I've played more of these than linear games) but your quote here is very intriguing to me:
"More importantly, how you interact with one layer often sets up uniquely tuned challenges with the other layers. By designing levels around these accidental, emergent, rippling effects, designers were able to create levels with neat emergent secrets and scenarios that are activated by player actions and guided by player curiosity and learning."
I'm very curious to see how you explore this in your next few articles!
I just read the Iwata Asks interview you linked to. Interesting, but I want more depth! I'm sure you'll give it to us soon. :)
@axcho
I'm glad that statement has piqued your interest. The plan I have for my next year and a half or so of articles is all written out. They will be the culminating work of the past 4 years of writing. Everything is so clear to me now.
I'll give you the depth you ask for. And it only gets deeper from here.
Ahhh linearity is something my thoughts turn to a LOT recently. Mostly because it only seems to be brought up when people feel it's a negative. When complementing a linear game people don't seem to being up how the linearity helped it, but when they don't like a game being linear is often at the top of the list of complaints.
I can't wait to see you tackel it as you often use very clear definitions, which is defiantly lacking in the discussion surrounding linearity. Super Metroid and Grand Theft Auto are both non-linear is very distinct ways. And even with Super Metroid is the person talking about the micro linearity? they way you move back to point A to get to point C? Or are they talking about the macro linearity in the form of the sequence braking? VMega Man for example well being very linear at the micro level hasa high amount of macro non-linearity. But I'm rambling.
Oh and be careful, in my experience when you think you have everything all figured out is when life throws you the curve ball.
@NathanS
Yeah I've been very disappointed how the industry has trashed linearity recently. But it's not like I look to the industry for any real insight about game design or how to talk about it.
You're right, there are a lot of different levels a game can be linear. In my article, I'm not only going to get to the bottom of the term but also what people are trying to express when they poorly throw around the term currently. Should be fun.
I'd be less confident if I didn't have the kind of GDC experience I did. It's funny that you mentioned having it "all figured out." I have been working on some universal theories. lol
Hmm.
@ Richard Terrell (KirbyKid)
"...what people are trying to express when they poorly throw around the term currently. Should be fun."
Indeed, this is probably were my greatest interest in the whole thing lies, but I always tend to want to examine things from a more, anthropological, point of view.