It’s important to note that sometimes we actually want power creep. As we introduce new content, it helps if it is a little stronger or does something different than its predecessors to get the players excited about it. Sometimes our power formula for the Bruiser class is a little too far off the rest of the classes, and we want to adjust Bruisers upward to bring them more into line with everybody else. Maybe we’re releasing an expansion to an established MMOG and want to effectively reset a years’ worth of raid gear progression so that everybody gets to even footing and all the new content isn’t invalidated for the hardcore from the get go. What we want to limit is unintended power creep.
Whenever we design a game system, we usually have some high-level macro goals for that system. For example, if we are creating a [leveling system], we have to think about how many total hours the player will play the game, how many total levels we want, how to distribute those levels over the course of that playtime, and how much power to give each player at each level to feel good about obtaining that level relative to the amount of gameplay effort it took to obtain it. This top-down design view sets us up with certain driving directives and an overall power curve - a mapping of player level to power level.
So, given that we have this curve of expected power level, we can start placing individual elements on the graph. Elements above the power curve are probably overpowered. Elements below the power curve are underpowered. The thing is that we want a variety of elements for players to play with - we want some things to be a little overpowered and some things to be underpowered on purpose. This distribution provides texture and allows players to improve their own playing skill by learning how to identify the things that are better and using them. The range of acceptability is a band that expands outward from the curve itself to encompass all of the “acceptable” power levels. These criteria also help us identify and avoid outliers like the red dot below.
The unifying factor here is that we’ve created some sort of mathematical formula to calculate how powerful something should given its level. We can use this formula to generate the numerical values (along with some percentage variance to adjust it up or down while still keeping it within the band of acceptability) for new abilities or elements to keep them roughly in line with the expected power curve. Most of this sort of tuning is done with a [spreadsheet like Excel] to handle the math.
This, of course, doesn’t take into account the “softer” aspects of these game elements - maybe this ability has too much synergy with that one, or the environment where this enemy spawns is just too oppressive for the player’s power level because the player can’t evade the attacks, or whatever. Something that makes sense by the formula might still be too strong or too weak. This is where playtesting comes in. We typically run playtests as often as possible in order to ferret out these issues so that we can adjust them if they are too egregious.
The FANTa Project is currently on hiatus while I am
crunching at work too busy.
Got a burning question you want answered?