Thursday, January 3, 2008

Challenge Level of Random Spawns in a Star Trek MMORPG

A short article on The Fine Art of Gameplay Balance over at got me thinking about two popular ways of providing randomly-spawned tactical-level challenges, and about which of these approaches might work for Star Trek Online.

As the article points out, one approach is that taken by (among others) TES 4: Oblivion: the power level of randomly-spawned opponents is calculated to roughly match the current power level of the player's character. (It's actually more table-driven; the character's power level and location are used to randomly select an appropriate type of opponent from a table, and the overall game difficulty setting is used to calculate the opponent's offensive and defensive power. But you get the idea.)

In this approach, you're never hit with a random challenge that's either too strong (so that you can't win, which isn't fun) or too weak (so that you never lose, which is boring). The challenge level is always roughly what you can handle. Because it allows any unlocked area to be explored instead of limiting the character to "safe" areas, this approach to random challenges can usually be found in open-world single-player games.

Contrast this with how the challenge level of randomly-spawned opponents a character might encounter is determined in most MMOGs: it's by zone. A new character starts out in a relatively safe area, immediately outside of which are a mix of very low-power spawns. After gaining a few levels by defeating these random opponents, they no longer present a challenge -- to continue to gain useful XP, the player must move the character to a different location in which the challenge level is closer to the character's current power level. Some of these locations will contain randomly generated opponents capable of defeating the character instantly; the player simply has to learn which zones may not be safely entered (yet).

Both of these approaches to tactical-level random challenges have strengths and weaknesses in various contexts. What I'd like to do in this thread is talk about some of those strengths and weaknesses, and to discuss which of these approaches (or others) might be the best fit for the kind of game we'd like Star Trek Online to be.

To get us started, here are some questions.

1. What are the strengths of the zone-based approach to challenge definition? For MMORPGs in particular, what are the weaknesses of segregating challenges into specific zones?

2. What are the strengths of the dynamic approach to creating challenges appropriate to a character's (or group's) current power level? What are the weaknesses of this approach?

3. It appears that ST:O will follow the convention of past MMORPGs by defining the challenge level of random spawns by zones ("sectors"). What if ST:O used an Oblivion-style dynamic challenge generation system? Could that work? Would players accustomed to the zone-style challenge approach accept a dynamic system in a MMORPG? Are there any changes that could be made to a dynamic system that would make it acceptable to most gamers in general? Is there some other way of generating random tactical challenges other than zones or dynamic calculation that might work for a MMORPG like Star Trek Online?

4. What about using grouping or other "AI" to increase the challenge level of a random encounter, rather than simply scaling up the offensive/defensive power numbers of a single opponent? Would that be appropriate for Star Trek Online, or is it too much work (or too much challenge) for not enough payoff? What are some of the cool effects or problems of challenge-by-grouping that should be considered?

5. So far this has been about tactical-level challenges. What about randomly-generated operational- or strategic-level challenges for characters who enjoy a wider field of play? Is the zone-based approach still workable? Would a dynamic strategic challenge generator be better? (And what might that look like?) Or is there some other way of generating strategic-level gameplay challenges that we haven't considered?