Sunday, November 21, 2010

Standalone Complex

A continuing trend across many games genres is a tendency towards every combat encounter within a game being its own discrete entity, and with no real continuity in terms of resource management across these encounters.  Okay, I used big words there that make this sound more confusing than it actually is, but I'm trying to be specific. But let's quickly make it simple and take the example of a first person shooter...

This has resource management continuity

Health was a finite resource in Doom. You started off with 100 health, and every bit of damage you took subtracted from that total, but you could restore health (and even exceed your starting "maximum") by picking up items. If you did badly on a particular section or particular group of enemies, you began desperately searching for medkits to get your health back up to a safer level.

Jump forward to modern titles like Modern Warfare, Gears of War, Mass Effect or Uncharted 2, and you don't have a health number. When you get injured, your screen goes red (to varying degrees) indicating you've taken damage and need to bunker down to recover. When you finish one encounter, you're effectively at full health again and ready for whatever might come your way.

Dragon Age attempts to introduce a small middle ground, by penalising characters who fall in battle with "injuries" that hinder their performance in combat until the player uses an "injury kit" or returns to camp to heal. But for the most part, these injuries don't make a big difference upon a character's ability in combat. Maybe people who play on nightmare difficult might notice a more significant penalty, but I never really noticed any appreciable difference in fighting prowess as a result of injuries.

So a party member is injured. Do I really care?

Each system has it's own strengths and weaknesses, so let's take a look at them. For Doom, it's possible for one bad encounter to turn the tide for the entire level. A player gets through the first few fights and finds them really difficult and so has very low health as a result.  The remaining fights in the level might be significantly easier, but the player can't get past them because they can't avoid being hit entirely, and they simply don't have the health to survive.  This is bound to lead to frustration for the player, as they might try dozens of times to get past the level before realising they'll had to load up an earlier save (providing they actually have one) to do the first bit of the level again. It's generally considered poor design to get the player into a situation where they can't win, and that's effectively what this kind of system does.

On the flip side, this kind of system actually makes the player care about each individual combat scenario more. They know they have a limited amount of resources/health available to them, and hence will be more invested in every encounter with the enemy, because they know it will have an impact as they progress through the level. Some could argue that ammunition provides a degree of resource management for a game, but unless the designers are being unfathomably stingy with ammunition, this isn't really much of an argument. I'd also contend that if as a designer you're being that stingy with ammunition, you're probably making a game that many players are not going to find fun.

We don't play First Person Shooters to go ammo hunting

Now let's look the game without resource management. As a designer, it is easier to make every encounter more balanced and achievable for the player, because you know exactly how durable they are coming into the fight. You know they will have full health and how much damage they can take before dying. There's no need to consider the effects of previous encounters within the level on their current state, and thus it's easier to precisely manage the difficulty curve for the player. This is great in terms of pacing, as it allows the designer to do things like giving the player some easy encounters to make them feel good about themselves before throwing them in the deep end by putting them in a really tight spot. It's easier to provide a balanced game experience, and reduces the possibility of players getting stuck in a situation in which they can't possibly prevail.

However, the main weakness of the system stems from the lack of continuity, in that each individual combat scenario can feel different. If you're simply gunning down enemies with little resistance and the combat feels like a procedure rather than exciting gameplay because the player is never in any danger, then the game has become dull. Because the player's performance in one battle doesn't affect what comes after it, the player can become somewhat nonchalant about the encounter, because once the enemies are dead, they don't matter. The player simply walks along and is quickly 100% ready for the next bit of combat, thus somewhat trivialising "ordinary" encounters that quickly come to feel like "filler" to pad out the game's playtime.

More fighting. Yay.

I touched on Dragon Age's attempt to take something from both styles. Health and stamina/mana (used for special abilities or spells) regenerate very quickly outside of combat, effectively meaning characters fully recover between individual sections of combat. However, anyone that falls in battle "comes back" (I guess they got better) once the fight is over, but with a small penalty to their statistics, making them less effective in subsequent fights. This is perhaps an interesting step in trying to get more continuity across encounters, but it fell somewhat flat from my experience. Besides, as long as a character isn't "killed" completely, even if they survive on the smallest sliver of health, there is absolutely no "penalty" given to them at all.

I'm not sure I like this trend for everything to be standalone. From a design perspective I understand the desire for it, but it reduces the tension for "ordinary" encounters with enemies, and makes them somewhat tedious. In a strange conundrum, resource continuity might allow designers to be lazier by adding more "filler" encounters with generic enemies, but at the same time, makes every such encounter more meaningful and interesting for the player because they provide actual consequences should the player perform badly. This is why the "random encounters" with creatures in older RPGs actually meant something, because you were potentially already low on health and mana and forced to fight for your life against a paltry band of goblins (okay, three half-ogres in this case)...

You have been waylaid by enemies and must defend yourself

If you restore a player to full fitness between each encounter, it might be easier to create balanced "special" encounters, but it is actually more demanding of the game/combat designer, because more sections of combat must be "special" in order to keep the player interested. Standard combat against generic enemies quickly becomes as bland as it sounds, because it simply doesn't matter to you as a player.

In trying to simplify their work by removing ongoing health meters, game designers have actually made their job (and that of the rest of their development team) harder because there have to be more "standout" sequences that are made unique by the mechanics, enemy AI, scripted events, the level design, or any other of a number of factors that combine to make a combat encounter. Without any sort of ongoing consequences, there has to be variety in most encounters in order to keep the player engaged and give them something to care and think about.


  1. "because you know exactly how durable they are coming into the fight."
    In an RPG, you only know that as a designer if your game is very linear, so you know exactly how strong they are. Open worlds mean your character can be powerful or weak at any given point.

    "I'd also contend that if as a designer you're being that stingy with ammunition, you're probably making a game that many players are not going to find fun."
    I disagree. See System Shock, or Thief (limited special arrows). Limited ammo works great, it simply gives a game a different feel or genre.

  2. Fair points, but if you've got a non-linear RPG, then there's likely some kind of inbuilt scaling mechanic. If not, you run into the problem of having encounters that are either far too hard or too easy for the player, neither of which help make for a fun game.

    I never played System Shock so I can't comment on it, but in thief you very rarely had to use special arrows. The point I was making that if you're giving the player resources that they must use in order to complete the game, if you make it such that a few small mistakes (i.e. missed shots) will make the game or an encounter impossible to beat, you need to rethink your design. Unless of course you are attempting to make your game ridiculously difficult... in which case you had better hope you know what you are doing.

  3. "If you restore a player to full fitness between each encounter, it might be easier to create balanced "special" encounters, but it is actually more demanding of the game/combat designer, because more sections of combat must be "special" in order to keep the player interested."

    That's as if I'd said "V/O and good graphics make things harder by obliging us to write good stories" :) I see good encounter design as one of the most valuable things a game can have, so it's not much of a downside if your mechanics oblige you to do more set-pieces. Classic Week has about 30 of them, and they were so much fun to make!

    While some players may find trash encounters to be tense in a resource-based system, mostly you're testing their quicksave/quickload habits rather than an in-game skill. The choice between making your encounters easy or requiring players to reload a save from a long way back (assuming they have one) is not a good one. Long-term resources serve exploration fun games best (eg modern Fallouts) - I don't think they have much to offer a tactical fun game.

    That being said, trash still has some importance in pacing and demonstrating the growth of character power. I think a lot of WoW players reconsidered their position on trash after Tier 9 and the amazing trashless raid (which you'd run four times a week >_<). Classic Week probably doesn't have enough, particularly the zero-trash Tier 6 and 7 instances where you may not get a sense of how powerful your party has become because the bosses are tuned for it.

  4. When I was working on nwn2 I considered the majority of encounters as pacing. I like harder more puzzle like encounters but I wanted the game to be enjoyable to people more used to nwn2.

    Its more important to test the game with people to see if they enjoy it. Its easy to get caught up in philosophy.

  5. Jye: I guess I somewhat implied that "set pieces are bad" with that line. I like set pieces and having encounters that are unique, but I'm mindful that this creates more work for the designer. I haven't played with DAO's combat mechanics as much as you have, nor experimented with all the different ways to make them more interesting.

    I'll also say that I feel it's a little harder to create unique combat while keeping within the bounds of Dragon Age's lore. It doesn't feel like there's the same scope for fantastical scenarios when compared to something like the D&D realm. Or perhaps I'm simply not being imaginative enough with encounter ideas. I probably need to play classic week (and other games with lots of set pieces - any suggestions?) to get my head rolling with ideas for encounters/mechanics that I can put my own unique spin on.

    Corey: I agree that there is no substitute for playtesting to determine whether something is fun or not. I remember employees from Epic games saying that they would give the current build of UT2K4 a run every single day to determine whether it was "fun" and what might have been a bit broken.

    But it's still important to understand the reasons behind the particular design decisions that you're making and the advantages and disadvantages of those choices. This analysis helps you guide the direction and development of your game/mod, and also helps you identify what things did or did not work in existing creations, whether they belong to you or someone else. That should enable the creation of more enjoyable and engaging games, which is the ultimate aim.

  6. System Shock and Thief worked because they designed around limited ammo. With System Shock, the RP system encouraged you to have a primary and a secondary specialization. And since most/all encounters had multiple ways of defeating them, if you ran out of "ammo" for your primary (the intelligence spec used hacking tools, so I put ammo in quotes), you could resort to secondary skills.

    Thief worked the same way in that multiple approaches are designed in.

    My own NWN2 campaign handles open world by using player skill to warn them if encounters are going to be too easy or difficult. Storm of Zehir pretty much does the same thing, trash mobs will avoid you and the Overland Map design and location of story areas encourages the player not to run to the killer areas.

    Personally I hate extensive level scaling (see the youtube vids of completing Oblivion at level 2 or 3). I like the possibility of getting into trouble outside the boss.

  7. @AmstradHero: It's really tough to go past WoW for PVE encounter design, particularly from a Dragon Age point of view. There's enough common ground between the systems and between pausable singleplayer and realtime multiplayer that a lot of the techniques can be brought over.

    You played during Vanilla, IIRC, but I don't know if you did anything like Naxxramas when it was current. I didn't play then, but I've been progression raiding all of WotLK and it's been really valuable to see encounters in their prime. It's a reasonably large time commitment to stay current though; you're looking at a night a week in Cataclysm with a good group, I think, probably more to get set up.

    Classic Week is probably a good substitute for a Dragon Age player, not because I'm a brilliant encounter designer, but because it's essentially a remix album of my favourite encounters. It's a challenging module, but fairly tightly tuned, so it's never unfair. Golems of Amgarrak's final encounter is really great too, but I think it suffers a bit if you don't bring the right toolkit over from Awakening.

    As a bit of a left-field suggestion - Final Fantasy XIII. A lot of JRPGs have neat boss fights, but the turn-based context isn't quite right for DA development. XIII has a *superb* battle system that's closer to a cooldown style, and while it doesn't have enough truly great encounters for that system to shine, it's still worth checking out. In particular its gentle but very long learning curve is worth observing, even if you wouldn't go that slow normally.

    Re the lore: I have a fairly cavalier attitude to the DA lore when designing encounters, but I think a more careful approach could still use a lot of standard encounter tricks.

    For example, Thedan mages can't teleport, and I don't *think* we see any NPCs doing teleporting stuff in gameplay. Classic Week's first encounter breaks this rule lazily (the boss "jumps through the fade"), but most of the other places I use *the mechanic* could be reconciled with that piece of the lore. I periodically teleport the players to Charybdis in the Scylla/Charybdis encounter, but that's Charybdis "sucking them in" and it could be animated thusly. Lots of adds just appear, but they could be more carefully placed near doors or out of LoS.

    You can also consider that the lore is a encounters even in Bioware's hands. Cauthrien is undoubtedly an imposing soldier in the setting, but her personal might in her signature encounter is superhuman. In Darkspawn Chronicles, Alistair/Morrigan/Barkspawn/Leliana as a team are able to rapidly demolish the Archdemon, because it's a better encounter that way. And that's not even considering the gaminess introduced by level scaling.

    Plus there's fertile ground for making stuff up. We don't really have a comprehensive picture of what demons and spirits can do, so I feel vaguely justified for the weird stuff that goes on in Coldhearth in service to the encounter design :)

  8. I didn't get up to Naxx. AQ got opened on my server near the tail-end of my time - I did ZG, MC, Onyxia, AQ20, 1st boss of AQ40 and BWL. I've been occasionally tempted to go back, but I can't justify the time commitment required to experience that content firsthand. I imagine some time spent on one of the WoW wikis could probably tell me most of the relevant encounter information though.

    But as I said, I haven't been concentrating on encounter design a great deal, dedicating my efforts towards story, choice and consequence, characters and level design. Encounter design has been put on the back burner for the time being, though I'm very conscious of the fact that I do need to dedicate some serious effort to it in order to spice up Dragon Age's combat mechanics. I don't want to create another Deep Roads.

  9. That's a pretty decent raiding syllabus. You'll probably find when you do get around to encounter design that a lot of the stuff you remember boils down well to 4 "players".