Explain good/bad encounter design to me

[Jaesun]

The Final Battle in Pools of Darkness.

 

[thesisko]

In a given encounter, changing the difficulty level shouldn't mean that his "press a for awesome skill of doom of darkness" suddenly deals double damage. Instead he should be able to cast something else that's even worse, better yet, that counters whatever you might use to survive it in the first place.

I'm talking the other way around though. As in, he already does all that fancy stuff, but on easier difficulties you'll take less damage, immunities become % resistances, enemy buffs aren't as strong so you get away with not dispelling them etc.
Why wouldn't it work to initially tune for a good challenge and then tone it down by changing numbers?

 

[Gord]

So it sounds like Sawyer "gets it", but still people find the encounter design in MotB to be mediocre. So, what was lacking? And shouldn't be enough to play with the numbers to tune an encounter up/down if it's interesting and engaging at its core?

My biggest problem with MotB were the epic levels. Aside from making the world implausible, they somewhat screw up balancing.
Magic users in particular are somewhat handicapped suddenly, as most spells don't grow much in power after lvl 20, while epic-level enemies tend to have a lot of resistances and high hitpoints.
It tends to reduce the worth of magic users as compared to fighters, whose increase in power is much more worthwhile in epic levels - especially if coupled with the shitload of high-level loot you get.

While the story and setting was quite imaginative and good, I would have prefered it to be more in the range of lvl 15 to 25 top.

OK, admittedly this rambling isn't much about encounter design.

 

[Crichton]

I think mostly it boils down to two things:

1) The "right" moves shouldn't be immediately obvious; most of the time this means having a variety of enemies in any given encounter that are dangerous enough to take into consideration.

Examples of this done poorly are too numerous to list, every fight with 4 identical orcs or 3 identical wargs or 8 identical spiders etc. Encounters of x many units of crap and one fragile magic user have the same issue, the obvious move is to focus on the magic user.

Examples of this done well are usually fights with rival "parties" al a BG2, the severed hand fight in IWD2, the goup arena fights in DA:O or the fights at the docks of NWN2. Though it's worth noting that good encounter design can't fix a bad ruleset. DA:O suffers a lot of "I Win" spells like mana clash.

2) Anything gets boring if you throw enough of it at the player. Too much of any particular situation just grows to be old hat after a while.

Examples include the aforementioned 1d10 of enemy_X fights, but also boss fights with HP bloat (it's as if you have to fight the same boss three times in a row) and even the repetition of otherwise-well designed fights beyond a certain point. DA:O has huge hordes of identical encounters; often they're ok the first time you face them, but you can expect to see that exact same encounter eight more times in this one area.

Simple rule sets suffer more from this, DA:O is a good example, but so is Baldur's Gate. There are just only so many ways to do things when every fighter is like every other. It doesn't really matter if the player is fighting bugbears, hobgoblins, orcs or gnolls if they're all just different artwork for the same 1HD no special rules creature.

The best way to get around this is to work hard at giving each creature type it's own personality. Magic helps a lot with this, orc shamans should use support magic to pump up the warriors while hobgoblins focus more on direct damage. Having good stealth rules can also give encounters a different feel; bugbears should have more stealthy units than gnolls or orcs. A change in a small number of units can make a big difference. Even just giving goblins in a cave a web-throwing spider every now and again can change the whole complexion of the encounter.

The only good way around this is playtesting. You don't know how obvious something is or how much is too much unless you actually play through the whole game, not just little portions of it.

 

[Awor Szurkrarz]

Huh, 9 posts and I don't have anything to add. What's happened to the codex? Oh hey it's Jasede, how are you bro?

Anyway, none of OP's examples are entirely correct. BG2 has a bunch of chaff encounters, inappropriately leveled thugs and such, and not much of it is challenging. It has it's moments, but a big problem is cutscene-invulnerability and other cheesy shit like that. Personally I liked BG1 more - maybe a group of goblins isn't that interesting but encountering some roving goblins in the woods makes sense and when you're low level, your characters might die to bad luck if nothing else.

BG1 didn't have any goblins.

 

[Saark]

I'm talking the other way around though. As in, he already does all that fancy stuff, but on easier difficulties you'll take less damage, immunities become % resistances, enemy buffs aren't as strong so you get away with not dispelling them etc.
Why wouldn't it work to initially tune for a good challenge and then tone it down by changing numbers?

Because when designing an encounter you always keep at least one possible solution to that encounter in mind. Devs are too afraid to design encounters without giving the player that solution in the first place any more. Of course you can imagine some super heavy fight that throws all these nifty mechanics at the player. You still want "everybody" to beat it though, don't you? So to make a game challenging, you will most likely have to start at the "everybody will be able to beat this" foundation, then increasing several parameters of that fight. Be it additional enemies, increasing spell power, gaining resistances/immunities or something else.

No one really cares about it though. Why design the games' fights a second, third/fourth [...] time again for increasing difficulty when you can simply scale them to the players level through an algorithm or just double their hp/damage when switching over to the "hard mode"? They don't think of combat as a tactical experience. It has to be cool, it has to be awesome. It doesn't have to be challenging. That would just upset the generic customer they want to cater to. I cruised through Faggotage Origins on nightmare without any problems (besides uninstalling it after the derp roads cause it blew chunks), while every bribed reporter stated that the game was unplayable hard even on normal.

And if you actually wanted to create challenging fights in a new game, you would either have to sacrifice the thought that it will become at least somewhat popular, or scratch every single idea you had about challenging and engaging combat. You can only have one of these two. And for the challenging combat, well I already elaborated that in my first post. Of course you can think of a challenging fight and tune it down. But its just so much simpler and in my opinion more effective to start at the basics, thinking of what you might want this fight to be about. To increase the difficulty, grant them new abilities with increased strength, stronger henchman or mercenaries that also have new perks or abilities the harder you want the fight to be.

 

[thesisko]

Of course you can imagine some super heavy fight that throws all these nifty mechanics at the player. You still want "everybody" to beat it though, don't you? So to make a game challenging, you will most likely have to start at the "everybody will be able to beat this" foundation, then increasing several parameters of that fight. Be it additional enemies, increasing spell power, gaining resistances/immunities or something else.

Wouldn't practically everyone be able to beat Baldur's Gate 2 if the default difficulty restricted all enemies to trivial buffs and spells, reduced NCP damage 75%, removed random encounters when resting, added auto-rest when out-of-combat and knocked companions unconscious instead of killing them? And it probably would still be more interesting at this difficulty level than Dragon Age, even if it was no more difficult.

And if you actually wanted to create challenging fights in a new game, you would either have to sacrifice the thought that it will become at least somewhat popular, or scratch every single idea you had about challenging and engaging combat.

Wouldn't it generate good publicity and better reviews if the game became known for having a challenging-but-fair "Hard/Insane" mode, while still being easy enough at default difficulty?

 

[Jasede]

I'm a friend of the "design the game for your hardest difficulty, then remove monsters/encounters/abilities/resistances for easier difficulties" philosophy.

The hardest difficulty should always be the "true game", not some scaled abomination.

 

[Saark]

Wouldn't practically everyone be able to beat Baldur's Gate 2 if the default difficulty restricted all enemies to trivial buffs and spells, reduced NCP damage 75%, removed random encounters when resting, added auto-rest when out-of-combat and knocked companions unconscious instead of killing them? And it probably would still be more interesting at this difficulty level than Dragon Age, even if it was no more difficult.

It's not like everyone was already able to beat BG2 since it wasn't that hard a game on the lower difficulty levels to begin with. When compared to DA:O everything is more difficult. We're talking about encounter design though, and not a single encounter in DA:O had any diversity. That's the biggest difference between those 2 games. The whole concept of buffs and the breaching of buffs made BG2 so much more "complex", and that's only one difference. Out of like a hundred. Spamming health potions, no perma-death and hp regeneration out of combat won't change that fact.

Wouldn't it generate good publicity and better reviews if the game became known for having a challenging-but-fair "Hard/Insane" mode, while still being easy enough at default difficulty?

If reviews nowadays were actually based on the games content... Maybe. Since pretty much every single reviewer is too dumb to pop moles though... No. What you might consider being easy is too damn hard for them. So you better have some Day-One DLC to make up for it. And it better contains gay nigger Krogans having sex. That sure as hell gives you more publicity than challenging combat.

 

[King Crispy]

I'd like to shift the conversation a bit, if I might, towards endgame encounter design, the ineffable Final Boss -- Foozle himself.

I loved encountering The Master. Although the fight could be considered nearly impossible by some first-time players of Fallout and/or by those who chose a particular type of build but constructed it poorly, for example, and also that there was at least one way to "cheat" in beating him, I still think this was one of the best end boss encounter designs ever. The multiple ways to deal with him, his unique design, his dialog and the creepiness and megalomania that he exuded -- it all added up to one of the most memorable fights from all of my RPG years. It was so interesting it demanded an immediate reload to try something different once you took him out, whatever way, just to see the other options and outcomes.

I actually hated the final fight in New Vegas. Legate Lanius was certainly an interesting character, and the Speech method of dealing with him was mildly rewarding, but it was over so quickly that one never really felt they even barely accomplished much after dealing with him. "Don't attack, come back later." "Okay, see you later then." basically sums it up. Fighting him was equally lame unless possibly for an uber-melee build, because the terrain on which you have to maneuver is much too highly restricted, with his tent on top and your stupid companions blocking any retreat. So it's face him head on (and die) or jump like an idiot Mario to try to give yourself some room. A lot of potential lost here, and it's a shame because that was a character that deserved more than he got.

There are many, many others. Please give some of your favorite and most hated endgame encounters.

Edit: Jaesun actually first brought up endgame fights, so credit to him.

 

[Orgasm]

Variety and Challenge. Even variety in challenge and challenge in variety. Sums up pretty much all the tldr.

 

[Gord]

I actually hated the final fight in New Vegas. Legate Lanius was certainly an interesting character, and the Speech method of dealing with him was mildly rewarding, but it was over so quickly that one never really felt they even barely accomplished much after dealing with him.

On my second playthrough the final fight was actually General Oliver with a couple of NCR Veteran Rangers. This fight, including the whole setup was done better, imho.
I agree that the Lanius encounter feels very anti-climatic. It's maybe one of the more detrimental side-effects of having fully-voiced games.
Somehow a lot of conversations feel somewhat forced and hurried, especially when speech-skills are involved, though I'm not sure if it is only because developers want to save some money for voice-over or if modern design philosophies state that too much text will scare away too many players.

Only yesterday I played the quest where you have to deal with Alice Hostetler and while the quest itself offers a couple of different approaches the final dialogue with Alice herself was pretty stupid, aggravated by terrible voice-acting in my localized version.
It was pretty much this:
"I hate my life and you cannot convince me of anything else"
"But think of what you really want"
"I don't know what I want and my parents are stupid"
"But hey still love you"
"You are right, everything's fine now, kthxbye!"

Yeah, so anyway, I blame poor voice-acting and far too reduced dialogue for the anticlimatic Lanius fight.
Also the fight itself will be over relatively fast. FONV's system is simply not one for drawn out, highly tactical battles.

 

[Overweight Manatee]

For D&D games and related systems it should be really easy to scale encounters for different difficulty levels. The idea of a challenge rating for encounters has been around for decades. You still need good AI and enemy variety, but simply balancing encounters should be easy. Since most enemies should ideally be using similar rules as that players, it should be a piece of cake to simply make an algorithm that scales enemies up and down or adds and subtracts characters easily. You could also easily take advantage of this to semi-randomize an encounter. Say, for the first game the lvl 10 boss has 2 lvl 8 priests, a lvl 8 mage, and 3 lvl 6 fighters with him. Next game, he has a lvl 9 priest, 2 lvl 8 thieves and 2 lvl 7 fighters. Assuming you have a decent auto-leveling system to create these enemies, and if they follow rules similar to players you should be able to simply steal whatever builds your playtesters are making for their characters, all you need to do is place them in a decent formation to start the battle and have a decent AI controlling them.

 

[thesisko]

For D&D games and related systems it should be really easy to scale encounters for different difficulty levels.

So why has this been done so poorly? At least that's the impression I get, since people regard certain games as "too hard" and others as "too easy". For instance, ToEE, IWD2 and SoZ are regarded as being pretty hard.
Would it, for instance, have been impossible to tune MotB to be as hard as IWD2 on the hardest setting but still have the same difficulty on normal?

Why did the other devs, according to Sawyer, stop him from requesting "IWD2 levels of difficulty", instead of saying "OK, but only for the hardest setting"?

 

[Mastermind]

Fighting him was equally lame unless possibly for an uber-melee build, because the terrain on which you have to maneuver is much too highly restricted, with his tent on top and your stupid companions blocking any retreat.

I took the fucker out without breaking a sweat as a half-ass sniper. Then again, I didn't give the fucker a chance to start talking.

 

[Roguey]

Why did the other devs, according to Sawyer, stop him from requesting "IWD2 levels of difficulty", instead of saying "OK, but only for the hardest setting"?

Time and resource management on an expansion-sized schedule and budget. Plus NWN2's hardest setting makes enemies do double damage and that's just wrong.

 

[Surf Solar]

Interesting discussion. While mostly we're talking BG2 or D&D games with a wide variation of critters, how would you design interesting encounters in a fallout game, which lacks bigger variation of critters and so on? Don't want you to do my homework for me, I am just interested in your opinion, as I'm doing some optional "dungeons" at the moment and want to make it as interesting as possible. :great:

 

[zeitgeist]

Interesting discussion. While mostly we're talking BG2 or D&D games with a wide variation of critters, how would you design interesting encounters in a fallout game, which lacks bigger variation of critters and so on? Don't want you to do my homework for me, I am just interested in your opinion, as I'm doing some optional "dungeons" at the moment and want to make it as interesting as possible. :great:

If we're strictly talking about combat encounters, you can compensate for the relative lack of variety in enemy types by a sufficiently complex combat system, a great variety of party customization (equipment, skills, etc.), decent NPC AI, and the playing field designed to interact with the NPCs and PCs in such a way that tactical combat is encouraged (if not outright required).

 

[Mastermind]

Interesting discussion. While mostly we're talking BG2 or D&D games with a wide variation of critters, how would you design interesting encounters in a fallout game, which lacks bigger variation of critters and so on? Don't want you to do my homework for me, I am just interested in your opinion, as I'm doing some optional "dungeons" at the moment and want to make it as interesting as possible. :great:

You don't need critters at all to have a large variety of enemies. Different weapons and abilities can provide a great deal of interesting situations. The most interesting BG2 fights were usually against other NPCs, not monsters.

 

[Overweight Manatee]

For D&D games and related systems it should be really easy to scale encounters for different difficulty levels.

So why has this been done so poorly? At least that's the impression I get, since people regard certain games as "too hard" and others as "too easy". For instance, ToEE, IWD2 and SoZ are regarded as being pretty hard.
Would it, for instance, have been impossible to tune MotB to be as hard as IWD2 on the hardest setting but still have the same difficulty on normal?

Why did the other devs, according to Sawyer, stop him from requesting "IWD2 levels of difficulty", instead of saying "OK, but only for the hardest setting"?

I can think of three reasons:

1. Game system just not being worth it. Even if you did randomly level and scale all of the enemies, fights wouldn't feel much different. This is the case with DA:O. Even if you implemented a perfect CR system that completely randomized every encounter to a specific difficulty, you would have the problem that every mage and every non-mage essentially fought the same way and all your hard work would mean jack shit. This is probably the case with about 80-90% of RPGs, only ones with particularly deep gameplay systems can work with it in the first place.

2. Lack of AI. AI development in the past decade has been basically non-existent if not in decline for RPGs (the majority of other genres as well). Either because its simply not as glamorous and will never sell a game better than BLOOMFUCKINGEVERWHERE, or because developers want to make games that fellate players with how awesome they are rather than games that can actually outsmart a player and make them feel dumb. If you have a complex gameplay system, especially one where enemies have a semi-random assortment of spells and abilities, you have to make a pretty decent AI. Its not impossible, but it requires work. SCS for BG/BG2 does a really good job of this.

3. Time. You need time to figure out what to balance and play test stuff a lot, and since by its nature you have to do a lot of work at the tail end of development it would probably drag out the development schedule by a month or so. You can't start working on it halfway through development, if the gameplay mechanics change then you pretty much have to start from scratch. Game systems based upon established rulesets like D&D would most likely cut down on the length of time needed considerably (though not entirely - no cRPG has a 100% D&D accurate ruleset of course).

Interesting discussion. While mostly we're talking BG2 or D&D games with a wide variation of critters, how would you design interesting encounters in a fallout game, which lacks bigger variation of critters and so on? Don't want you to do my homework for me, I am just interested in your opinion, as I'm doing some optional "dungeons" at the moment and want to make it as interesting as possible. :great:

You don't need critters at all to have a large variety of enemies. Different weapons and abilities can provide a great deal of interesting situations. The most interesting BG2 fights were usually against other NPCs, not monsters.

What model or sprite the enemy uses has nothing to do with the enemy variety in terms of gameplay. That is wholly up to enemy abilities and strength/weaknesses. BG2 fights were interesting because enemies had a variety of spells and effects to throw around semi-intelligently. Unfortunately Fallout is rather gimped in this regard, damage types are irrelevant for the most part and the only distinction you need to make is whether an enemy is melee or ranged and how good they are at making your health go to zero. Because of this, and the limitations of FO AI, I wouldn't say that its really possible to make the combat encounters great in the same way that a BG2 encounter can be. The gameplay system to support that just isn't there. IMO, the best you can do is to have enemies that are using the environment well. Place short ranged high-initiative enemies around blind corners, snipers that hold a position at the end of a long corridor, etc.

 

[Saark]

IMO, the best you can do is to have enemies that are using the environment well. Place short ranged high-initiative enemies around blind corners, snipers that hold a position at the end of a long corridor, etc.

This is pretty much it. Most fights are in open space, where you can scatter to avoid are of effect attacks, be it a grenade or a fireball. For example in FO:NV I had no problems whatsoever taking out the Deathclaw Mother with my sniper build because I could snipe her from pretty much the end of the world. With the same build I couldn't even enter their cave to face the legendary deathclaw though, simply because I stood no chance whatsoever when it came to facing two smaller versions of them, let alone one that needed like 6 armour piercing bullets to the head.

Unfortunately more tactical games with a top down view don't seem to use the environment very much. Coming back to FO1 or BG2 I can't think of any fight that actually was severely limited by the space you could fight in. It would be pretty easy to design an interesting encounter that way though. Take a small corridor with a sharp bend. Place some knights or paladins right around the corner, with like a cleric and mage a little farther away and possibly one or two thieves standing right next to them so you can't take them out that easily with your assassin yourself. Since every sane person would start throwing spells like cloudkill to damage the melee enemies, every archer/mage would walk right into it to face you as well while you "pull" them. Why not make them go back instead to their supportive characters, healing up and taking formation again. The wizard might even be smart enough cast True Sight since he would guess they were scouted. You could also put in a blockade or so to limit the place your melees can walk even more or add a hidden door where your sorcerers would usually stay which opens and frees a horde of critters to swarm and interrupt them. And from that point on one of their thieves starts to roam and scout too.

It would also make for more interesting skill checks, like opening a hidden gateway to engage from a more favorable position, be it through speech or some other non-combat related skill. Hell you could even work in some sort of blackmail-assassination attempt, being a check choosing your stealth and intimidation skills. If it is high enough (or if his courage fails), you can force one of the enemies to betray their own people since you would slice his throat otherwise.

All of this only works if you don't have your generic enemy trash mob though, instead handcrafting special encounters like that. Granting your enemy a specific AI-set that combines their skills with the environment in a way that comes close to what a player might do himself. I suppose no one actually has the time to do that though.

 

[sea]

Unfortunately more tactical games with a top down view don't seem to use the environment very much. Coming back to FO1 or BG2 I can't think of any fight that actually was severely limited by the space you could fight in.

Fallout actually had a lot of situations like this. Using choke points is essential to managing the late-game enemies (it's cheesy, I admit), melee characters are there to knock your ass around while sharpshooters do the real damage, and sometimes traps exist to mess you up in the middle of encounters, usually appearing where you least want them. Playing with locked doors can also be extremely entertaining, again cheesy, but still important as far as environment design for fights goes. It's not exactly Panzer General, but I've found most of the problems with Fallout's combat don't come from the problems in positioning and environment design, but rather they come from poor weapon balance and the fact that aiming for the eyes usually trumps everything else.

Dragon Age: Origins, if not for its shitty character system and lack of variety, also would have had better encounters in places than Baldur's Gate 2 - it's one of the few RPGs in its style I've seen where positioning in the environment actually did matter and where the AI would often exploit your weaknesses (archers pinning your party down from above while mages hit them with fireballs, then tanks running in to mop up). Baldur's Gate 2 unfortunately puts most of its fights in tight quarters and unless you exploit the fog of war mechanics or use meta knowledge, the only real tactics come down to prebuffs since things like positioning, flanking etc. are basically handled automatically (or even made irrelevant with certain skills), and most fight arenas are too confined for elevation to matter. For all its issues in living up to Baldur's Gate 2, encounter design was actually where Dragon Age improved in some fairly substantial ways.

 

[RK47]

For New Vegas, a combination of Sniper Build plus Motion Sensor Explosives did the trick for most of my encounters. Especially when it involves caves. Drop a few at the entrance. Fire a shot inside, kill one, run out from the rest, backpedal like mad as the rest of the mobs exit the caves and trigger the mines.

 

[Saark]

Dragon Age: Origins, if not for its shitty character system and lack of variety, also would have had better encounters in places than Baldur's Gate 2 - it's one of the few RPGs in its style I've seen where positioning in the environment actually did matter and where the AI would often exploit your weaknesses (archers pinning your party down from above while mages hit them with fireballs, then tanks running in to mop up). Baldur's Gate 2 unfortunately puts most of its fights in tight quarters and unless you exploit the fog of war mechanics or use meta knowledge, the only real tactics come down to prebuffs since things like positioning, flanking etc. are basically handled automatically (or even made irrelevant with certain skills), and most fight arenas are too confined for elevation to matter.

I totally agree. The problem with DA:O really derives from its system. It's just not as complex as a game that is based on some pen and paper which is being developed for like 20 years already. A shame, since it's one of the few games that really could've turned out great combat wise. Unfortunately they had to throw in crappy filler combat that doesn't even make sense in a game that handles everything like an mmo, overpowered mages and boss fights that turned out to be pretty boring. Not to mention the whole "cut scene -> repositioning" thing. The only fight I ever had to decrease difficulty was that strange book-thief fight on the roads. Positioning your mage and archer and then moving in with your character for the talk, getting ported in, clustered and raped by three consecutive fireballs 5 reloads in a time is just plain stupid. If I see trouble ahead, I should be able to expect a combat situation and prepare myself.

 

[thesisko]

Here's an idea:

Why not put DLC to good use and sell a different difficulty tuning, playtested by hardcore players instead of the usual idiots. Maybe if developers got extra cash for a real hard mode they wouldn't need to phone it in. I know I'd gladly pay $10 to have a MotB tuned by Sawyer.