i'm coming to the conclusion that it takes more that just a big seamless level (IE an "open world") to make a good game.

Absolutely! Several designers are of that persuasion (I'm pretty sure Raph Koster and Richard Bartle are of that mindset), and I'm on that page also. I apply that to content as well - I've come to that conclusion that it takes more than just MORE enemies (or even more different enemies) to make a good game. It takes more than just MORE items to make a good game. We need better worlds, not bigger worlds. Better items, not more items. Better enemies, not more enemies. Better dungeons, not more dungeons. Better NPCs, not more NPCs.

Ofcourse, we do need enough items/dungeons/NPCs/monsters to provide enough content, so more is better, as long as it's more of quality content. A single room, or a single NPC, and a single item isn't enough content, no matter how good it is. But after you hit a certain point then you need not just more content but also better content, I think.

Basically, I want content-dense worlds, not larger worlds, where content = quality content, not mass-produced content. When the world is so dense it can't fit more content into it, then expand the world vertically and horizontally. I don't want it just sprawled out across a flat plane stretched over a heightmap.

To me, "better" is not "more of what is boring", but instead "more enjoyable". I'd be delighted with games that are only 10% of the size of the GTAs and Farcrys and EQs and WoW, if the game was more enjoyable.

a big seamless level with a few hundred hard coded non-respawning spawn points is nothing more that a giant doom level. shoot all the badguys and then its just an architectural walk through.

Yeah, and even if a big seamless level is filled with buttloads of content, that doesn't make the content well-designed or the world well-shaped. It just artificially stretches out and waters down the content, with longer walking periods and annoying flying creatures between each enjoyable encounter.

Right now I'm kinda in a "Open compact world" mindset. The world should be content-huge but area-small. But on the same token, content needs to be refined down to quality content and should just have more content merely for the sake of more content. I'd rather have ten great encounters than 100 mediocre encounters, for example. Or 10 great quests than 100 boring quests. 10 great dungeons then a hundred bland dungeons.

respawns, random encounters, quest generators, house design (a la the sims), and replacemnt of important NPCs (such as dead merchants killed by dragons in skyrim) all seem to be called for.

Well, here's my thinking: Procedural generated areas is fun (think Minecraft), but even better than random generation is finely-crafted / hand-crafted areas. While hand-crafted areas have less replayability, I personally enjoy them more.

I think procedural generation is an amazing tool in the hands of area designers, but that the area designers shouldn't delegate entire control to it. (In-general. In some games, pure random-gen makes perfect sense - e.g. RTS arenas, roguelikes, and so on).
Procedural generation, in my opinion, should be used to help area-designers create hand-crafted areas faster and better.

I feel the same way when it comes to content within the area: Procedural quest generation is fun, but I think ultimately we'll head in the direction of using procedural quest generation to help designers hand-craft even better quests.

Procedural monster generation is fun, but I think procgen will be used ultimately to help aid designers in creating better monsters and monster placements

I agree (to some extent) about respawns and (to some extent) NPC replacements and (to some extent) random encounters. But again I think procedural generation will work together with, and be heavily guided by, and fill in the gaps left from, the designers.

they finally add random encounters in skyrim and what is it? you encounter a dragon in town while hauling loot back from the dungeon to the store! wait a sec there guys - isn't that dragon supposed to be at the bottom of the dungeon, not waiting for you back at town once you get out? <g>.

ROFL. He was just there buying half a hog - you interrupted his shopping trip. :lol:

You stopped-and-frisked him purely off the color of his scales, you racist! :P

>> and annoying flying creatures between each enjoyable encounter.

you mean like these? <g>

http://images.akamai.steamusercontent.com/ugc/41973090953611681/B2C79CEEA3226ED65611D24E650E38121147ABA7/

i remember reading an article where the reviewer said the same thing about skyrim dragons. and they're right, epic dragon battles become nuisance encounters. i think its because of where and how they spawn. i can't rememeber how many times i've come out the back end of a dungeon that opens onto a little spot high up the side of a mountain where you can only fast travel from - and then they give you a dragon encounter - when your pinned down on the side of a mountain.

>> Or 10 great quests than 100 boring quests.

as of last night, while playing new vegas, i came to the conclusion that most "quests" in games these days are not really "quests", they are merely "errands". Courier Girl has pretty much stopped following quests. she's gotten to the point where she can't complete any more because they require a diplomat (speech), thief (lockpick), medic (medical) or are mundane "fetch quests": questor A needs scrap metal, doesn't tell you where a source is for sure, and gives you no money to buy it with. same idea with questors B, C etc. "bring me 10 charred skeever hides for 100 gold" is not a quest - that's grocery shopping! "bring me the ring of the dark lord - pried from his dead, cold hands - in exchange for a kingdom of your own". now THAT's a fetch quest! anything less is just BS. i suspect this is what happens when mere dilettantes are put in charge of design.

>> Well, here's my thinking: Procedural generated areas is fun

seems like re-play-ability is not a priority for you. but if you just play a game once, and it only has a some tens of hour of content, you never really get to advance over time - which is what RPGs are (supposed to be) about. i'm looking for a game to invest my entertainment hours in long term, not some lighter fare to merely burn though while killing time.

>> You stopped-and-frisked him purely off the color of his scales, you racist! :P

hey man - dangerous times call for drastic measures! <g>.

not really - profiling really just makes you overlook the clever badguys. its the one you don't see coming that gets you. and the first thing a clever badguy is going to do is not match the profile.

>> In reality how would that large animal get there, through those narrow windy passageways full of orcs?

though the caverns at the back end of the dungeon of course! <g> the ones that open up to the cave high up the side of the mountain where it rises from the sea. the wizard's ruined keep sits perched atop the crags, almost ready to sluff off into the ocean itself. and below it, the dungeon to end all dungeons (100 levels, 5000+ rooms, 2500+ encounters at random locations, plus a few hard coded ones such as the great orc hall). they say that on a moonlit night full, you can occasionally see the great beast taking to wing from the cave, out to pick up a little snack - a nice fat cow - or perhaps a nice fat farmer.

now doesn't that sound like a cool world to play in? sure does to me. and yet there's no mention of quests, large worlds, exploration, storyline - any of that. why? cause the fact is you don't really need any of that stuff to make a cool D&D (IE RPG) game.

i really should do a D&D game. nothing that's been done yet (that i'm aware of) really does the original core game justice.

i remember reading an article where the reviewer said the same thing about skyrim dragons. and they're right, epic dragon battles become nuisance encounters. i think its because of where and how they spawn.

Great insight! I'll have to ponder that some, but I intuitively feel you are on to something there.

as of last night, while playing new vegas, i came to the conclusion that most "quests" in games these days are not really "quests", they are merely "errands"

True dat. Game designers have repurposed the word 'quest' to mean 'errand' instead of 'adventure'.

And then to create "big quests", designers just chain together a few dozen errands.

To some extent, it may be more interesting and enjoyable for a quest to simply be, "Kill the necromancer", and the necromancer being on the other side of the world where it takes you five hours of gameplay to actually travel there and kill the guy, with the 'quest' part being all the emergent adventuring you naturally do in-between, growing on the way, instead of a pre-scripted zig-zagging of quest nodes.

Lord of the Rings, the Frodo's quest had three steps:
1) Escape from The Shire and make contact with Strider
2) Reach Rivendell and make contact with the Elrond and Gandalf
3) Toss the ring into Mount Mordor

The adventuring is what happened in-between.

>> Great insight! I'll have to ponder that some, but I intuitively feel you are on to something there.

i had 6 year experince as a classic edition D&D Dungeon Master before i got into making PC games.

like modern PC games, it has a "go anywhere - do anything - at anytime" play style.

when designing the game, its a little different than just writing a script for a play. you're writing/designing a scene: "dragon encounter scene". but you don't know when, or from what direction, or under what circumstances the player will partake in that scene. so your design must anticipate and handle all possible cases, much the way you must anticipate and handle all cases when writing AI or quests. in skyrim, its simply a case of a poor job of DM'ing.

i'm starting to come to the conclusion that games use folks who are essentially artists for many things they are not trained for. artist types commonly create outdoor levels, yet they know nothing of geography or geology, resulting in the unrealistic mountains you see in games. others here have commented on this in the past. and they are not architects either - leading to the whacked out building designs which are so common. maze level design is cognitive baggage from wof3d 1.0 we still have yet to shake off in favor of more believable environments. many level designers are artists, not DMs (IE game designers). but some are true DMs. a level done by a good DM won't seem like a level at all. and it definitely would not have dragon encounters that could be in any way construed as being "a mere nuisance". <g>.

>> To some extent, it may be more interesting and enjoyable for a quest to simply be, "Kill the necromancer", and the necromancer being on the other side of the world where it takes you five hours of gameplay to actually travel there and kill the guy, with the 'quest' part being all the emergent adventuring you naturally do in-between, growing on the way, instead of a pre-scripted zig-zagging of quest nodes.

not just to some extent - that is THE way to do it.

i'm coming to realize its all about the journey, not the destination.

>> Lord of the Rings, the Frodo's quest had three steps:

1) Escape from The Shire and make contact with Strider
2) Reach Rivendell and make contact with the Elrond and Gandalf
3) Toss the ring into Mount Mordor

actually, i think it was even less that than that:

quest: destroy the one ring.

step 1: meet gandalf at the inn outside the shire after he's talked to Saruman (and we all know how well that went <g>).

so right away they were SNAFU, and from there on it was all "emergent" winging it, until new plans were made at the council of Elrond. so then they started following a quest marker again until they ran into the orcs, and then it was all emergent winging it again, until the next chance to make a plan. which of course would only last 5 minutes before something changed again.

in the end, you'll have 2 kinds of players: shooter, and RPG.

shooter players evolved from 2d platform players. wolf3d v1.0 was an attempt to make a commander keen platformer with a realtime 1st person perspective. shooter players are looking to complete levels, and follow quest/mission markers/waypoints.

RPG players evolved from players of D&D, traveller, GURPs, and similar RPG games. they are all about long term play and becoming more powerful over time.

there are more shooter players than RPG players. so popular RPGs, in an effort to expand their user base, often add shooter element to broaden the appeal, and will even reduce the RPG elements to increase appeal to the shooter fans at the cost of the rpg fans (ahem, bethesda!).

you know what they say, "you can please all the people some of the time...."

how about this one:

"embrace the epic, and eschew the mundane."

all i have to say, if i was a team lead, the first time someone came to me with a 10 skeever hides quest, i'd say, "you obviously have no f*cking clue about what the h*ll you're supposed to be doing, get the f*ck out - RIGHT NOW!" and i'd make sure everyone heard me - so there'd be no more of such nonsense.

but then again, i'm a hard core perfectionist, and find that the standards of performance in most workplaces are not up to my personal standards.

you know the saying about "hard to soar like an eagle when you're flying with turkeys".

back when i worked for the air force, about two weeks after i started, my boss' boss pulled me aside and said "we love the work you're doing but you need to slow down". i responded "i only have one speed". from then on they started calling me "stormin' norman" and this was BEFORE Schwarzkopf! <g>.

i've actually quit jobs in the past cause i couldn't handle the incompetence at the place.

i'm coming to the conclusion that it takes more that just a big seamless level (IE an "open world") to make a good game.

As a gamer who chaffed at 90s postage-stamp sized levels, I absolutely love that the tech has made it possible to have almost open worlds in so many games (Borderlands for me is a perfect example of this, hitting that sweet spot between open & too open). But it's clear that filling those worlds is becoming a bigger and increasingly expensive problem.

a big seamless level with a few hundred hard coded non-respawning spawn points is nothing more that a giant doom level. shoot all the badguys and then its just an architectural walk through.

adding 100 hard coded non-replayable quests only delays the inevitable.

adding hard coded non-replayable crafting of player homes (elder scrolls), villages (farycry 5 primal), or estates (witcher 3) also just delays the inevitable.

Agree, especially because the world itself has no purpose for the player once all the fetching and monster killing is done. Then homicidal depression kicks in as the world reveals how meaningless it is without a hard, final closure to cap everything and justify all the time you've spent (you've rescued the prince and saved the lands or whatever).

Have you ever noticed in contrast how genius Minecraft is in this respect? The world itself is the reason for your existence in it, and its many initial permutations and ways it can be reconfigured only enhances that existence. I would love to see something like this in a different context.

respawns, random encounters, quest generators, house design (a la the sims), and replacemnt of important NPCs (such as dead merchants killed by dragons in skyrim) all seem to be called for.

Rogue-likes have done a much better job here than other genres I think. I dream in this direction often but have been over the years frequently struck by how HIDEOUSLY complex all the moving parts are depending on the level of quality the game is aiming for. I once very naively thought that with a robust procedural framework you could knock out the artist and was deeply dismayed to find that there's a marked difference between art and GOOD art. And beyond that lies good, MEANINGFUL art (something Minecraft can't broach-- e.g., of all the rivers, why is *THIS* river special? What makes it stand out from other rivers? Is it the Mississippi or the Rhine? It doesn't even have a name!)

they finally add random encounters in skyrim and what is it? you encounter a dragon in town while hauling loot back from the dungeon to the store! wait a sec there guys - isn't that dragon supposed to be at the bottom of the dungeon, not waiting for you back at town once you get out? <g>.

One aspect of proc-gen that I don't feel gets enough love is the whole meta reason for content existing. Yes, maybe you can generate a nice, twisty level filled with spawn points that you use to populate in stuff, but WHY is that stuff there? Why is some stuff grouped with other stuff? Can the stuff tell a story? Is that story coherent? Thinking of a Doom promo I've seen, you'd want things generated to say to the player, "Oh, this guy went over here and was obviously trying to open this door when he got killed (which must be why there's a corpse here) and that's why this key card thing is over there..."

The layers of context you can get into here are fascinating.

To some extent, it may be more interesting and enjoyable for a quest to simply be, "Kill the necromancer", and the necromancer being on the other side of the world where it takes you five hours of gameplay to actually travel there and kill the guy, with the 'quest' part being all the emergent adventuring you naturally do in-between, growing on the way, instead of a pre-scripted zig-zagging of quest nodes.

Lord of the Rings, the Frodo's quest had three steps:
1) Escape from The Shire and make contact with Strider
2) Reach Rivendell and make contact with the Elrond and Gandalf
3) Toss the ring into Mount Mordor

The adventuring is what happened in-between.

I like these a lot more and they are more rare, I think in part because it's much harder to ensure quality. If the player does something to glitch the system or even something that's really clever which exposes unexpected deficiencies in the level / challenge design or (oh so common) stupidity in the AI and somehow ends up just plunking the ring into the lava they're going to be disappointed by the colossally anti-climactic experience. I think games have become overrun with rote, unimaginative chores and obedient instruction following not only because it sells but because it's easier to program and QA for. This becomes even more true if there are supposed to be multiple strategies for accomplishing the same thing (a non-combat and combat path, for example).

It's not impossible, but I do think the iterative testing and refining required to make it good is one reason why we don't see it as often as the alternative.

>> But it's clear that filling those worlds is becoming a bigger and increasingly expensive problem.

recently i've been noticing how disrespectful of player's time video games are compared to tabletop. in D&D, if my players said "we leave the dungeon", i'd say "ok", make the appropriate random encounter checks, and that would be it. they would not have to physically traverse each foot of the path back out in real time. and if they wanted to explore to the west, i'd do an encounter check, and move them 1 days's travel, telling them anything of interest along the way in case they wanted to stop and check it out.and just like that they would have crossed skyrim 4 times. but i would not make then sit there in real time while empty wilderness unfolds before them.

fpsrpgs have no accelerated time like flight sims to skip ahead to the action. i often use SGTM console commands in elder scrolls and fallout to accelerate time when forced to manually walk somewhere. needless to say, accelerated time is one of the first things i put in Caveman.

>> Agree, especially because the world itself has no purpose for the player once all the fetching and monster killing is done. Then homicidal depression kicks in as the world reveals how meaningless it is without a hard, final closure to cap everything and justify all the time you've spent (you've rescued the prince and saved the lands or whatever).

games that are designed that way might be able to use closure.

what i propose is a game that's not designed that way. a living world with purpose. no fetching, unless its the epic kind. and a never ending supply of action.

from my notes on the subject: "a continual supply of new and compelling content which is of interest to the player, and appropriate for their level."

continual - so you don't run out of content - IE stuff to do.

new and compelling - can't be the "more of the same". can't be "horse armor" - IE new yet worthless content.

of interest to the player - i don't want to buy the parts for a pre-designed house in skyrim - i already have houses and can only sleep in one at a time, and any single container in any of them can hold all my worldly possessions. i want to design my own castle.

appropriate to their level - i already have houses. i want to design my own castle. and raise an army. and attack my neighbors, and have them attack me, and encounter my neighbors in the wilderness and do champion's battle for prisoner and ransom. sometimes they win, i become their "guest prisoner" and pay ransom, sometimes i win, and they are my "guests" for a few days. or perhaps we bag and tag each other's followers, that kind of stuff. and all along there's adventuring to be done: allies need help in epic quests, constant dungeon adventuring to afford the army, the odd dragon menacing the kingdom from time to time, etc.

you start with a shooter (level maps, combat, monsters, npcs). add the rpg elements (stats, skills, classes, experience, inventory items and world objects [like houses], magic, techno, etc).

then instead of designing it with hard coded spawn points that don't respwan, or only some respawn, or respawn infrequently, or only respawn only a couple of badguys, and hard coded quests, you design it with random encounters. random spawn points, and quest generators.

and you don't design it as "ok, they do the quests, clear the spawn points, and that's the game - sounds like a plan, lets do it".

you design it as "they continually become more powerful, and the world renews itself, instead of getting used up by the player.". but this has no closure.

at some point, there's a limit to how powerful is practical. once you're a god, what's left?

character retirement is about the only "solution" i've seen, and it isn't much of one if you ask me.

but the important thing is the player should be willing to and want to retire , at least for a while, and try a different character in the same world - a different play style, or a different role to take on. and here's the important part:

when thy do this, it must be a new experience like the first pay through with the first character. they should not know where points of interest are in the world from previous play throughs, nor where the badguys spawn, nor when, nor where the good treasure is, none of that.

IE the game should be designed for replay-ability, not for one shot disposable consumption, like a novel or a movie.

and attention needs to be paid to long term play. bayguys can't kill off all the merchants over time (skyrim). in general the world simulation must exhibit long term stability. this is the critical design criteria which is not even considered in most fpsrpg titles.

the arms race aspect of player level/playing time and new content is the only part for which there seems no solution.

>> Have you ever noticed in contrast how genius Minecraft is in this respect?

believe it or not, i have yet to try it first hand. although i've read much about it.

at its core, its virtual design software - in this case landscaping etc. similar to the appeal of house design in the sims.

the brilliance of this is that you can't use up design software, as it not content you consume, its a tool you use to modify content.

you can no more "use it up" than you can "use up" blender or autocad.

and then they tie the ability to use the tool to gameplay rewards - IE simoleans gets you house parts. so you play the game to use the tool, which enhances further gameplay (hopefully you improved your house), in a never ending cycle. the sims breaks down when the game runs out of content. it takes much longer til you've exhausted the capabilities of the house design tool.

but editing a world and adventuring in it are not the same thing. the same way that exploring is not the same thing as action.

i feel a number of games have emphasized a number of gameplay elements such as exploration and quests, while somewhat neglecting the core aspect which apparently now has suffered so much its now derisively referred to as "grinding".

they seem to be too conditioned to getting that level up reward - when it seems that its really about the journey (the grind) and not the destination (retirement). i'm not really sure, as i've never been tempted to even try something that one might typically associate with grinding and fetch quests (IE MMOs). i have no need for PvP or co-op, which is about all an MMO offers over a single player FPSRPG. and who wants to hassle with connecting just to play single player?

>> Rogue-likes have done a much better job here than other genres I think. I dream in this direction often but have been over the years frequently struck by how HIDEOUSLY complex all the moving parts are depending on the level of quality the game is aiming for. I once very naively thought that with a robust procedural framework you could knock out the artist and was deeply dismayed to find that there's a marked difference between art and GOOD art. And beyond that lies good, MEANINGFUL art (something Minecraft can't broach-- e.g., of all the rivers, why is *THIS* river special? What makes it stand out from other rivers? Is it the Mississippi or the Rhine? It doesn't even have a name!)

i don' quite follow. its a bit different than level and spawn point type design. probably more complex, or at least more design ramifications to consider, even if implementation is simple. but i don't really see it increasing the amount of artwork. what's in the game is what's in the game, and you have to draw everything in the game. you're not really changing whats in the game, just where and when and how it comes and goes.

>> One aspect of proc-gen that I don't feel gets enough love is the whole meta reason for content existing. Yes, maybe you can generate a nice, twisty level filled with spawn points that you use to populate in stuff, but WHY is that stuff there? Why is some stuff grouped with other stuff? Can the stuff tell a story? Is that story coherent? Thinking of a Doom promo I've seen, you'd want things generated to say to the player, "Oh, this guy went over here and was obviously trying to open this door when he got killed (which must be why there's a corpse here) and that's why this key card thing is over there..." The layers of context you can get into here are fascinating.

that's just lazy / bad procedural content generation. good procedural content generation should be nigh on indistinguishable from hand rolled. it also depends on how much "window dressing" the game has. window dressing content that paints a picture and tells a story is nice, but often unnecessary. if its cool the first time and unnoticed or bypassed in subsequent play throughs, odds are its unnecessary. by the same token, randomly scattered stuff may be unnecessary too.

this is another example of things that video games do differently than table top. in D&D, i would say, "its a room, 20x20, misc junk scattered about, nothing of value, doors to the north and east. what do you do?" the junk in the room is largely irrelevant to gameplay. i could spice it up a bit with some backstory: "the junk includes a shattered piece of marble used as a fire hearth, bits of charcoal and food crumbs lie nearby. there are broken iron spikes near the north door, and some bloodied rags in the SW corner. it appears a group of adventurers were holed up here, with a fire, and a wounded party member. the north door show signs of failed attempts at spiking it shut. blood smears on the ground form a trail leading through the north door." my players would be like "that' s nice - looks like badguys to the north. we go north!". or perhaps they're all wounded to some degree, laden with treasure, low on spells and potions, and at a deeper dungeon level than they really ought to be, and think twice about that north door. only then does the "window dressing " become important - but it might just save your life then. but with things like monsters that level with player level, and all dungeons accessible at a given time are of the same difficulty (IE all monsters are of about same level as player) you can't really do that. so since every door is of relative equal threat, there's no use for window dressing to indicate one as being a more or less safe path. backstory for backstory's sake and immersion is about the only other use for "clutter", hand placed or otherwise. fallout has an interesting mechanic in that clutter is also essentially gold pieces lying on the ground to be picked up. but it exposes another video game insult to the player's time, making them pick up their treasure one gold piece at a time in real time. definite pain in the A.

>> I like these a lot more and they are more rare, I think in part because it's much harder to ensure quality. If the player does something to glitch the system or even something that's really clever which exposes unexpected deficiencies in the level / challenge design or (oh so common) stupidity in the AI and somehow ends up just plunking the ring into the lava they're going to be disappointed by the colossally anti-climactic experience. I think games have become overrun with rote, unimaginative chores and obedient instruction following not only because it sells but because it's easier to program and QA for. This becomes even more true if there are supposed to be multiple strategies for accomplishing the same thing (a non-combat and combat path, for example). It's not impossible, but I do think the iterative testing and refining required to make it good is one reason why we don't see it as often as the alternative.

i'm not sure that a string of smaller steps in a quest line vs one overall goal is easier to make bulletproof. like any bulletproof code, quest code simply has to handle all possible cases. i suppose it could be considered a divide and conquer strategy, we confine the player at each step to a handfull of possible outcomes, to guarantee which cases are the possible cases to ultimately handle at the end of the quest line - and thus handle all possible cases by limiting the possibilities. this could reduce the amount of forethought and anticipation needed. quest code and AI code are very similar. you wan to both to be bulletproof (handle all cases). so for both you must anticipate all possible cases to be handled.

its definitely more work to implement a bunch of little quests rather than just a main final goal quest. and more work to test. but it is one way to try to cover all cases (by reducing the possible cases to cover). sad thing is, even doing so, may game shops still fail, even when resorting to things like unkillable questors.

needless to say, by reducing the number of possible cases to handle at the end, they also reduce the number of ways the quest can be completed.

a better approach might be to give the player the ultimate goal quest, and have the intermediate quests just out there in the game world as side quests that just happen to aid in your overall goal. any required steps the plauer would be made ware of, or would discover over time. much of this sort of thing means the player needs the ability to ask people in the game world about things, like where can i find person x, location y, or item z. as one of my players used to say "i don't know where the dungeon is? Ok... Person in the street - which way to the dungeon?"