Thats true, loot is the superficial need, but the real goal of a game is enjoyable game play, waaay to much emphasis on the cheap fix of throwing loot at a game misses this point alltogether. The best instances I have ever played have had a mixture of mobs and events which gives lots of scope for using all the skills you have frequently - fun and satisfaction ffrom playing your char. Old wow instances are like that, GW1 instances, and to a lesser degree GW2 instances. If you run through an instance and just spam a button and consider it meaningless then the instance failed.
Re randomisation, as poster suggested above, it can be possible to have an instance that is carefully designed and feels unique but has randomised elements, expensive to develope but rewarding.
rpg/mmorg history: Dun Darach>Bloodwych>Bards Tale 1-3>Eye of the beholder > Might and Magic 2,3,5 > FFVII> Baldur's Gate 1, 2 > Planescape Torment >Morrowind > WOW > oblivion > LOTR > Guild Wars (1900hrs elementalist) Vanguard. > GW2(1000 elementalist), Wildstar
The problem is one of quality. They look like random dungeons.
^^THIS. Obviously random map generation is essential to 4x games like Civilization, but they are merely ok in tradtional dungeon crawlers like the Diablo series. They take on a samey-samey quality and it becomes a game of "find the door". "Oh look, it's over here in this map."
Of course I'm pretty down on instancing in general, so I'm biased, but if I had to have instances, I'd rather have them well-thought out, creative and semi-logical rather than slight variations of chaos.
Who says that randomly generated zones have to be instanced? Who says you can't have a thousand (or a million!) randomly generated zones linked together to make an enormous open world?
You could, however it wouldn't work in a 'real' MMO as it would just seem like a crappy world compared to a planned one. As a result, many of the advocates of it would probably dismiss it once it was in an MMO that strayed from what they consider right and proper design.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein "Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
The problem is one of quality. They look like random dungeons.
^^THIS. Obviously random map generation is essential to 4x games like Civilization, but they are merely ok in tradtional dungeon crawlers like the Diablo series. They take on a samey-samey quality and it becomes a game of "find the door". "Oh look, it's over here in this map."
Of course I'm pretty down on instancing in general, so I'm biased, but if I had to have instances, I'd rather have them well-thought out, creative and semi-logical rather than slight variations of chaos.
Who says that randomly generated zones have to be instanced? Who says you can't have a thousand (or a million!) randomly generated zones linked together to make an enormous open world?
You could, however it wouldn't work in a 'real' MMO as it would just seem like a crappy world compared to a planned one. As a result, many of the advocates of it would probably dismiss it once it was in an MMO that strayed from what they consider right and proper design.
What's not an MMO about that? You'd stick every single player in the entire game in a single enormous world, without having multiple instances of any area of the game. If you get more players, you add more zones to make a bigger world yet. You could make it completely seamless, too. If that's not "massively multiplayer", then there is no such thing. And it's trivially online.
The problem is one of quality. They look like random dungeons.
^^THIS. Obviously random map generation is essential to 4x games like Civilization, but they are merely ok in tradtional dungeon crawlers like the Diablo series. They take on a samey-samey quality and it becomes a game of "find the door". "Oh look, it's over here in this map."
Of course I'm pretty down on instancing in general, so I'm biased, but if I had to have instances, I'd rather have them well-thought out, creative and semi-logical rather than slight variations of chaos.
Who says that randomly generated zones have to be instanced? Who says you can't have a thousand (or a million!) randomly generated zones linked together to make an enormous open world?
You could, however it wouldn't work in a 'real' MMO as it would just seem like a crappy world compared to a planned one. As a result, many of the advocates of it would probably dismiss it once it was in an MMO that strayed from what they consider right and proper design.
What's not an MMO about that? You'd stick every single player in the entire game in a single enormous world, without having multiple instances of any area of the game. If you get more players, you add more zones to make a bigger world yet. You could make it completely seamless, too. If that's not "massively multiplayer", then there is no such thing. And it's trivially online.
I didn't say it isn't an MMO or that it isn't massively multiplayer. I said your idea would work in an MMO, but not in the standard EQ/WOW style MMO. In a scripted world MMORPG, a randomly generated zone would be rather lackluster or even incomplete when compared to a dev-crafted, scripted zone of planned content created for a specific purpose.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein "Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
Quantity has a quality all its own. Which is the point of randomization, after all.
But to repeat what I said earlier, it's not a matter of you randomize things or else you don't. Rather, it's a question of what exactly you decide to randomize and how you decide to randomize it.
Really, though, the question isn't so much randomized versus not randomized. Rather, game designers should constantly be thinking, can I randomize this little piece of the game without making it worse? Sometimes the answer is "no". But the answer should be "yes" a lot more often than most game designers are willing to do it.
Everyone's job looks easy to someone who doesn't know what the job entails.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein "Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
Don't forget that I'm working on not merely randomly generating a game world, but randomly generating all of the artwork for it, too. So naturally, I think that a lot more things should be randomized than most people. But not loot; random loot is stupid.
Everything in the universe is random; it's only a question of random with what probability distribution. This outcome with probability 1 and all others with probability 0 is a perfectly good probability distribution. I think that trying for it is done too much, though.
Incidentally, tessellation in its purest form is procedurally generated vertex data. While it can be used for the geometrically intuitive approach of subdivisions of simplicial complexes, that's hardly the only possible use. If you want to do procedurally generated anything that the GPU needs to be aware of, tessellation is likely to make it a whole lot easier.
Don't forget that I'm working on not merely randomly generating a game world, but randomly generating all of the artwork for it, too.
I don't doubt you can create a random zone and use a lot of jargon when talking about it.
I am simply addressing your various statements on the matter:
"Who says that randomly generated zones have to be instanced? Who says you can't have a thousand (or a million!) randomly generated zones linked together to make an enormous open world?"
"You'd stick every single player in the entire game in a single enormous world, without having multiple instances of any area of the game. If you get more players, you add more zones to make a bigger world yet."
"But the answer should be "yes" a lot more often than most game designers are willing to do it."
My response is that you could probably do that for something other than your typical EQ/WOW style MMO. In those games, what you suggest, would either create a lackluster environment, an environment with little relevance, or an environment that would require so much post work that it would be the same as crafting it manually to begin with.
I'm speaking from my experiences working directly with team members that create game worlds for a living. It's not that limitless worlds of amazingly procedurally generated awesomeness don't exist because of some kind of dev resistance to doing it. There's reasons that devs don't do it, especially in massively multiplayer RPGs where the absence or deficiency of narrative, scene management and a host of other aspects (pathing, traffic, fidelity, etc) become very apparent very quickly.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein "Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
I'm not arguing that game developers should usually make randomly generated zones for an entire zone. While I'd like to see that done more than it is, I wouldn't want to see every game do that for its entire game world.
But little things like randomizing a mob's spawn point slightly (e.g., this mob will spawn at a random point within 30 feet of here) or randomly picking which mob will spawn out of several possible can help a lot. Some games already do those things to some degree--and entirely within the context of a traditional MMORPG--but I'd like to see them done more.
^^THIS. Obviously random map generation is essential to 4x games like Civilization, but they are merely ok in tradtional dungeon crawlers like the Diablo series. They take on a samey-samey quality and it becomes a game of "find the door". "Oh look, it's over here in this map."
Actually in D3, the interesting part of the randomization is not so much of the random layout of the dungeon, but the randomly generated elite/champions.
If they roll the "right" abilities, the fight can be extremely challenging, even if you have good gear. And the nice thing is ... different combinations of abilities are challenging to different classes.
You're assuming that the randomization will be done in a particular way. That is one way to do it, but it's hardly the only way.
Who says that you can only have a few options for a boss? If you've got 50 different boss skills available, then picking 5 at random for a particular boss to have gives you over 2 million possible combinations right there. If you want to restrict things to one from this category, one from that category, and so forth, you could still easily have many thousands of combinations.
Boss stats give you even more randomization possible. You could set the sum of a boss's attributes and pick the attributes randomly such that they total that sum. Or you could fix the sum of the squares of the attributes or whatever.
If you do only have a handful of particular random bosses available, then who says every boss has to give the same loot? You could track which bosses usually get killed versus which ones usually get skipped, and give dramatically better loot for the latter.
Nobody said anything about "only a few options". D3 has 24 traits and probably about as many monster types, which have variations on the core stats (damage/health) and inherit a random combination of the traits (molten, jailor, etc). I'm sure it results in far more than 2 million combinations, but the strategies written for D3 only have to worry about describing those ~50 core pieces.
Honestly D3 shows how easily randomized content can break in two completely different directions regarding strategy websites. On one hand, once you've written a strategic overview of the 50 components, you've written an overview for the entire game. On the other hand, the tactics you'd learn from that page will only matter when you're facing mobs right at the Balance Sweet Spot: fights where player skill is the difference between success and failure. But due largely to monster randomization (and also loot randomization), you're almost never in the Balance Sweet Spot and skill is largely irrelevant because you're either going to faceroll the mob or be utterly destroyed (or have a very long and tedious fight.)
Which is sort of a shame, because if there wasn't any progression to Diablo 3 and you were always in the sweet spot, the monster design and some of the traits are actually really well done. But they virtually never get a chance to play out as intended.
As for the dynamic loot thing, I thought of that too but avoided mentioning it. It could indeed work. But it's a very dangerous decision. Basically you make a design decision to implement a complicated and hard-to-balance system (randomization) and then commit to a second complicated and hard-to-balance system (dynamic loot adjustment) just to try to get things right.
"What is truly revealing is his implication that believing something to be true is the same as it being true. [continue]" -John Oliver
still waiting my DM v.2.0 ,maybe Ultima Online 2 can do it.
The problem with the AO missions is that there were only a couple of mission types and the setpieces were so few that every "random" mission looked virtually the same as every other one. It wasn't all that random either, there were standard mission builds which rarely deviated all that much, plus the mobs were pretty much predictable, it was the same mobs over and over and over and over.
Plus the fact, to get the highest tokenboards, you had to run over 1000 missions, which was an absurd grind.
If start a mmo that have random dungeons it would be very strange who change them some unknown force a godlike person who have power to completely change dungeosn would be awkward i say and unrealistic the least.
Make dungeons so that if you kill mobs inside there gone forever but other mobs or npcs can populate this dungeon see it as there new home.
Make dynamic system where npc's and mobs roam arround can also change game world build things.
So when you next time enter a dungeon suddenly totally different npc's or mobs took as there new habitat.
Improve AI make npc's and mobs more creative so they also can change enviroment to there liking.
Real Dynamic game world with superb AI is what we need.
Sounds good but I've never played a random dungeon so maybe I would hate it??? I thought PvP would liven up a game once. Had no idea it would ruin every game I touched from forum pleas to constantly change character stats so every time I logged in I had no idea what my char was able to do (usually much less than it could do before). My guess is any dungeon is great if I can get others to come along. I've played many games where we all just gave up on the dungeons. Usually because they're so dependent on the trinity which is what makes them so boring and hard to party up for.
Players want some meaningful story, so some parts of a MMO need to be handcrafted, but especially those instanced dungeons every MMO now has, or most, could be improved by some degree of randomization. At least that's my 5 cents.
People don't ask questions to get answers - they ask questions to show how smart they are. - Dogbert
Players want some meaningful story, so some parts of a MMO need to be handcrafted, but especially those instanced dungeons every MMO now has, or most, could be improved by some degree of randomization. At least that's my 5 cents.
The problems arise when you start thinking just how you're going to implement that randomization.
Take traps, for example, how do you determine where to put them so that they make sense and players wont have to walk with a minesweeper all the time? Or do you want them to act like they're in a minefield all the time? And if you want to make the trap placement make sense, players will identify those possible locations very quickly.
It all sounds fine on paper, but the implementation is where all the problems come in.
Same with bosses, as Axehilt wrote, either it is a number or a set of random traits or the bosses are premade. It is pretty much either or, and player will again write walkthroughs how to beat a boss with such and such set of traits.
In the end you gain very little if anything. In my mind, I find it hard to justify the extra effort that goes into it.
I skate to where the puck is going to be, not where it has been-Wayne Gretzky
Honestly D3 shows how easily randomized content can break in two completely different directions regarding strategy websites. On one hand, once you've written a strategic overview of the 50 components, you've written an overview for the entire game. On the other hand, the tactics you'd learn from that page will only matter when you're facing mobs right at the Balance Sweet Spot: fights where player skill is the difference between success and failure. But due largely to monster randomization (and also loot randomization), you're almost never in the Balance Sweet Spot and skill is largely irrelevant because you're either going to faceroll the mob or be utterly destroyed (or have a very long and tedious fight.)
Which is sort of a shame, because if there wasn't any progression to Diablo 3 and you were always in the sweet spot, the monster design and some of the traits are actually really well done. But they virtually never get a chance to play out as intended.
This is an interesting discussion & interesting view. I think the important point is this:
You don't need to be in teh sweet spot in every encounter to have fun. In fact, many strategy guide will help you maximize your "coverage" ... robust against many types, and try to minimize the bad elite rolls that will kill you. It is not an unfun experience when you face-roll most mobs, and suddenly you have to run for life. That kind of random challenge keep one on the toes, and at the same time, keep one feeling powerful.
Let me give you a concrete example. I do many Act 3 keep 2 legendary runs on MP8. (I can do mp10 but that won't be efficient). To made it efficient, i intentionally go more glass canon on my wiz, and use glass canon, and shock armor (instead of energy). I kill normal mobs very fast, and most elites ok. Occasionally i have to run (RD mobs!) and fight smart to kill them. Very occasionally, they kill me in 2 sec (vortex me in, caged, fire/acid and also RD).
I like this set up. The situation has strategic depth while i can feel poweful. Most of the time, i mow down things, but i have to keep on my toes, because one wrong affix, and one wrong move, i can be death. Secondly, i can tune the challenge vs efficiencies to my liking. When i feel like it (and i have), i turned off glass canon, use prismatic energy armor ... my dps goes down a lot, but survivability goes up. So i go to MP10, and kill elites, slowly and carefully, to show that i can do it.
That kind of strategic choice to fit my mood (want a big challnege, or want farming fast with the occasinoally death, or want to farm safely) is very fun to me. That is why i keep varying my build according to what i want to do in a particular day.
the FPS game i'm making also happens to be a roguelike.
only single player for the first release.
sequel/stage-2 would probably have a smallish overworld, and definitely multiplayer.
my life's goal is to eventually turn it into a virtual world, where most things are handcrafted. mostly by players. its voxel based (minecraft style), so its quick and easy to edit things, with an extremely low (actually non-existent) barrier to entry. i'd imagine at least %50 of the players would build as well as play. i intend to build a lot of areas myself, as i've always wanted to, but instead i've put that time into learning/programming. it will be a joy to build things after all these years.
but along with handcrafted dungeons, you'd still have unlimited randomly generated dungeons as an option. and the generated dungeons would be the only "instances" in the game (with options to allow invasions, or late join-in-progress pickup co-op).
thats the reason i'm starting with a roguelike. i want the game to have unlimited content from day one. and atm, i'm only aware of one other FPS roguelike (which is a Doom 3 mod).
Heroes Call on Android has fixed spawn points, but randomizes the spawn groups on each dungeon load.
Even a system as simple as this helps keep the runs fresh.
Ken Fisher - Semi retired old fart Network Administrator, now working in Network Security. I don't Forum PVP. If you feel I've attacked you, it was probably by accident. When I don't understand, I ask. Such is not intended as criticism.
Comments
Thats true, loot is the superficial need, but the real goal of a game is enjoyable game play, waaay to much emphasis on the cheap fix of throwing loot at a game misses this point alltogether. The best instances I have ever played have had a mixture of mobs and events which gives lots of scope for using all the skills you have frequently - fun and satisfaction ffrom playing your char. Old wow instances are like that, GW1 instances, and to a lesser degree GW2 instances. If you run through an instance and just spam a button and consider it meaningless then the instance failed.
Re randomisation, as poster suggested above, it can be possible to have an instance that is carefully designed and feels unique but has randomised elements, expensive to develope but rewarding.
rpg/mmorg history: Dun Darach>Bloodwych>Bards Tale 1-3>Eye of the beholder > Might and Magic 2,3,5 > FFVII> Baldur's Gate 1, 2 > Planescape Torment >Morrowind > WOW > oblivion > LOTR > Guild Wars (1900hrs elementalist) Vanguard. > GW2(1000 elementalist), Wildstar
Now playing GW2, AOW 3, ESO, LOTR, Elite D
You could, however it wouldn't work in a 'real' MMO as it would just seem like a crappy world compared to a planned one. As a result, many of the advocates of it would probably dismiss it once it was in an MMO that strayed from what they consider right and proper design.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein
"Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
What's not an MMO about that? You'd stick every single player in the entire game in a single enormous world, without having multiple instances of any area of the game. If you get more players, you add more zones to make a bigger world yet. You could make it completely seamless, too. If that's not "massively multiplayer", then there is no such thing. And it's trivially online.
I didn't say it isn't an MMO or that it isn't massively multiplayer. I said your idea would work in an MMO, but not in the standard EQ/WOW style MMO. In a scripted world MMORPG, a randomly generated zone would be rather lackluster or even incomplete when compared to a dev-crafted, scripted zone of planned content created for a specific purpose.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein
"Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
Quantity has a quality all its own. Which is the point of randomization, after all.
But to repeat what I said earlier, it's not a matter of you randomize things or else you don't. Rather, it's a question of what exactly you decide to randomize and how you decide to randomize it.
Everyone's job looks easy to someone who doesn't know what the job entails.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein
"Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
Don't forget that I'm working on not merely randomly generating a game world, but randomly generating all of the artwork for it, too. So naturally, I think that a lot more things should be randomized than most people. But not loot; random loot is stupid.
Everything in the universe is random; it's only a question of random with what probability distribution. This outcome with probability 1 and all others with probability 0 is a perfectly good probability distribution. I think that trying for it is done too much, though.
Incidentally, tessellation in its purest form is procedurally generated vertex data. While it can be used for the geometrically intuitive approach of subdivisions of simplicial complexes, that's hardly the only possible use. If you want to do procedurally generated anything that the GPU needs to be aware of, tessellation is likely to make it a whole lot easier.
I don't doubt you can create a random zone and use a lot of jargon when talking about it.
I am simply addressing your various statements on the matter:
"Who says that randomly generated zones have to be instanced? Who says you can't have a thousand (or a million!) randomly generated zones linked together to make an enormous open world?"
"You'd stick every single player in the entire game in a single enormous world, without having multiple instances of any area of the game. If you get more players, you add more zones to make a bigger world yet."
"But the answer should be "yes" a lot more often than most game designers are willing to do it."
My response is that you could probably do that for something other than your typical EQ/WOW style MMO. In those games, what you suggest, would either create a lackluster environment, an environment with little relevance, or an environment that would require so much post work that it would be the same as crafting it manually to begin with.
I'm speaking from my experiences working directly with team members that create game worlds for a living. It's not that limitless worlds of amazingly procedurally generated awesomeness don't exist because of some kind of dev resistance to doing it. There's reasons that devs don't do it, especially in massively multiplayer RPGs where the absence or deficiency of narrative, scene management and a host of other aspects (pathing, traffic, fidelity, etc) become very apparent very quickly.
There isn't a "right" or "wrong" way to play, if you want to use a screwdriver to put nails into wood, have at it, simply don't complain when the guy next to you with the hammer is doing it much better and easier. - Allein
"Graphics are often supplied by Engines that (some) MMORPG's are built in" - Spuffyre
I'm not arguing that game developers should usually make randomly generated zones for an entire zone. While I'd like to see that done more than it is, I wouldn't want to see every game do that for its entire game world.
But little things like randomizing a mob's spawn point slightly (e.g., this mob will spawn at a random point within 30 feet of here) or randomly picking which mob will spawn out of several possible can help a lot. Some games already do those things to some degree--and entirely within the context of a traditional MMORPG--but I'd like to see them done more.
Actually in D3, the interesting part of the randomization is not so much of the random layout of the dungeon, but the randomly generated elite/champions.
If they roll the "right" abilities, the fight can be extremely challenging, even if you have good gear. And the nice thing is ... different combinations of abilities are challenging to different classes.
Nobody said anything about "only a few options". D3 has 24 traits and probably about as many monster types, which have variations on the core stats (damage/health) and inherit a random combination of the traits (molten, jailor, etc). I'm sure it results in far more than 2 million combinations, but the strategies written for D3 only have to worry about describing those ~50 core pieces.
Honestly D3 shows how easily randomized content can break in two completely different directions regarding strategy websites. On one hand, once you've written a strategic overview of the 50 components, you've written an overview for the entire game. On the other hand, the tactics you'd learn from that page will only matter when you're facing mobs right at the Balance Sweet Spot: fights where player skill is the difference between success and failure. But due largely to monster randomization (and also loot randomization), you're almost never in the Balance Sweet Spot and skill is largely irrelevant because you're either going to faceroll the mob or be utterly destroyed (or have a very long and tedious fight.)
Which is sort of a shame, because if there wasn't any progression to Diablo 3 and you were always in the sweet spot, the monster design and some of the traits are actually really well done. But they virtually never get a chance to play out as intended.
As for the dynamic loot thing, I thought of that too but avoided mentioning it. It could indeed work. But it's a very dangerous decision. Basically you make a design decision to implement a complicated and hard-to-balance system (randomization) and then commit to a second complicated and hard-to-balance system (dynamic loot adjustment) just to try to get things right.
"What is truly revealing is his implication that believing something to be true is the same as it being true. [continue]" -John Oliver
The problem with the AO missions is that there were only a couple of mission types and the setpieces were so few that every "random" mission looked virtually the same as every other one. It wasn't all that random either, there were standard mission builds which rarely deviated all that much, plus the mobs were pretty much predictable, it was the same mobs over and over and over and over.
Plus the fact, to get the highest tokenboards, you had to run over 1000 missions, which was an absurd grind.
I'd want something much, much better.
Played: UO, EQ, WoW, DDO, SWG, AO, CoH, EvE, TR, AoC, GW, GA, Aion, Allods, lots more
Relatively Recently (Re)Played: HL2 (all), Halo (PC, all), Batman:AA; AC, ME, BS, DA, FO3, DS, Doom (all), LFD1&2, KOTOR, Portal 1&2, Blink, Elder Scrolls (all), lots more
Now Playing: None
Hope: None
If start a mmo that have random dungeons it would be very strange who change them some unknown force a godlike person who have power to completely change dungeosn would be awkward i say and unrealistic the least.
Make dungeons so that if you kill mobs inside there gone forever but other mobs or npcs can populate this dungeon see it as there new home.
Make dynamic system where npc's and mobs roam arround can also change game world build things.
So when you next time enter a dungeon suddenly totally different npc's or mobs took as there new habitat.
Improve AI make npc's and mobs more creative so they also can change enviroment to there liking.
Real Dynamic game world with superb AI is what we need.
Not simple and boring predictable randomizing.
I hate instancing thus randomization is no option.
Also, handmade >> randomization, anyway.
I guess it is a matter of balance.
Players want some meaningful story, so some parts of a MMO need to be handcrafted, but especially those instanced dungeons every MMO now has, or most, could be improved by some degree of randomization. At least that's my 5 cents.
People don't ask questions to get answers - they ask questions to show how smart they are. - Dogbert
The problems arise when you start thinking just how you're going to implement that randomization.
Take traps, for example, how do you determine where to put them so that they make sense and players wont have to walk with a minesweeper all the time? Or do you want them to act like they're in a minefield all the time? And if you want to make the trap placement make sense, players will identify those possible locations very quickly.
It all sounds fine on paper, but the implementation is where all the problems come in.
Same with bosses, as Axehilt wrote, either it is a number or a set of random traits or the bosses are premade. It is pretty much either or, and player will again write walkthroughs how to beat a boss with such and such set of traits.
In the end you gain very little if anything. In my mind, I find it hard to justify the extra effort that goes into it.
I skate to where the puck is going to be, not where it has been -Wayne Gretzky
This isn't a signature, you just think it is.
This is an interesting discussion & interesting view. I think the important point is this:
You don't need to be in teh sweet spot in every encounter to have fun. In fact, many strategy guide will help you maximize your "coverage" ... robust against many types, and try to minimize the bad elite rolls that will kill you. It is not an unfun experience when you face-roll most mobs, and suddenly you have to run for life. That kind of random challenge keep one on the toes, and at the same time, keep one feeling powerful.
Let me give you a concrete example. I do many Act 3 keep 2 legendary runs on MP8. (I can do mp10 but that won't be efficient). To made it efficient, i intentionally go more glass canon on my wiz, and use glass canon, and shock armor (instead of energy). I kill normal mobs very fast, and most elites ok. Occasionally i have to run (RD mobs!) and fight smart to kill them. Very occasionally, they kill me in 2 sec (vortex me in, caged, fire/acid and also RD).
I like this set up. The situation has strategic depth while i can feel poweful. Most of the time, i mow down things, but i have to keep on my toes, because one wrong affix, and one wrong move, i can be death. Secondly, i can tune the challenge vs efficiencies to my liking. When i feel like it (and i have), i turned off glass canon, use prismatic energy armor ... my dps goes down a lot, but survivability goes up. So i go to MP10, and kill elites, slowly and carefully, to show that i can do it.
That kind of strategic choice to fit my mood (want a big challnege, or want farming fast with the occasinoally death, or want to farm safely) is very fun to me. That is why i keep varying my build according to what i want to do in a particular day.
the FPS game i'm making also happens to be a roguelike.
only single player for the first release.
sequel/stage-2 would probably have a smallish overworld, and definitely multiplayer.
my life's goal is to eventually turn it into a virtual world, where most things are handcrafted. mostly by players. its voxel based (minecraft style), so its quick and easy to edit things, with an extremely low (actually non-existent) barrier to entry. i'd imagine at least %50 of the players would build as well as play. i intend to build a lot of areas myself, as i've always wanted to, but instead i've put that time into learning/programming. it will be a joy to build things after all these years.
but along with handcrafted dungeons, you'd still have unlimited randomly generated dungeons as an option. and the generated dungeons would be the only "instances" in the game (with options to allow invasions, or late join-in-progress pickup co-op).
thats the reason i'm starting with a roguelike. i want the game to have unlimited content from day one. and atm, i'm only aware of one other FPS roguelike (which is a Doom 3 mod).
---------------------------
Corpus Callosum
---------------------------
Heroes Call on Android has fixed spawn points, but randomizes the spawn groups on each dungeon load.
Even a system as simple as this helps keep the runs fresh.