The problem with human beings is that they are capable of the most horrific cruelty or will inflict every type of nasty without rules. So AI is better for me as it is NOT ruled by ego or the feeling of wanting to grind people down as far as it can go. I don't trust humans not to be assholes.
You are wise to be wary, my friend. I enjoy a lot of different types of games, I do not enjoy ruining someone's day just because I can, however, I have learned to be like that, I enjoy PvP for a lot of the same reasons being expressed here I do not enjoy just walking in and getting an easy win. Sadly, I am a tiny minority inside a small fraction of gamers.
But, back to the point, this is why a dynamic AI would make the PvE so much more fun. Instead of: Ok, when his rage meter is full, Healer, throw your big HoTS, ranged make sure you take up the Dps, and all physical except Tank, fall Back.
It will be: Hey, we have no idea how this thing is going to react, just play your best, Good luck!
The problem with the latter, is that there never becomes an Easy Win, and PvE players don't want that, they directly want to learn the mechanics and then be able to farm the content for loot, because let's be honest, it is the loot that moves them.
As far as PvP players go, the fact that hack and cheat programs is a multi-million dollar industry speaks for itself about that demographic.
The thing is, with an Dynamic AI, is gearing it to groups of players, in which case it either needs to be a static power level, Much like a Chess AI program, where anyone above the skill level will know how to win, and anyone below will lose.
Now in a single player game, the AI can adjust itself to the player, so that it can gauge the skill/ability level of the player (Like a Chess Match) and then establish what level it needs to be at to give a challenging game to each individual player.
In a MMO, this is a lot more difficult, as in a group there will be people of various skill levels. This is in fact something that many Dev's talk about when they try to generate content. The disparity between the skill and damage output between hard core players, average players, and casual players, can be massive multipliers.
Just to use GW2, since that has a very flat power progression (That a lot people like to pick on being to easy/casual for players to get top end gear) and you would think that this would even things out, where hardcore players would not seriously outsrip the average players, but the GW2 Devs have openly said that Hardcore players (With the same gear as casuals) are putting out as much as 10x the damage that Average players can generate, just due to how they know how to game the system.
Now imagine an AI trying to work that out.
Now Imagine an AI trying to work that out in a group of mixed skill level players.
Now understand why if we asked them to do that.. they would just kill us all.
Egotism is the anesthetic that dullens the pain of stupidity, this is why when I try to beat my head against the stupidity of other people, I only hurt myself.
First, humans have yet to create artificial intelligence. We've created incredibly complex systems that can beat humans in specific circumstances and that can learn within specific parameters, but it's not artificial intelligence. Not yet, and not any time soon as far as I'm aware.
What we do have is complicated scripting that has the potential to deliver improved games.
Procedural generation is simply a type of scripting that makes use of the way computers generate random numbers. I think it can be an incredibly powerful tool, but it's hard to get it to work well and really depends what you are applying it to.
To give you an MMO example: SWG. The worlds of SWG were actually generated using procedural generation (the main terrain). Once the devs had gotten the procedural generation to work, they locked down the "seeds" for those worlds, then started manually adding the cities and POIs. The reason SWG did this was to save space. To hand-craft the planets, before even adding the cities, would have required something like 2.5GB just to sort the height maps and stuff for a single planet. That meant the game was simply too massive to ship if it was hand-crafted.
By switching to procedural generation, all they need to do was create the assets (which they had to do anyway) and then write the rules of how to generate a planet. Writing those rules is difficult and requires a lot of tweaking, but they did it. Whilst it is harder than hand-crafting, it is also cheaper and quicker.
So, I fully believe in the power of procedural generation, I just don't think we've seen it employed particularly well as of yet. No Man's Sky did a really good job of employing procedural generation for their planets, as in the quality was good. The main downside seemed to be that on a world that's generated randomly from the same set of assets, the results looked too similar. That's not a problem with procedural generation, it's a problem with the rules of procedural generation.
What I would like to see from procedural generation and scripting in an RPG:
1) Worlds generated by procedural generation.
With a wide enough set of assets and good rules, you can generate an infinite number of worlds to inhabit. You can generate a snowy world like Skyrim, or a European world like The Witcher, or something more high fantasy like Final Fantasy. Mountains, rivers, seas, swamps, forests, jungles, plains, deserts.....we are already at a position to do this technically, just no1 has done it yet.
2) Cities, town, villages and roads generated by procedural generation
I'm not sure whether this has been tried yet, but population centres in the real world do develop over time based on various unwritten rules, like access to resources, defensible features, separation of business and accommodation. Once you've figured out the rules for appointing roads and plots, generating the houses should be easy.
3) Cultures and NPCs also generated by procedural generation
This should probably be the easiest bit. Randomly creating names, gods etc should be pretty straight forwards. You would need to get the rules correct in terms of distrubution of jobs amoungst the NPCs, e.g. a city might need 5 blacksmiths, whilst a village might only have 0-1. This needs to tie in closely with generating the city - no point generating 20 smiths if there are only 2 forges...
The procedural generation, in effect, is just step 1: building the game world for us to inhabit. If done well, this can provide you with infinite game worlds to play in. If you allow modding, then the community can make it even more diverse and interesting. I would also want to tie in crafting resources as part of the world generation. Perhaps in world-001, steel is a rare resource and so the economies develop more advanced wood-based construction. Perhaps in world-002, there is an abundance of magic and so the cultures and economy becomes based on magic. Maybe in world-003, there is no magic, or wood, so it becomes a world of metal and leather.
When it comes to actually playing, I just want some advanced scripting designed to give us a simulated world. Each NPC should have it's own personality (assigned when generated), and everyone should live out their simulated lives. If a band of bandits is successful, then the game should make them setup a proper base in the countryside. If news of those bandits reaches the town, then if the town has a good guard, they should head out to kill them. Alternatively, this is what generates quests for the players - stuff happening in the simulated world generates quests. If rabbits are eating a farmers crops, then the farmer can give u a quest to kill those rabbits. If the town guard is expanding, it can issue a quest to craft 50 new swords and shields.
Is that level of scripting possible?
Yes. The tech is already here and not too difficult to write. The issue is that whilst the rules and scripts can be easily written, the consequences are hard to work out. If every NPC is simulated and lives in the world, if it is not perfectly balanced then it's possible that a bandit camp might take over the whole world within 50 days. Or if the ruler of a large city is given the personality of a dictator, they may well run that city into the ground and make that particular world very miserable to play in. Or if you don't generate enough NPC farmers, the local townsmen might starve to death.
The real world has plenty of naturally built-in self-balancing mechanisms, but in a virtual world the developers need to build them in. With so many moving parts in a simulated world, it is incredibly difficult to get it balanced properly. To make it fun for the player, you ideally want to program the world so that it always heads towards a sense of equilibrium. That way, the player's actions are more noticable and will have a greater effect on the world, but it also stops the worlds becoming unplayable.
In reading some of Raph Koster's work, I believe this was how UO was originally designed and built. However, as it was ground-breaking work, they didn't really understand how it would all work. So, they ended up in situations where NPC wolves killed everything else, or rabbits that had had enough success in game to become more powerful than players. Before they were given the opportunity to figure it out and make it work, the higher ups told them to remove those systems, so they did.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
The problem with human beings is that they are capable of the most horrific cruelty or will inflict every type of nasty without rules. So AI is better for me as it is NOT ruled by ego or the feeling of wanting to grind people down as far as it can go. I don't trust humans not to be assholes.
You are wise to be wary, my friend. I enjoy a lot of different types of games, I do not enjoy ruining someone's day just because I can, however, I have learned to be like that, I enjoy PvP for a lot of the same reasons being expressed here I do not enjoy just walking in and getting an easy win. Sadly, I am a tiny minority inside a small fraction of gamers.
But, back to the point, this is why a dynamic AI would make the PvE so much more fun. Instead of: Ok, when his rage meter is full, Healer, throw your big HoTS, ranged make sure you take up the Dps, and all physical except Tank, fall Back.
It will be: Hey, we have no idea how this thing is going to react, just play your best, Good luck!
The problem with the latter, is that there never becomes an Easy Win, and PvE players don't want that, they directly want to learn the mechanics and then be able to farm the content for loot, because let's be honest, it is the loot that moves them.
As far as PvP players go, the fact that hack and cheat programs is a multi-million dollar industry speaks for itself about that demographic.
The thing is, with an Dynamic AI, is gearing it to groups of players, in which case it either needs to be a static power level, Much like a Chess AI program, where anyone above the skill level will know how to win, and anyone below will lose.
Now in a single player game, the AI can adjust itself to the player, so that it can gauge the skill/ability level of the player (Like a Chess Match) and then establish what level it needs to be at to give a challenging game to each individual player.
In a MMO, this is a lot more difficult, as in a group there will be people of various skill levels. This is in fact something that many Dev's talk about when they try to generate content. The disparity between the skill and damage output between hard core players, average players, and casual players, can be massive multipliers.
Just to use GW2, since that has a very flat power progression (That a lot people like to pick on being to easy/casual for players to get top end gear) and you would think that this would even things out, where hardcore players would not seriously outsrip the average players, but the GW2 Devs have openly said that Hardcore players (With the same gear as casuals) are putting out as much as 10x the damage that Average players can generate, just due to how they know how to game the system.
Now imagine an AI trying to work that out.
Now Imagine an AI trying to work that out in a group of mixed skill level players.
Now understand why if we asked them to do that.. they would just kill us all.
lol it would be better if you didn't quote my posts.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
That is essentially the idea I like, the AI being able to recognize and adjust to a party composition essentially "on the fly" amongst other things.
If you want a new idea, go read an old book.
In order to be insulted, I must first value your opinion.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is where things get tricky, this idea was started by someone called Mark Glickman, and he designed something called the Glicko Scale, this was used for Chess games, so that players could have their number upon the scale, and face other players of close to or equal numbers to ensure challenging matches were played, and in chess tournaments you didn't have grandmasters playing against noobs. This was also a means to eliminate things like a "Lucky Game" or just having a "Bad Day" as the scale was built across many games.
The thing here, is while a wide range of difficulties could exist, like the various levels of Chess, with an AI, you are asking the computer to pre-select the difficulty.
Now for a single player game, this stupid easy for the AI to handle, it just gauges your prior performance, based against the varying difficulties it tossed at you as you played, it knows where you sit on the difficulty scale and can always provide a good game for you to play, exactly like Computer would handle a Chess Opponent.
It becomes a mess when you have people of varying skills, if they are close, then it is not a problem, but the more diverse the individual players the harder it is for the AI to make a good encounter. I would wager it would be exactly like it would be for a Human DM handling a D&D game, to design an encounter for their group, and needing to handle both the general luck of the group and dealing with a wide breadth of players from Just learning Noob to Ultra Munchkin Hardcore.
Thankfully in MMO's cliques form, so this becomes less a problem, but you still end up say a situation where a Veteran player opts to help out a new player, the AI is now faced with the challenge of trying to balance the encounter to be fun for them both, and we all know how this ends, the new player ends up as paste against the wall in the first fight.
Which is where the real issue of AI starts, think of how hard it is for a Human to balance an encounter against varying skill levels, even in small groups like friendly D&D games, which we have been dealing with since the 70's. I think it would be a lot harder than it sounds as you add more players to the mix.
Egotism is the anesthetic that dullens the pain of stupidity, this is why when I try to beat my head against the stupidity of other people, I only hurt myself.
Now understand why if we asked them to do that.. they would just kill us all.
lol it would be better if you didn't quote my posts.
That is not a problem to stop quoting and responding to you. Consider it Done.
Egotism is the anesthetic that dullens the pain of stupidity, this is why when I try to beat my head against the stupidity of other people, I only hurt myself.
To give you an MMO example: SWG. The worlds of SWG were actually generated using procedural generation (the main terrain). Once the devs had gotten the procedural generation to work, they locked down the "seeds" for those worlds, then started manually adding the cities and POIs. The reason SWG did this was to save space. To hand-craft the planets, before even adding the cities, would have required something like 2.5GB just to sort the height maps and stuff for a single planet. That meant the game was simply too massive to ship if it was hand-crafted.
You seemed to tallk about world which is randomly generated. For some reason I thought the OP is talking about how the world is constantly changing and how the asset would behave on its own.
For example you could have an orc settlement. And it is destroyed by centaur and become a centaur den. NPC would be doing different things every day, have different dialogue every day. Or assign procedure generated quest.
Ungood said: The problem with the latter, is that there never becomes an Easy Win, and PvE players don't want that, they directly want to learn the mechanics and then be able to farm the content for loot, because let's be honest, it is the loot that moves them.
As far as PvP players go, the fact that hack and cheat programs is a multi-million dollar industry speaks for itself about that demographic.
The thing is, with an Dynamic AI, is gearing it to groups of players, in which case it either needs to be a static power level, Much like a Chess AI program, where anyone above the skill level will know how to win, and anyone below will lose.
Now in a single player game, the AI can adjust itself to the player, so that it can gauge the skill/ability level of the player (Like a Chess Match) and then establish what level it needs to be at to give a challenging game to each individual player.
In a MMO, this is a lot more difficult, as in a group there will be people of various skill levels. This is in fact something that many Dev's talk about when they try to generate content. The disparity between the skill and damage output between hard core players, average players, and casual players, can be massive multipliers.
Just to use GW2, since that has a very flat power progression (That a lot people like to pick on being to easy/casual for players to get top end gear) and you would think that this would even things out, where hardcore players would not seriously outsrip the average players, but the GW2 Devs have openly said that Hardcore players (With the same gear as casuals) are putting out as much as 10x the damage that Average players can generate, just due to how they know how to game the system.
Now imagine an AI trying to work that out.
Now Imagine an AI trying to work that out in a group of mixed skill level players.
Now understand why if we asked them to do that.. they would just kill us all.
Playing Overwatch vs AI isn’t super terrible if we are talking pure combat. And thats probably on the low side of good AI combat. It has a lot of untapped potential. I think combat would be one of the easiest things to work out concerning AI, but I agree it would most likely try the tolerance of one side or the other, referring to the hardcore vs. casual, but there are three difficulty levels to make up for that. Having the missions remain intriguing using AI would be much more difficult, imo, but also more worthwhile.
Post edited by Palebane on
Vault-Tec analysts have concluded that the odds of worldwide nuclear armaggeddon this decade are 17,143,762... to 1.
I'm curious how some of you think computers work if you think that this or that isn't "real" AI. All they do is follow some set formulas. As best as we can tell, that's generally how the universe works, too, at least up to some randomness on a quantum scale. For all we know, that might even be how human intelligence works--and perhaps even with much effect from the weird, quantum randomness.
Some people are more impressed with AI if the formulas are complicated enough that no one understands why it gives the answers that it does. I'd argue that you shouldn't be. The famous acronym is KISS: Keep It Simple, Stupid. When a black box that you don't understand kicks out results that are obviously wrong, that is not fun to debug.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
The idea behind what I was saying is you don't get the same fight every time. I really dislike the fact that fights can be memorized, like a dance (as someone else pointed out) and then it's just following your script to victory. The only challenge is getting everyone on the same script (and I admit, in current MMORPG climate that is a true challenge) and then dancing to victory. I just think it would be cool to see a Boss or even a regular monster not behave the same way every time.
Again, just my opinion.
If you want a new idea, go read an old book.
In order to be insulted, I must first value your opinion.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
This reminds me of Anthem, where the players discovered that the Rewards and Challenges were based on a players prior performance, Basically it was all "AI" set, custom for the individual, so they would feign incompetence just to stomp some later challenge to get max rewards.
Egotism is the anesthetic that dullens the pain of stupidity, this is why when I try to beat my head against the stupidity of other people, I only hurt myself.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
The idea behind what I was saying is you don't get the same fight every time. I really dislike the fact that fights can be memorized, like a dance (as someone else pointed out) and then it's just following your script to victory. The only challenge is getting everyone on the same script (and I admit, in current MMORPG climate that is a true challenge) and then dancing to victory. I just think it would be cool to see a Boss or even a regular monster not behave the same way every time.
Again, just my opinion.
Exactly. In the system I showed, lets say a group of players is against a group of Orcs, in a dungeon room. The Orcs have their own AI, but it's also supplemented by the "Dungeon AI" (or whatever you want to call it), which is more "choices based on what's in that room for them to use. In other words, the Orcs have "choices" on a list, and a weighted die roll. Weighted by the current circumstances.
So at first the Orcs are doing normal stuff, most likely. Sword attacks, arrows, maybe some spells if they have them, maybe throwing some Explosion Bombs. Each round, choices are made for each Orc, but as a unit in the AI. It may end up all doing the same thing, an even mix, or any combination.
As the battler progresses, die rolls determine what the Orcs do, but it's weighted on the current circumstances. If they've taken a lot of damage, they may heal themselves (removing attacks), or they may not, or something in the middle.
Lets say that the Orcs have taken losses and they are now outnumbered 2 to 1, and there are no dead Players. Under this system, it's highly likely that the Orcs will roll "Flight", and try to flee the battle. The Dungeon AI system may have provided them a choice to head down a corridor deeper into the Dungeon, but in an organized retreat. Or it may not, maybe they roll any of the unlikely fighting choices, or just break and run out any direction (individually).
Now, the Players may have blocked their escape, or not. The Orcs may retreat to another room where there's more Orcs, taking less damage because of their organized retreat as they go. Maybe along the way to that other room, they lose half their numbers and roll to just break and retreat, to that second room because it's the only way left that's open.
Once they get to that second room, anything can happen based on this system, as they've combined with the Orcs AI that is there. Or maybe they don't make it at all. And maybe the Orcs in the second room become aware of the fighting and start to react, or maybe not if it was a way away and there was some reason the sound didn't carry (like maybe a Player cast a Silence spell).
On the other hand, maybe none of this happens because the Orcs chose to do something early on that decimated the Player group (just the right combination of actions, unlikely but possible), and the battle went the other way. The Player retreat, hopefully organized, and make their escape out of the Dungeon and into the forest to lick their wounds and come up with a plan that might work.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
This reminds me of Anthem, where the players discovered that the Rewards and Challenges were based on a players prior performance, Basically it was all "AI" set, custom for the individual, so they would feign incompetence just to stomp some later challenge to get max rewards.
It's not based on prior experiences. It's on the spot, current circumstances, round by round. Why the hell even assume something like that? Especially when I never said anything close to that.
At least that was one of the more original Negative Nancy approaches to shooting something down.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
This reminds me of Anthem, where the players discovered that the Rewards and Challenges were based on a players prior performance, Basically it was all "AI" set, custom for the individual, so they would feign incompetence just to stomp some later challenge to get max rewards.
It's not based on prior experiences. It's on the spot, current circumstances, round by round. Why the hell even assume something like that? Especially when I never said anything close to that.
At least that was one of the more original Negative Nancy approaches to shooting something down.
Good Hell Dude, Chill the Fuck Out.. I was only commenting on your point about the idea of game world being made to order feeling gamey, much like what happend with Anthem.
Egotism is the anesthetic that dullens the pain of stupidity, this is why when I try to beat my head against the stupidity of other people, I only hurt myself.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
This reminds me of Anthem, where the players discovered that the Rewards and Challenges were based on a players prior performance, Basically it was all "AI" set, custom for the individual, so they would feign incompetence just to stomp some later challenge to get max rewards.
It's not based on prior experiences. It's on the spot, current circumstances, round by round. Why the hell even assume something like that? Especially when I never said anything close to that.
At least that was one of the more original Negative Nancy approaches to shooting something down.
Good Hell Dude, Chill the Fuck Out.. I was only commenting on your point about the idea of game world being made to order feeling gamey, much like what happend with Anthem.
Well, it made no sense at all. Now, that could be an actual strategy in a particular battle. Sort of lulling the enemy into an action, such as charging the player group, a trap of sorts? Maybe the players use weaker weapons and an organized retreat, and setting a trap for the Orcs. That would be cool strategy stuff. But that's not what you said.
And why should I chill, when every god damned time I start talking about new ideas I get this kind of negativity, throwing shjt at the walls to see what might stick in the minds of other readers?
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
This reminds me of Anthem, where the players discovered that the Rewards and Challenges were based on a players prior performance, Basically it was all "AI" set, custom for the individual, so they would feign incompetence just to stomp some later challenge to get max rewards.
It's not based on prior experiences. It's on the spot, current circumstances, round by round. Why the hell even assume something like that? Especially when I never said anything close to that.
At least that was one of the more original Negative Nancy approaches to shooting something down.
This reminds me of rubber-banding. Racing games employ it successfully to slow the enemies down if you're behind or speed them up if you're ahead.
It's good idea for driving games, but for your standard MMO combat it would be awful. In hard fights the players would intentionally suck at the start so that they'd get some bonus for being behind, then blow all their cooldowns at the end.
Or if it's done in instance-level the players would intentionally suck the easy fights to make the hard fights easier. The players will game this kind of systems.
First, humans have yet to create artificial intelligence. We've created incredibly complex systems that can beat humans in specific circumstances and that can learn within specific parameters, but it's not artificial intelligence. Not yet, and not any time soon as far as I'm aware.
What we do have is complicated scripting that has the potential to deliver improved games.
Procedural generation is simply a type of scripting that makes use of the way computers generate random numbers. I think it can be an incredibly powerful tool, but it's hard to get it to work well and really depends what you are applying it to.
To give you an MMO example: SWG. The worlds of SWG were actually generated using procedural generation (the main terrain). Once the devs had gotten the procedural generation to work, they locked down the "seeds" for those worlds, then started manually adding the cities and POIs. The reason SWG did this was to save space. To hand-craft the planets, before even adding the cities, would have required something like 2.5GB just to sort the height maps and stuff for a single planet. That meant the game was simply too massive to ship if it was hand-crafted.
By switching to procedural generation, all they need to do was create the assets (which they had to do anyway) and then write the rules of how to generate a planet. Writing those rules is difficult and requires a lot of tweaking, but they did it. Whilst it is harder than hand-crafting, it is also cheaper and quicker.
So, I fully believe in the power of procedural generation, I just don't think we've seen it employed particularly well as of yet. No Man's Sky did a really good job of employing procedural generation for their planets, as in the quality was good. The main downside seemed to be that on a world that's generated randomly from the same set of assets, the results looked too similar. That's not a problem with procedural generation, it's a problem with the rules of procedural generation.
What I would like to see from procedural generation and scripting in an RPG:
1) Worlds generated by procedural generation.
With a wide enough set of assets and good rules, you can generate an infinite number of worlds to inhabit. You can generate a snowy world like Skyrim, or a European world like The Witcher, or something more high fantasy like Final Fantasy. Mountains, rivers, seas, swamps, forests, jungles, plains, deserts.....we are already at a position to do this technically, just no1 has done it yet.
2) Cities, town, villages and roads generated by procedural generation
I'm not sure whether this has been tried yet, but population centres in the real world do develop over time based on various unwritten rules, like access to resources, defensible features, separation of business and accommodation. Once you've figured out the rules for appointing roads and plots, generating the houses should be easy.
3) Cultures and NPCs also generated by procedural generation
This should probably be the easiest bit. Randomly creating names, gods etc should be pretty straight forwards. You would need to get the rules correct in terms of distrubution of jobs amoungst the NPCs, e.g. a city might need 5 blacksmiths, whilst a village might only have 0-1. This needs to tie in closely with generating the city - no point generating 20 smiths if there are only 2 forges...
The procedural generation, in effect, is just step 1: building the game world for us to inhabit. If done well, this can provide you with infinite game worlds to play in. If you allow modding, then the community can make it even more diverse and interesting. I would also want to tie in crafting resources as part of the world generation. Perhaps in world-001, steel is a rare resource and so the economies develop more advanced wood-based construction. Perhaps in world-002, there is an abundance of magic and so the cultures and economy becomes based on magic. Maybe in world-003, there is no magic, or wood, so it becomes a world of metal and leather.
When it comes to actually playing, I just want some advanced scripting designed to give us a simulated world. Each NPC should have it's own personality (assigned when generated), and everyone should live out their simulated lives. If a band of bandits is successful, then the game should make them setup a proper base in the countryside. If news of those bandits reaches the town, then if the town has a good guard, they should head out to kill them. Alternatively, this is what generates quests for the players - stuff happening in the simulated world generates quests. If rabbits are eating a farmers crops, then the farmer can give u a quest to kill those rabbits. If the town guard is expanding, it can issue a quest to craft 50 new swords and shields.
Is that level of scripting possible?
Yes. The tech is already here and not too difficult to write. The issue is that whilst the rules and scripts can be easily written, the consequences are hard to work out. If every NPC is simulated and lives in the world, if it is not perfectly balanced then it's possible that a bandit camp might take over the whole world within 50 days. Or if the ruler of a large city is given the personality of a dictator, they may well run that city into the ground and make that particular world very miserable to play in. Or if you don't generate enough NPC farmers, the local townsmen might starve to death.
The real world has plenty of naturally built-in self-balancing mechanisms, but in a virtual world the developers need to build them in. With so many moving parts in a simulated world, it is incredibly difficult to get it balanced properly. To make it fun for the player, you ideally want to program the world so that it always heads towards a sense of equilibrium. That way, the player's actions are more noticable and will have a greater effect on the world, but it also stops the worlds becoming unplayable.
In reading some of Raph Koster's work, I believe this was how UO was originally designed and built. However, as it was ground-breaking work, they didn't really understand how it would all work. So, they ended up in situations where NPC wolves killed everything else, or rabbits that had had enough success in game to become more powerful than players. Before they were given the opportunity to figure it out and make it work, the higher ups told them to remove those systems, so they did.
Yeah this is a good explanation of my thinking. People always complain about how generic it could be but it has to start somewhere. It can improve with time which is another advantage. Not to mention the quest we have now as filler is already generic.
I like the idea of having problems and solving them. It was also have to change the ideas of we play. Being local to an area instead of constantly move feeding on content as a vagabond.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
This reminds me of Anthem, where the players discovered that the Rewards and Challenges were based on a players prior performance, Basically it was all "AI" set, custom for the individual, so they would feign incompetence just to stomp some later challenge to get max rewards.
It's not based on prior experiences. It's on the spot, current circumstances, round by round. Why the hell even assume something like that? Especially when I never said anything close to that.
At least that was one of the more original Negative Nancy approaches to shooting something down.
This reminds me of rubber-banding. Racing games employ it successfully to slow the enemies down if you're behind or speed them up if you're ahead.
It's good idea for driving games, but for your standard MMO combat it would be awful. In hard fights the players would intentionally suck at the start so that they'd get some bonus for being behind, then blow all their cooldowns at the end.
Or if it's done in instance-level the players would intentionally suck the easy fights to make the hard fights easier. The players will game this kind of systems.
Bingo, and On Point, and this is exactly what happened with Anthem once players learned that things were being custom set to their individual skill level by an "AI GM" program, they feigned incompetence to get better rewards and game the system.
Great idea on paper, but, it does not work so well against the reckless abandon of players.
Egotism is the anesthetic that dullens the pain of stupidity, this is why when I try to beat my head against the stupidity of other people, I only hurt myself.
As to using AI / advanced scripting to control enemy difficulty, there's nothing to say you can't program the AI to provide the players with fun. Just because an AI has the capability to always play at the top of its ability, doesn't mean it has to. If the AI's goal is to provide fun and a challenge, it could easily provide us with a wide range of difficulties and modify the enemies to each player's ability.
This is pretty much what I'm after in the AI system I described on page 2. It has randomness but based on the NPCs own nature and abilities. Each battle would be different against the same enemy of like group. And they can make mistakes that players can take advantage of, or do something players didn't expect. I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present." I want to know what I've beaten or lost to, and be able to judge my progress based on that. Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
This reminds me of Anthem, where the players discovered that the Rewards and Challenges were based on a players prior performance, Basically it was all "AI" set, custom for the individual, so they would feign incompetence just to stomp some later challenge to get max rewards.
It's not based on prior experiences. It's on the spot, current circumstances, round by round. Why the hell even assume something like that? Especially when I never said anything close to that.
At least that was one of the more original Negative Nancy approaches to shooting something down.
This reminds me of rubber-banding. Racing games employ it successfully to slow the enemies down if you're behind or speed them up if you're ahead.
It's good idea for driving games, but for your standard MMO combat it would be awful. In hard fights the players would intentionally suck at the start so that they'd get some bonus for being behind, then blow all their cooldowns at the end.
Or if it's done in instance-level the players would intentionally suck the easy fights to make the hard fights easier. The players will game this kind of systems.
Bingo, and On Point, and this is exactly what happened with Anthem once players learned that things were being custom set to their individual skill level by an "AI GM" program, they feigned incompetence to get better rewards and game the system.
Great idea on paper, but, it does not work so well against the reckless abandon of players.
Well it is not abused by players,no system is abused by players if it is as intended or done well.How often do we see design issues where it looks like no testing or thought at all was put into them?I mean if players are so easily finding the issues/exploits what have the dev teams been doing,do they have any quality control,any testing,do they listen to feedback?
Anthem is a poor example for anything because that was a lazy game design,a lazy effort,just a big business operation using it's clout to promote a crap game,similar to promoting a game that is tagged with a brand,like Star Wars or Warcraft or Final Fantasy.
Game design over the past 10 years seems to be more about an assembly line ordeal,toss in onto the belt and get er done as fast as possible with no care for it's quality of design systems.
It is as simple as the AI system being done well.ANY form of scaling is garbage so that is not a good idea for an AI system.I think i mentioned it already,but yeah it would take a bit of effort put into each database but not really that difficult,just too much usually for the lazy developers i have been witnessing. @above poster,let's hear YOUR ideas rather than insult others...afraid or don't have any design ideas?No problem disagreeing with others ideas,that should be expected but to just say BS and not offer anything at all is very lame and leans towards you not actually having anything yourself to offer.
Never forget 3 mile Island and never trust a government official or company spokesman.
Well it is not abused by players,no system is abused by players if it is as intended or done well.How often do we see design issues where it looks like no testing or thought at all was put into them?I mean if players are so easily finding the issues/exploits what have the dev teams been doing,do they have any quality control,any testing,do they listen to feedback?
Anthem is a poor example for anything because that was a lazy game design,a lazy effort,just a big business operation using it's clout to promote a crap game,similar to promoting a game that is tagged with a brand,like Star Wars or Warcraft or Final Fantasy.
Game design over the past 10 years seems to be more about an assembly line ordeal,toss in onto the belt and get er done as fast as possible with no care for it's quality of design systems.
It is as simple as the AI system being done well.ANY form of scaling is garbage so that is not a good idea for an AI system.I think i mentioned it already,but yeah it would take a bit of effort put into each database but not really that difficult,just too much usually for the lazy developers i have been witnessing. @above poster,let's hear YOUR ideas rather than insult others...afraid or don't have any design ideas?No problem disagreeing with others ideas,that should be expected but to just say BS and not offer anything at all is very lame and leans towards you not actually having anything yourself to offer.
Since it's random, and on a round by round basis, it would be hard for players to carry out some system to abuse it. But it's actually a point that they can try. That's tactics. But it's not predictable and there's no set outcome for the MOB reactions.
Well it is not abused by players,no system is abused by players if it is as intended or done well.How often do we see design issues where it looks like no testing or thought at all was put into them?I mean if players are so easily finding the issues/exploits what have the dev teams been doing,do they have any quality control,any testing,do they listen to feedback?
Anthem is a poor example for anything because that was a lazy game design,a lazy effort,just a big business operation using it's clout to promote a crap game,similar to promoting a game that is tagged with a brand,like Star Wars or Warcraft or Final Fantasy.
Game design over the past 10 years seems to be more about an assembly line ordeal,toss in onto the belt and get er done as fast as possible with no care for it's quality of design systems.
It is as simple as the AI system being done well.ANY form of scaling is garbage so that is not a good idea for an AI system.I think i mentioned it already,but yeah it would take a bit of effort put into each database but not really that difficult,just too much usually for the lazy developers i have been witnessing. @above poster,let's hear YOUR ideas rather than insult others...afraid or don't have any design ideas?No problem disagreeing with others ideas,that should be expected but to just say BS and not offer anything at all is very lame and leans towards you not actually having anything yourself to offer.
I agree that Anthem was not a great game, but it was a good example of trying these kinds of systems where there is some AI trying to scale the encounters to the players, and it did a great job of showcasing some of the glaring problems that exist in those kinds of ideas, and how players respond to many game ideas and how willing they are to try and game the systems.
I believe the line is "The best laid plans"
Egotism is the anesthetic that dullens the pain of stupidity, this is why when I try to beat my head against the stupidity of other people, I only hurt myself.
Well it is not abused by players,no system is abused by players if it is as intended or done well.How often do we see design issues where it looks like no testing or thought at all was put into them?I mean if players are so easily finding the issues/exploits what have the dev teams been doing,do they have any quality control,any testing,do they listen to feedback?
Anthem is a poor example for anything because that was a lazy game design,a lazy effort,just a big business operation using it's clout to promote a crap game,similar to promoting a game that is tagged with a brand,like Star Wars or Warcraft or Final Fantasy.
Game design over the past 10 years seems to be more about an assembly line ordeal,toss in onto the belt and get er done as fast as possible with no care for it's quality of design systems.
It is as simple as the AI system being done well.ANY form of scaling is garbage so that is not a good idea for an AI system.I think i mentioned it already,but yeah it would take a bit of effort put into each database but not really that difficult,just too much usually for the lazy developers i have been witnessing. @above poster,let's hear YOUR ideas rather than insult others...afraid or don't have any design ideas?No problem disagreeing with others ideas,that should be expected but to just say BS and not offer anything at all is very lame and leans towards you not actually having anything yourself to offer.
I agree that Anthem was not a great game, but it was a good example of trying these kinds of systems where there is some AI trying to scale the encounters to the players, and it did a great job of showcasing some of the glaring problems that exist in those kinds of ideas, and how players respond to many game ideas and how willing they are to try and game the systems.
I believe the line is "The best laid plans"
That it can be done badly hardly means that it cannot be done well.
For scaling difficulty, I prefer giving players a difficulty slider and letting them put it wherever they want. If they set it too high or too low, it's their own fault.
Quizzical said: For scaling difficulty, I prefer giving players a difficulty slider and letting them put it wherever they want. If they set it too high or too low, it's their own fault.
I agree with you. There are many players who really can’t handle that responsibility though. For them, gaming the system is the goal, the thing to overcome. Sliders make that a generic experience. For me it’s ideal because some days I want more of a challenge and some days I am more laid back, but I don’t necessarily want to do separate or different content/areas.
Games that don't use sliders often cause me to get creative and drop gear/lower stats or use less ideal setups when I want a challenge, which is another form of gaming the system, really, but is also another thing the above type player seems highly adverse to.
Vault-Tec analysts have concluded that the odds of worldwide nuclear armaggeddon this decade are 17,143,762... to 1.
Comments
As far as PvP players go, the fact that hack and cheat programs is a multi-million dollar industry speaks for itself about that demographic.
The thing is, with an Dynamic AI, is gearing it to groups of players, in which case it either needs to be a static power level, Much like a Chess AI program, where anyone above the skill level will know how to win, and anyone below will lose.
Now in a single player game, the AI can adjust itself to the player, so that it can gauge the skill/ability level of the player (Like a Chess Match) and then establish what level it needs to be at to give a challenging game to each individual player.
In a MMO, this is a lot more difficult, as in a group there will be people of various skill levels. This is in fact something that many Dev's talk about when they try to generate content. The disparity between the skill and damage output between hard core players, average players, and casual players, can be massive multipliers.
Just to use GW2, since that has a very flat power progression (That a lot people like to pick on being to easy/casual for players to get top end gear) and you would think that this would even things out, where hardcore players would not seriously outsrip the average players, but the GW2 Devs have openly said that Hardcore players (With the same gear as casuals) are putting out as much as 10x the damage that Average players can generate, just due to how they know how to game the system.
Now imagine an AI trying to work that out.
Now Imagine an AI trying to work that out in a group of mixed skill level players.
Now understand why if we asked them to do that.. they would just kill us all.
That is essentially the idea I like, the AI being able to recognize and adjust to a party composition essentially "on the fly" amongst other things.
If you want a new idea, go read an old book.
In order to be insulted, I must first value your opinion.
The thing here, is while a wide range of difficulties could exist, like the various levels of Chess, with an AI, you are asking the computer to pre-select the difficulty.
Now for a single player game, this stupid easy for the AI to handle, it just gauges your prior performance, based against the varying difficulties it tossed at you as you played, it knows where you sit on the difficulty scale and can always provide a good game for you to play, exactly like Computer would handle a Chess Opponent.
It becomes a mess when you have people of varying skills, if they are close, then it is not a problem, but the more diverse the individual players the harder it is for the AI to make a good encounter. I would wager it would be exactly like it would be for a Human DM handling a D&D game, to design an encounter for their group, and needing to handle both the general luck of the group and dealing with a wide breadth of players from Just learning Noob to Ultra Munchkin Hardcore.
Thankfully in MMO's cliques form, so this becomes less a problem, but you still end up say a situation where a Veteran player opts to help out a new player, the AI is now faced with the challenge of trying to balance the encounter to be fun for them both, and we all know how this ends, the new player ends up as paste against the wall in the first fight.
Which is where the real issue of AI starts, think of how hard it is for a Human to balance an encounter against varying skill levels, even in small groups like friendly D&D games, which we have been dealing with since the 70's. I think it would be a lot harder than it sounds as you add more players to the mix.
For example you could have an orc settlement. And it is destroyed by centaur and become a centaur den. NPC would be doing different things every day, have different dialogue every day. Or assign procedure generated quest.
Vault-Tec analysts have concluded that the odds of worldwide nuclear armaggeddon this decade are 17,143,762... to 1.
Some people are more impressed with AI if the formulas are complicated enough that no one understands why it gives the answers that it does. I'd argue that you shouldn't be. The famous acronym is KISS: Keep It Simple, Stupid. When a black box that you don't understand kicks out results that are obviously wrong, that is not fun to debug.
I laid out the foundation as the basis, only, and I fear that you all didn't realize the scope of what can be done with it.
One point of contention though, I really don't like the idea of modifying the enemies to fit the players. I want the enemies I face to be within a factor, known or unknown "at present."
I want to know what I've beaten or lost to, and be able to judge my progress based on that.
Otherwise, it's a world that's "made to order", and feels gamey rather than interesting.
Once upon a time....
Again, just my opinion.
If you want a new idea, go read an old book.
In order to be insulted, I must first value your opinion.
The Orcs have their own AI, but it's also supplemented by the "Dungeon AI" (or whatever you want to call it), which is more "choices based on what's in that room for them to use.
In other words, the Orcs have "choices" on a list, and a weighted die roll. Weighted by the current circumstances.
So at first the Orcs are doing normal stuff, most likely. Sword attacks, arrows, maybe some spells if they have them, maybe throwing some Explosion Bombs. Each round, choices are made for each Orc, but as a unit in the AI. It may end up all doing the same thing, an even mix, or any combination.
As the battler progresses, die rolls determine what the Orcs do, but it's weighted on the current circumstances. If they've taken a lot of damage, they may heal themselves (removing attacks), or they may not, or something in the middle.
Lets say that the Orcs have taken losses and they are now outnumbered 2 to 1, and there are no dead Players. Under this system, it's highly likely that the Orcs will roll "Flight", and try to flee the battle. The Dungeon AI system may have provided them a choice to head down a corridor deeper into the Dungeon, but in an organized retreat. Or it may not, maybe they roll any of the unlikely fighting choices, or just break and run out any direction (individually).
Now, the Players may have blocked their escape, or not.
The Orcs may retreat to another room where there's more Orcs, taking less damage because of their organized retreat as they go. Maybe along the way to that other room, they lose half their numbers and roll to just break and retreat, to that second room because it's the only way left that's open.
Once they get to that second room, anything can happen based on this system, as they've combined with the Orcs AI that is there. Or maybe they don't make it at all. And maybe the Orcs in the second room become aware of the fighting and start to react, or maybe not if it was a way away and there was some reason the sound didn't carry (like maybe a Player cast a Silence spell).
On the other hand, maybe none of this happens because the Orcs chose to do something early on that decimated the Player group (just the right combination of actions, unlikely but possible), and the battle went the other way. The Player retreat, hopefully organized, and make their escape out of the Dungeon and into the forest to lick their wounds and come up with a plan that might work.
Once upon a time....
Why the hell even assume something like that? Especially when I never said anything close to that.
At least that was one of the more original Negative Nancy approaches to shooting something down.
Once upon a time....
Now, that could be an actual strategy in a particular battle. Sort of lulling the enemy into an action, such as charging the player group, a trap of sorts? Maybe the players use weaker weapons and an organized retreat, and setting a trap for the Orcs.
That would be cool strategy stuff.
But that's not what you said.
And why should I chill, when every god damned time I start talking about new ideas I get this kind of negativity, throwing shjt at the walls to see what might stick in the minds of other readers?
Once upon a time....
It's good idea for driving games, but for your standard MMO combat it would be awful. In hard fights the players would intentionally suck at the start so that they'd get some bonus for being behind, then blow all their cooldowns at the end.
Or if it's done in instance-level the players would intentionally suck the easy fights to make the hard fights easier. The players will game this kind of systems.
I like the idea of having problems and solving them. It was also have to change the ideas of we play. Being local to an area instead of constantly move feeding on content as a vagabond.
Great idea on paper, but, it does not work so well against the reckless abandon of players.
Just more of the same BS from some of you.
Once upon a time....
Anthem is a poor example for anything because that was a lazy game design,a lazy effort,just a big business operation using it's clout to promote a crap game,similar to promoting a game that is tagged with a brand,like Star Wars or Warcraft or Final Fantasy.
Game design over the past 10 years seems to be more about an assembly line ordeal,toss in onto the belt and get er done as fast as possible with no care for it's quality of design systems.
It is as simple as the AI system being done well.ANY form of scaling is garbage so that is not a good idea for an AI system.I think i mentioned it already,but yeah it would take a bit of effort put into each database but not really that difficult,just too much usually for the lazy developers i have been witnessing.
@above poster,let's hear YOUR ideas rather than insult others...afraid or don't have any design ideas?No problem disagreeing with others ideas,that should be expected but to just say BS and not offer anything at all is very lame and leans towards you not actually having anything yourself to offer.
Never forget 3 mile Island and never trust a government official or company spokesman.
But it's actually a point that they can try. That's tactics. But it's not predictable and there's no set outcome for the MOB reactions.
Once upon a time....
I believe the line is "The best laid plans"
For scaling difficulty, I prefer giving players a difficulty slider and letting them put it wherever they want. If they set it too high or too low, it's their own fault.
Vault-Tec analysts have concluded that the odds of worldwide nuclear armaggeddon this decade are 17,143,762... to 1.