Heh, when I read that blog entry, I thought it was too true. My fiancee keeps asking me why I kill all the cute little "Yodas." Honestly, the only reason I kill them is because they run at me first.

I actually prefer worlds where "traditional" fantasy memes are avoided, eg. the black and white "orcs are evil and elves are good" mentality. I think it's very possible to befriend orcs. Why not? A good example of a storyline that involves non-evil orcs would be the WTactics project's Red Banner Faction. Basically, the orcs became so fed up with oppression, they turned a new leaf and began fighting not to conquer but for equal rights with the humans.

If Flare's main theme is survival then by all means violence is not only required it is necessary since  savage beasts (or lesser races) are motivated only by primal instinct.
As for the other more (or less) intelligent world inhabitants the player should have a saying , aether try to communicate with the orcs blocking the path or let his sword do all the talking.

Different creatures should have different motivations and being a "good guy" and giving some food to a hungry orc or a goblin will not only make game world feel more alive , but perhaps that goblin or orc will reveal some information or put in a good word for you back at the Goblin camp. On the other hand being a "bad guy" by robing some wounded hunter might have negative results but it might help you get noticed by the "right people".Using magic regardless of intention will cause mage hunters to come for the player and when they catch him they might let him go depending on how he used his magic powers.
 The human settlements are small so "breaking the law" should result in exile further more killing another human  should result in death or some kind of irremovable marking.
The penalty  of commuting lesser crimes should  result in a sentence of hard labor (or high risk jobs).

I remember having the option to help orcs or zombies (or alternatively grind them to pulp) in the outstanding game with the ridiculous name Divine Divinity. It was only possible at predefined points in the game as part of a quest but I think there were quite a number of these. It would definitely be cool to move away from the black and white stereotypes common in this genre.

It's bad enough if the game forces you to kill goblins because every goblin has exactly one method of interaction - attack you on sight. It's worse if the game forces you to kill goblins because you need to pick their pockets in order to get the money to become better so you can achieve later missions. It's even worse if the game forces you to kill goblins because you just literally need to kill goblins or you don't get better (XPs). Currently Flare is doing all three, but the last is the most tasteless.

That said, I'm not sure it's a good idea to try to change the fundamental paradigm of the basic adventure. Probably better to build a new one from scratch with these ideas in mind, and leave the basic adventure for those who like that sort of thing.

One way to minimise the killing would be a grind-free version of the adventure. At least then you'd only have to kill a few people in the boss lair or whatever, who the narrative can establish are definitely evil, rather than needing to murder everyone who happens to live in the same county. By "grind-free" I mean an adventure in which XPs and gold are awarded for completing missions, in much larger amounts than usual, and no XPs or gold are awarded for killing monsters or looting barrels, etc.. Perhaps the monsters don't respawn (need to think about that). Any player who found they couldn't progress could go to the mentor-philanthropist and get free XPs and gold - the more you do of this, the lower skill level you are effectively playing on. I'd like to think it was fairly easy to mechanically modify the standard adventure to give one that works like this, am I right?

Ever played games like planetscape torment or baldurs gate? most xp comes from resolving quests but you requiere a heavy focus on story and it get tedious to go around if the story doesnt keep up or is not interesting enought (think on how you can do interesting every "fecht X of XXXX"). That means developments goes heavily into creating the world and quests, making the combat phases less important and secondary; while Diablo and Flare goes into the "hack and slash" genre, lots of action, story is there just to point you to the next thing to kill.

That conversion is doable, but at the risk that is way more easy to mess up and end up with a boring game if you go the story-heavy route, also limits the playerbase, which nowadays is full of one-neuron violence junkies.

That said, the reputation oriented behavior villages could be a good idea, but for more freely oriented games with a persistant world, in which you may choose a few to dedicate yourself and end the game with (gothic games or geneforge saga), while on diablo games its just a hub to get services until you get to the next one in story.

Danimal,

I suspect the reason Flare's playerbase is 1NVJs is that this is the sort of campaign Flare is. With a non-grind option you may be able to broaden the playerbase.

Of course you don't want to alienate your fan base. So by all means keep the grind version, and make the non-grind version an option selectable at game start. A fork would be a really unfortunate thing so it might be best to have both parameter values in the scenario definition and let the game choose which one to use according to a flag the player sets at startup. If that's considered too much overhead for the main campaign to wear, then we could work out rules for mechanical translation of the grinding campaign into a non-grinding one, though that's not trivial.

Well Flare is built to be an Action RPG and nothing more. Combat is the main gameplay and always will be. Story-heavy games require branching dialog and should be built in another engine.

I say this because Flare has always been about small, controlled scope. Otherwise it's almost impossible to actually finish and engine and make a game using it. Flare is already 3.5 years old, and even slight additions in scope threaten to expand the project by years. Consider, for example, that FreeDroid's dialog engine alone is about 10k lines of code -- nearly a third of the size of all of Flare. Flare gets a lot done in a tiny amount of code because of its focused purpose.

Combat is central to the Action RPG genre, but there are certainly ways to make it less violent. Imagine a cartoony game in Flare where you're a wizard kid shooting a bubble wand at tiny monsters, and instead of dying the monsters turn into candy.

Or to make the enemies really deserving of violence. Undead and demons are easy. Natural beasts less so, as they're mostly acting territorially. Sentient humanoids are really questionable unless they're actively part of some larger evil plot (e.g. they're at war, slaughtering humans). I don't like the idea of actual human enemies at all, and I don't think I'm a mature enough writer to have an effective human villain.

With Wandercall I have a story designed that most, if not all, of the enemies are obvious evil. Where they aren't obvious evil, they're metaphysical -- personal demons instead of literal demons.

In paper RPGs, races like goblins/orcs have often been inherently evil - and in that sense, killing them is no more unethical than killing demons.

Another way of looking at it I think is that the races are essentially at war - or alternatively, it is a matter of survival.

I thought it interesting how in Morrowind (and presumably the other Elder Scrolls games), creatures like Orcs were friendly, and the enemies you encountered were either other humans or undead/demonic/magical. Though I don't think it avoids the issue of killing (indeed, in that game, you are killing other humans), just the races behave differently.

In theory, I like the idea of an RPG that has a focus on things other than killing, and isn't simply about killing room after room full of monsters. However I think writing such a thing is far from easy - you need a far larger amount of ideas, just for a small amount of gametime. Plus it doesn't change that killing enemies is still a core part of most RPGs - it's like saying it should be possible to complete a shoot 'em up without killing things :)

One advantage of having XP awarded on smaller things like killing creatures, and having them respawn, is it makes it easier to balance the gameplay - if it's too hard for a player, they can always get stronger by killing a few more creatures.

With paper RPGs where you have a human GM, it's easier to deal out the XP at the end of a quest (or major part), because a human can better use their judgement to decide how much a person deserves or needs to balance the game.

There's two conversations happening here. Some people are looking for something more story heavy. The mod I was imagining, though, would still be a game about combat. It's just about minimising the repetitiveness of the combat, making a shorter game.

I found the blog post and this discussion thread very interesting.  I'm not opposed to hack & slash games (Nethack is one of my all-time favorites, which probably dates me a bit).  But I too am often uncomfortable with the basic requirement to be the worst mass murderer in history in order to "win."

I understand that a nonviolent (or less violent), story-driven RPG would be a very different sort of game, and would generally require a whole lot more hand-written content.  Flare's not going to be that, nor was it intended to be.  That's fair; it should be what it is (and be as good at that as it can be).

However... I'm entering graduate school (as a non-traditional student; I've been working in industry for a couple decades) in AI, and I'm looking for a good environment and community to contribute to.  My vision is NPCs (including both friends & enemies) that have some real depth to them, and behave in human-like manners, beyond the simple reflexes, state machines, and scripts that even big-studio commercial games still rely on.

That ties in directly with the topic of this thread.  Suppose the goblins (and other humanoids) were mostly neutral, as Clint suggested, and their attitude and behavior changed based on their knowledge of you (the player).  This could include both actions they observe directly, as well as knowledge communicated by others in their community.  This would present obvious "sneak" opportunities where if you're going to do something vile, you better not leave any witnesses.  Conversely, if you do something good, it behooves you to make sure it's observed.

In addition to this basic disposition towards the player, I'd want to add simulated emotions, so intelligent enemies would act in more interesting ways — fleeing a hopeless battle, except when a loved one is in danger, for example; or being susceptible to bribery, threats, and persuation based on what each character knows and values.  I'd consider learning, too, though I honestly don't know if that would ever be noticeable, given the rather short life expectancy of most agents in a game like this.

All this doesn't require a lot more hand-written content... these aren't really story-driven changes; they're just more sophisticated behaviors, applied (to varying degrees, perhaps) to all agents in the game.

What do you think?  Is there room for (and interest in) this sort of deeper AI in an action RPG?

This sounds like a very interesting project. Here are my personal opinions about your proposed changes to Flare. Of course I only speak for myself here and Clint ultimately decides on what to include.

1. Regarding the actions/consequences system: It would be very difficult for action rpg players to appreciate the sophisticated behaviour that would be going on in the background. In fact, a player could probably play through the entire game withohut even noticing that their actions had any consequences at all. The main issue is that the only relevant choices that a player has in Flare is to attack or not attack an enemy or to take or not take an item. To the Action RPG player, it would be difficult to turn this into a moral decision. For the sort of behaviour you are describing, I think the game would need a more complicated dialogue system, such that a player can see that they are selecting one option or another. This way it would be clear that they are making a decision which could affect something else. Of course this in turn adds a lot of additional burden on game content creators.

2. The changes described are not really consistent with the vision of the game or the way it is developed. Please refer to the post above from Clint.

3. As mentioned previosuly, the main part of gameplay is killing enemies and combat. So if combat becomes optional (in the sense that you can choose not to kill something and in the sense that neutral enemies dont attack you becouse of your previous actions), what does the game become?

Sorry for the negative tone of my comments. I would be interested in following your progress with whatever you choose to do and of course your other ai ideas might be more appropriate for Flare.

Of course if you really like Flare and want to work with it, theres nothing to stop you from doing so, but you may have to be satisfied with the fact that your changes would not necessarily be accepted into the main project. If you did come up with something which is more suitable for the Action RPG genre and for the Flare game, maybe it would be pulled into the version 2 engine, but you should not count on it.

Perhaps you should also consider the ai possibilities for allies/companions. This is potentially more interesting than enemy AI, due to the enemy life expecancy that you mentioned. The "simulated emotions" behaviour you mentioned could be suitable here e.g. an ally is less/more likely to run away, protect the player, protect another ally, charge into combat etc depending on their personality. It would be easy to show personality and an emotional attachment between allies if they were able to partake in dialogue. The only thing about this is whether it would need to be any more complicated than a few threshold values and % chance variables for each ally.

Hi Ryan,

Thanks for your comments — I don't take them as negative at all.  They're very sensible, in fact I had similar concerns myself.  An Action RPG is built around the hack&slash mechanic; in such an environment, more realistic behavior of the things you're hacking and slashing probably only detracts from the game.  This genre might be traced back to Gauntlet, which was an incredibly fun game even though it consisted of little more than wading through waves of mindless enemies.  Slowing down to consider the consequences of your actions doesn't really add to that sort of gameplay.

Your suggestion about allies/companions is very insightful, though.  That's certainly something I'd thought about too, though above I was taking Clint's concerns about humanoid enemies as the launching point for discussion.  But you're probably right, characters that are going to stick around and travel with you for a while would have a lot more opportunity to show off their AI.