Dynamic Quest Design
HTML-код
- Опубликовано: 8 фев 2025
- I talk about how I would design and implement a dynamic quest system.
Videos I reference:
Writing My Own TTRPG: • Writing My Own TTRPG
Quest Implementation: • Quest Implementation
Arcanum Generated Dialog: • Arcanum Generated Dialog
Bartle's Player Archetypes: • Bartle's Player Archet...
Sometimes I Hate VO: • Sometimes I Hate VO
![Noob To Max With DRAGON REWORK In Blox Fruits [FULL MOVIE]](http://i.ytimg.com/vi/LnBMOinoOvA/mqdefault.jpg)
There is a cool game called *Shadows of Doubt*, which is basically a procedural investigation game that creates crimes for you to investigate. A neat "Radiant Quest" system.
I totally forgot about that game, it's probably one of the strongest examples of dynamic quest design in years!
I think there's a lot we can learn about how SoD did it, and how it all ties together with the world. One of its greatest weaknesses is in how repetitious it gets over time, and how fruitless it seems since every single citizen is a candidate for becoming a murderer.
So it would be great if they could resolve that in some way. I wonder if they could have a system for generating new citizens- people moving into town to replenish the population. Could even tie the businesses in with it, having them rise and fall. Lot of potential there.
@@Anubis1101 Yeah once you notice the patterns of such dynamic quest systems they tend to lose their value, if you ask me.
Wonder how complex one would have to make a system like that to not have their pattern revealed, real life simulation levels? ;'P
@@Gijontin Not at all, I think there's a long way we can go for relatively little effort, by making them generated in stages. Freelancer is a good (albeit simple) example of this: at higher levels, there's a chance your target might be at the wrong location, or your supposed backup may never arrive, or the enemy might try to flee to another location with more enemies.
That sort of multi-stage generation allows for exponentially more possibilities, to keep things fresh for much longer than the simple, straightforward quests we see more often implemented. Imagine going to steal an item, only to find it's already been stolen, and your target is now the thief that got there first. Imagine having to clear out a dungeon, but you find out the monsters have already spilled out of the dungeon and are now also attacking a town. Imagine being paid to escort someone, only to find out they're actually a criminal and lied to you about the job- and you now have to decide what to do with them.
These things are mechanically simple to generate using existing techniques, but lend a sort of uncertainty to otherwise straightforward, throwaway busywork quests. That uncertainty will do wonders for replayability, as long as the chances of things happening are balanced appropriately.
One of the things I love from Cain is how he talks in an abstract language similar to a bunch of big names of Software Engineering does ("you need this method", "you need this module", "your module must be aware of this item", etc) - but the way he speaks feels tangible. After 20+ years in this area, it is so refreshing to hear it and think "yeah, makes sense" instead of feeling someone wants to treat me like I'm dumb who needs to follow a set of magical rules.
Makes me remember and appreciate the quality of the side quests / bounty hunts in Witcher 3. Some were so dark and tragic, it was amazing.
I remember the memorable side quests in The Witcher 3 because they _weren't_ "dynamically" generated "stuff to do," but hand-crafted, self-contained stories, many with bearing on the overarching story. In other words, they were artistically created by humans as opposed to "generated" by AI or otherwise and, therefore, utterly uninteresting "filler content" or, worse, an "infinite (so-called) quest" machine. I did grow a little weary of following blood trails and/or footsteps, etc., which was a neat sleuthing mechanic otherwise, but that's just because there were so darn many of them and the sleuthing mechanic was perfect for supporting the gameplay of all those unique stories. Ergo the repetition was bearable. The quality of the side quests are, I think, among the many reasons, aside from technical, The Witcher 3 was widely considered groundbreaking and is lauded to this day as an exceptional example of game design. Afic, it set a new and, as yet, unsurpassed bar for *quality* "content".
Perhaps needless to say, quality is emphasized here as quantity over quality appears to rule in the industry today. To say there's an imbalance there would also be an understatement.
Watching this make me want to revive an old project I got side tracked from and abandoned, to make a procedural multi-stage quest line generator (or rather, choose your on ending style quest-network generator). I'm a big procedural generation nerd who learned to program rather than just "code" making dungeon generator and re-writing the Minecraft world generator, and had a lot of ideas how to do it but never finished.
Thanks so much for the advance video, Tim! I look forward to those future videos to explore this idea further, since I absolutely love thinking about and discussing dynamic elements in games. I think it's videogames' greatest strength! If it's dependent on future content though, I really don't mind if you push it off until those are done, I promise I won't be offended! I'll watch every video of yours either way.
There are so many interesting ways to implement this system that I've seen over the years. Freelancer has very basic varieties- kill all the enemies, kill one specific enemy, destroy installations... but does interesting things at higher levels where the mission may not go as advertised! Bethesda's radiant quest system has a ton of potential, and I think Fallout 4 did a fantastic job of leveraging the open world to make them appealing to players (even if it did get a bit naggy).
One really interesting case of MMOs using them is Mabinogi. There's all kinds of odd jobs to do in that game, from crafting supplies to sell to a local store, to killing monsters in a dungeon, to caravanning goods between cities (that one has a lot of depth, as you can buy your own backpacks or carts to hold stuff). They also have a fascinating narrative around it, where NPCs may or may not remember you, due to the strange way time works in that universe. I think tying other mechanics or aspects of the game into dynamic quests is important for making them more appealing and enjoyable to players. The sheer volume and variety of them, as well, works as a strength, allowing players to explore the world at their own pace and in their own way, and I wish more games put that sort of consideration into it.
Dwarf Fortress's Adventure Mode just came out of beta recently, so I've been playing that a lot. It's a very hardcore example since balance is difficult to quantify in that game, but the reward of going down forever in the history of that world because of your actions is a really interesting aspect of it. The idea that randomly generated worlds can not only create content for the player, but also respond in turn to the players actions is one of the greatest achievements of DF, and I've considered making my own mods to expand on their dynamic quests now that Adventure Mode is out. They've got a surprisingly robust engine for generation, so really the only thing stopping me is time.
For my own games, they'll be tied to the world state and whatever else is going on. I don't think I'll make use of generative AI at this stage, but I've got some working ideas for token-based dynamic dialog, and how to tie them in with other aspects of the game. I've got too many ideas and not enough time or skill to explore them all, but that's what I love about game development!
Awesome video! What I dislike about traditional quest design that's often paired with branching/linear storylines is that there's often a solution that "works better", because it's "designed for that solution". Almost all RPGs are playable as a generic knight type of character, but if you try to do something "different", you often have a hard time. Ever played Arx Fatalis as an archer... well there are 1 or 2 bows and one "rare" that you can get at the end of the game... if you don't know this, you're doomed. Baldur's Gate 1 with a Katana... there are not many Katanas out there, you're gonna have a hard time.
Narrative wise it's also the case, choosing the "neutral option" or a "mix" in BioWare games that came after Mass Effect, feels like you'll almost always get the "generic ending" that's for people that are not "pure good" nor "pure evil".
A dynamic quest system could adress those issues and "create" quests that fit to your playstyle etc. the huge problem I see is that there probably needs to be an ending and these often have boss-fights. New Vegas did this exceptionally well (of course), because it offers so many solutions to the given problem. Alduin is the easiest fight in Skyrim (you can idle in a corner somewhere, doing a shout here and there and just do the last hit), it's kinda the "safe route", because "anyone" can kill Alduin at any given time (which kills credibility and the narrative overall of course). Gothic had a mechanic based bossfight and solved the problem this way.
I'm personally thinking about a storylet approach, where each storylet has an "enter requirement", if these are fulfilled, you can "enter" the storylet (comparable to StoryNexus' structures and their games heavily use it). So if the player steals something, a storylet can have something like "hasStolen" as requirement that can end up in a thieves guild questline for example. Since these storylets are somewhat "loosely coupled" and not hard connected like it's often the case with classic branching/linear quest structures, I can dynamically react on items that are found BEFORE the player even accepted a quest for this item for example. Or I can have a mysterious creature that annoys some people and all people can tell the player about that creature instead of "only this specific NPC has this specific dialogue line about this specific mysterious creature" and all other people around don't "know" what's going on :D.
In my head it's nice and I've implemented it in Ink (a narrative scripting language) which works great, but I need to implement it in the game project I'm working on aswell... which will probably cause some headaches :D.
How would you solve the "paradigm" of "bosses should be a challenge for the player" with dynamic quest design, where you probably don't know "when" the player approaches a specific boss and you probably don't know how strong the player is at this point etc.? Are there games that address this? I sadly tend to "limit myself" in RPGs, because I fear that I might not be able to finish the game with the stuff I have in mind for my character.
To make sure bosses are a challenge, I would recommend level scaling them...within limits. I talk about that here: ruclips.net/video/za5u4j6lc-w/видео.html
Inkle’s structural ideas are killer, and Ink is a cool tool! Really Inkle’s approach to emergent storytelling vibes with me… of course it’s a ton of work, because you have to spend a lot of time imagining the possible game states that result from various elements interacting (and then debug why some emergent result is sideways). So you’ll do a ton of work that most people who play won’t see, which of course corporations are not known to be thrilled to support that sort of work. But if you can at all afford to do that kind of work, it will absolutely make the game more replayable and have a deeper sense of a living world. I think that type of work is one reason why people love BG3 so much, they went far out of their way to include content for people who were crazy enough to do certain things. In their case it becomes more like easter eggs since their narrative isn’t emergent. But it’s got some of that same feeling for the player "ah wow! They made this content for meee!"
Another great benefit of dynamic quests concerns content as reward (which was covered in a previous video). For example, the place in which to carry out the quest could be chosen automatically in order to make the player discover an area they had not explored, which could contain a unique secondary quest, or a village with a unique faction, etc. A very well done system could choose to put the dynamically generated quest at a point that takes into account both the player's current position (thus predicting the path he will probably take) and his skills and attributes. This way the player would have an "excuse" to get to a point, and along the way they could discover a whole series of things that they would otherwise never have seen.
6:57 speaking of botched quests. I recall a quest in fallout 3 where you ultimately would help a ghoul get access to a specific power armor. If you got it yourself you failed the quest and he would be angry with you.
Then they made a dlc where the prize was another set of same power armor. Picking that up still fails the original quest. I guess some item id overlapped...
I've only seen two implementations of this kind of procedurely generated quest design that I've particularly enjoyed: trade opportunities in Sunless Skies, and missions/information in Starsector.
Sunless Skies uses it as a method to allow the player to generate income while travelling and exploring, while requiring them to carry/acquire a certain amount of a certain item, adding an additional resource management element to the game. It also acts as an immersive shipping mechanic, since the game is about driving a train, and thus making deliveries.
Starsector does something similar, but targets regions that the player hasn't explored or completed yet, meaning you can be assigned missions or information telling you about valuable resources in a specific star system, or the location of a unique/worthwhile encounter.
The first example adds to a feature that is already in the game while solving the player's issue of "How can I pay for supplies/fuel?", and the second example guides/encourages the player to explore the colossal area available to them by highlighting potential encounters.
I'd also give an honourary mention to Uplink, which uses similarly generated quests as the bulk of its content in the form of jobs. It's a game about hacking, so they'll be jobs like... delete files in a target computer, steal data, edit someone's records, etcetera.
They're somewhat immersive, but vague enough that you get the vibe that your actions aren't having too much of an impact (thought you will see the results of your hacks on the news). What makes Uplink interesting, however, is that if you continue doing these "radiant" missions for long enough, you'll discover something fantastic:
They were hiding the main plot of the game. The procedurally generated quests were effectively just training for a series of much harder hacks to foil (or support) a conspiracy which you absolutely can miss if you aren't paying attention.
Something not mentioned that I think could help avoid the procedural "quests aren't interesting" factor is to have them expand (compound) on one another. You can still use the exact examples Tim gave, but have inputs into them taken from the results of the previous one. Example:
A - Explore and Report - Player explores a location, they can report on ore deposits, rare flowers, dangerous wildlife, etc. Skill dependent. If they don't have the skill or don't explore, quest chain ends.
B - Escort Quest - Player sent to escort a team to exploit natural resources in explored location. Size of team dependent on number of resources found.
B2 - If teamsize > X, the caravan is attacked en route as a juicy target by bandits.
C - Fed-Ex Quest - Player sent to take more goods to outpost team, or bring back local resources obtained by team. Weight dependent on size of team. Very big team might make it a wagon which makes it a second Escort Quest?
D - ETC
Because the inputs of each new quest is entirely based on the players performance in the previous, which itself is enabled by their character build, the actual quest details will be more varied and dynamic then just randomising the inputs. Not any extra real work either, so long as you design the system to store those in a character performance quest log type thing. You can further decrease the feeling of it being procedural by having it be more of a callback, so no escort immediately into a fedex to the same place. Have it separate by a few quests, or number of in-game days timer or something.
Wrote basically the same thing earlier but youtube ate it so re-writing it (and taking the opportunity to clean up the language). Just as an explanation in case youtube regurgitates it and I've got two very similar comments.
That’s a really good idea to chain quests that way. And it’s something you can only do with a pool of many quests, something a dynamic system would provide.
@@CainOnGames Thanks, and yeah that was my thinking - maximise the unique aspect of the system.
If the system is suitably complex you can also have weighted RNG on the initial quest in the chain (and follow ups) based on the guild or quest giver, so players who prefer certain types of content can seek it out in a logical manner - those who want to fight challenging enemies logically join the fighters guild, who gives quests based on fighting, and their exploration or fedex quests are through dangerous zones. Explorer types join the surveyors guild, and so on.
This is a very good idea, but the implementation is not as straightforward as you think. Quests are typically done as scripts written by quest designers. This would require that engine programmers write a system that's at least one order of abstraction higher than this so the quest implementation is generated from a meta-description of it, or alternatively provide a way for a quest designer to provide a meta-description and write a management layer that would consume it to match quests. Not impossible, but not as easy as one intuitively might think. If the first option is used, you've got the additional problem of the generated quests being formulaic. There is a reason I can't think of a game that does this at scale.
With that said, I absolutely agree that this would add a dimension that makes questing seem deeper and fresher, at least for the first few games that do it because of the novelty.
AI is gonna completely consume this style of Quest/Mission Creation.
It couldn't be worse than Starfield
I've debated the drama around ai vo for awhile, I think the future of ethical AI VO would be to license voice packs from VO performers. Where they get royalties or flat payments when their voice is used a certain number of hours on average in a singular project. This would be expensive but cheaper than hiring them to actually deliver the lines, and for non-guild (indy) projects it would be far cheaper than hiring a professional vo, but it wouldn't be free.
That seems to be the only sticking point between ai vo and the performers, the fact that their voice is being used and they aren't getting paid for it. Creating voice data that is clear, specific, and detailed would be a large upfront cost on the artist but then they would get a steady income off of it as long as it remains in use.
This is already being done, there are actors who license their voice under a standard license for ML-driven TTS and a few companies that offer access to them. Keep in mind that a lot of actors (and artists, and.. ) object to ML everything regardless because they're already having trouble making ends meet and reduced demand would push them over the edge even if ML use were to observe all possible legal and ethical norms.
Shadowrun for the Sega Genesis(Mega Drive) had repeatable quests as well and it allowed for emergent story telling by being deliberately vague as to who and what you were delivering/destroying. This is not exactly the same but this video reminded me of it.
Afternoon Tim isn't real. He can't hurt you.
Ooo recovering from botched quests is a great topic
Recorded, uploaded, and queued for March 17th...assuming nothing bumps it to be later
@CainOnGames yay!
Dragon's Dogma 2 has a forgery system where you can copy items. Often, these are unique and required by two different quest-givers. 8:47 -> they also did this a lot in DD2. Made me find all sorts of places.
Both Dragon's Dogma games have it, and it's one of the most unique systems I've seen in an RPG.
Subverting the game's own quest system like that is just an absolute joy.
Fun thing about that is that (at least in DD1) because the forgeries are, well, forgeries, they don't actually work. In some quests this means that giving in the forgery to hand the other somewhere else or keep it yourself can have negative effects. One notable one is the Wormking Ring, which is amazing for casters so you might want to keep it, but it also lets the King open the chest for your quest reward. So if you hand in a forgery the cutscene shows it failing to open the chest, and he says something about the "magic being worn out" and you just don't get a reward (the Ring is better for most characters anyway).
If your making an escorting an NPC quest always remember make the NPC slower than the player running but just too fast to walk beside them
It was an interesting question, and thank you for interesting answer!
I mean, this sounds a lot like what solo play in TTRPGs is all about. Set up some tables, give it a roll, and solve the problem in a way that makes sense in your game. Add an oracle table and you'll get your ideas going.
Check out shadowdark with its solo dark booklet. Uses d20 yes/no oracle with d100 tables to expand the answer. The core rules have every dungeon, terrain and town district encounter tables. I think their very simple mechanic would work well for adding procedural immersion to this game. It's very thin answers leaving it to the players to connect the dots without a gm.
Yes, but what makes this easier and more fulfilling in a TTRPG (whether solo or not) is that you have a flexible, human oracle who can decide in a non-formulaic way whether your solution passes, thereby accommodating a lot of variety. It is an open question how to reproduce this in a CRPG.
making these quest priorities content you haven't done yet sound like a good idea,
for the escort quests you could have a merchant caravan and you just sit down and enjoy some dialogue or scenery or music between the combat encounters, this might make escort quests more bearable
Yea the biggest issue I see with escort quests is they control the pacing, instead of the player. There are two ways I can think of to address this: design it so the player controls the pacing (eg: the NPCs follow at the players pace and run/teleport/etc to keep up), or make it so the player doesn't have to interact with the game until combat or some other interaction is needed, like the caravan example.
Freelancer was originally supposed to have dynamic escort quests- though they were cut, with only a few story missions having the mechanic, and it's a perfect example of the latter situation: you can simply enter formation with the caravan you're escorting, and sit back and let the caravan navigate to the destination until a combat encounter. The game lets you do that in the open world, as you can enter formation with any non-hostile NPC you see flying around, but you don't get paid for helping.
@@Anubis1101 The NPC taking away player control is one problem. An even bigger problem is primitive NPC behaviour (aka "AI"), which is a bane on all games that have companions/party members/etc. Some games attempt to partially solve this by allowing you to customise the behaviour of your party members, but even if that worked well (which it rarely does) that sort of investment is only justifiable for a character that's going to be with you for a while, not a one-off like in an escort quest. The number one complaint about those quests is how "stupid" the escored NPC behaviour is, how bad the pathing is, and therefore how difficult it is to keep them alive even under trivial threat.
Very often in many games dynamic quest systems feels empty/soulless and get boring suprisinly fast. Bethesda used this even in elder scroll arena. With Tim´s tip lets all make great quest systems ;)
exactly. Everytime it happens I either get uninterested in the game or I just ignore every side quests. Artificial longevity is one of the thing I hate the most in games nowadays.
I distinctly remember all the hype about Radiant quests and being thoroughly disappointed the moment I experienced them. They were primitive, formulaic, shallow and unimpactful. Not just after a while, but almost immediately.
The system of forgeries of quest items is found in the games of the Dragon Dogma series. For example, there is a character who asks to bring a demonic manuscript - and if you bring the original, he will summon a demon and die. And if you give a fake, then just put it on the shelf.
Another way to think about Dynamic Quests is the nature of Control.
In a Strategy Game the Player is used to Order units around to do something, but if you think about it you can use Quests to make Players do something similar.
If there was Meta-Strategic Simulation Layer to the game then you can give the equivalent of a Command to the Player by generating a Quests to achive what the AI considers a Strategic Objective.
Of course you have to incentivize the player with the right Reward but that is the cost of using someone as powerful as a Player.
Strategy games are definitely a genre where dynamic quests haven't made much of an appearance, and that's a shame. Some SRPGs make use of it to a degree, especially ones that lean heavily on random world generation like Disgaea and its Item and Chara worlds.
I think squad-based RTS games have done it a few times as well, though I've only played a handful of them.
I would love to see RTS games or hybrid strategy games like Valkyria Chronicles make use of them, I think there's a lot of potential there.
@@Anubis1101 I am not saying to add Quests to a Strategy Game.
What I am saying is to use the Players as if they were a Unit in a Strategy Game, and the only way to do that is through this kind of Dynamic Quests for a Faction that achieves the objectives of that faction.
You can't just Command and Force Players to do things, that violates their Agency and probably pisses them off.
But if you make it into a Quest and add some XP, Gold, Rewards and Faction Rep you can make the player do anything. Even to their own detriment.
The rival leader is a problem? Just let the player assassinate them.
@@adrixshadow I seem to remember a game from some years ago that was a hybrid RTS/FPS, though I never actually played it.
Might be worth investigating.
@ C&C Renegade?
I was thinking this more for a RPG or maybe something like Kenshi or Mount and Blade.
@@adrixshadow Yea that might've been it.
Forgot about M&B. That's a game that really does mix those kinds of things, though it's always felt a bit janky to me. That might highlight why we don't see this done more in games, though: it's expensive. Developing the strategy bit itself is one thing, but then tying it into the world and single-player aspects of the gameplay are an entirely different beast. You're essentially making two games in one, and that means double the time spent testing and tweaking such high-level mechanics.
I plan to explore hybrid strategy mechanics in my own games as well, so I'll definitely be looking into this more myself.
Gear Head RPG is the best project that has development around this for people who are interested.
Can't post a link but there is a blog from the developer that has very intresting insights.
Tim, can you please make a vid on what was it like to implement design an questing tools back in the days. I think we can all learn a thing or two. And where do you think is the line between core programming and gameplay expressionism, and how not to get too stiff or limited in the process. Thank you.
I did that here: Quest Implementation
ruclips.net/video/pw6tOlmwwYE/видео.html
And I explain the scripting here: Arcanum Scripting
ruclips.net/video/coPkWyJIl-M/видео.html
And the tool building here: Building Tools
ruclips.net/video/_P4-2Nv_Or8/видео.html
Lately I've been designing a quest system that generates sidequests - but the catches are several.
First, quests may or may not have hidden "escalations". Radiant quests in let's say Skyrim are very head-on. Why "kill X in location Y" should always be the same i dunno, i disagree with it:) Target may know or not know about you, it may even not be there and you would have to investigate first, target may escape after the fight and it would become a chase sequence, etc. Simple assassination mission may then become several types of missions by the grace of random added steps/permutations
Second, options aplenty, we must have various ways of doing things. Such quests are to be intertwined with faction reputation system which in turn facilitates different methods of going about the quest, resulting in rep change depending on actions and outcome. If your kill request targets a member of hostile faction, you are clear to engage, true. But if this target faction is not at war with the quest giver, several options are available. You may need to be covert, you may want to implicate another faction, you may openly do it and live with the consequences. Or you may negotiate, switch sides etc. So, based on quest type + escalations present + faction relations there are several choices and corresponding outcomes.
Third, stakes. Faction repercussions must be a thing, more than just some rep change. Factions must have various states they can be in, so that quests have tangible effect. So, the point is about deepening the outcome effect. If you outright kill a leader of a faction, it must lead to various outcomes like electing new boss, creating splinter groups, organizing vendettas against you, or forced peace talks with the quest giver, etc etc. Several of those together, even. Also, with different faction states, with your relations, the combination may determine the types of quests available, resources provided to carry out the job, rep gained etc.
Fourth, as with hidden "escalations", hidden "agendas" might also take place. Like if you get a kill request, but the quest giver faction wants you dead instead this way, the quest morphs even further. Just imagine telling questgiver you succeded, but they wanted you to be gone. All depending on reputations, faction agendas and other moving parts within the game systems. Any quest may be a set-up, or smth else - an uncovering of an agenda leading to different activities. As if you were tasked to kill bcs someone is an enemy lieutenant, but then you find evidence he was an undercover cop, which your questgiver would not want to kill, spawning unwanted/unforseen outcomes and putting the faction in an interesting state. You bringing the results back to the quest giver does not need to be "mission accomplished", it may spawn a new situation created by the choices/findings in the previous quest. Which also needs to be optional - you may uncover a deeper plot, but you can miss it and go about your day without engaging by missing or ignoring it.
It's not all, but those are my main points, all of them intertwined. Just with one quest base type you can see how different the game would become, how didferent each quest may then be, all seemingley being "kill X in location Y" on the surface.
Perhaps in a year there will be a prototype from me, applied to a small "toy", to show rather than tell. That would erase my main gripe with radiant quests being same-y and inconsequential and having more depth.
Ooo wait here me out. You have to get a water chip but many water chips are broken after the years since the bombs fell.
Fallout Forever
Hi Tim. Have you run into any particular technical challenges when making dynamic questing systems? Does such a system tend to require more computational resources or present particular resource management obstacles?
I have never implemented a dynamic questing system. The closest dynamic system I made was Arcanum's generated dialog, which has a video linked from this one's description.
I think games like Balatro prove that dynamic content can be fun, but it all depends on the execution. Shame how skyrim's bland radiant quests kinda poisoned people's perception of the idea
Bethesda has done some of the best and worst examples of dynamic quests. I think this is because what makes them great depends on other parts of the game, like navigating the world, or the general variety and accessibility of encounters the game supports.
"Another settlement needs our help."
Timothy Cain I love your videos ❤ you created one of the best game franchises ever.
It is my strong expectation that a map and menu(and/or strategy/sim) type game will create a dynamic rpg experience with actual choice and consequence first. If that doesn't yet exist then you can be assured that a 3D voiced walk around rpg won't do it any time soon.
Greetings, Tim!
What are your thoughts on game demos? They were quite prevalent in the 1980s and 1990s but became less popular in the 2000s. It seems like the traditional area-limited or time-limited game demo lost its place in the mainstream to be replaced by trailers, streamers, reviews, the free parts of freemium games, and Steam sales.
During your time making games, how did you and those around you plan for and implement demos of your games? What are notable examples of games where the demo helped sell the game as a game maker and player and why? When were demos not worth it as a game creator and player?
Do you believe where applicable more traditional game demos should push for a return to popularity or generally remain as they are? Why?
To be clear, this question is primarily about playable demos and not just trailers and watching others play.
Thankee!
Informative!
In your botched quest video will you talk about softlocking games? Like in Zork when you eat the garlic item you can no longer finish the game due to a vampire later on.
Yes. There were some early RPGs that became unfinishable in the way you describe, and that led to my quest design in Arcanum that allowed failed quests to be unbotched.
As a fan of roguelikes I wonder how much could you push dynamism. Dynamic quests are fun but what about dynamic recipes, dynamic creatures, maybe even dynamic magic systems?
Dynamic recipes mean each run potion of health takes different ingredients e.g. run 1 heart + water = hp potion, run 45 water + mint + lemon eye = hp potion.
Dynamic creatures could be enemies generated out of exchangable blocks. E.g. fire + golem = fire golem, and ice + skeleton= ice skeleton.
Dynamic magic is like something from Morrowind, but each creature is given their flavored magic. Fireball for fire golem.
I think a lot could be done with creatures plus color and fx changes to make dynamic creatures. But dynamic recipes raises questions about what changes the recipe. Does it change over time, or are the recipes generated for a particular PC (whether a single player game or in an MMO) and remain stable after discovery? And in an MMO, how would you make the recipes "fair", in the sense that Player A's recipe contains ingredients that are just as difficult or expensive to acquire as Player B's?
@@CainOnGames In Roguelikes, which are almost exclusively single player games with random seed, it would vary with the random seed. Whether the seed is generated per run or per world is to be determined. Albeit, per-world makes more sense. The idea is stable per discovery but unique to a person (i.e. each player sees their own version of world).
In a MMO, if that's even possible, you'd need some kind of internal evaluation e.g. Potion of Healing must contain water, herb and organ of regenerating monster, total gold value of ingredients must be < 100, or some combination.
An even crazier suggestion is to give herbs random effects, some good some bad, and then use tools to try to remove some effect or combine them a la Morrowind. A potion in Morrowind basically dependent on ingredients having same effect, so it there was more than one way to skin the health potion cat.
Dwarf Fortress does a TON of randomly-generated stuff, including musical instruments. The hard part with random items is establishing how to give it possibility without being too unbalanced.
One example would be cooking mechanics. In the Mineral Town Harvest Moon games, you have a standard recipe, but you can greatly enhance (or degrade) the outcome by adding extra ingredients to the process. Since the core recipe doesn't change, it limits the amount of variation to a somewhat predictable range.
The real limitation of dynamic content is art assets/production values. You can do recolors and special add on body parts or w/e but that's pretty blatant to the player. You can absolutely do all the things you mentioned data/mechanics wise. The graphics and sounds are the limitation.
Here's an idea regarding the dynamic potions suggestion: it could be useable in an apothecary setting, where you're treating others. The recipe for the potion might not necessarily be different, but you'd have to adjust the portions to match the patient and their illness (e.g. avoiding having too much of one ingredient as to avoid poisoning them, or solving a deficiency they might have.
If you've ever had a blood test, take a read of the results for an idea of what I mean.
Additional items (you mentioned mint + lemon eye) could be use for unique statuses or afflictions a patient might have (even something as simple as making medicine taste less bitter for a child). In this particular setting, I guess it would be more that the patient's requirements would be dynamically generated rather than the potion.
Considering this video and another recent one you made both touched on dynamically generating things like quests and NPC dialogue, I wanted to ask: Do you have any experience or familiarity with multiplayer RPGs like Space Station 13? Do you have any opinions on the effectiveness of such games at generating emergent experiences? There are problems with games like this but they've provided me with the closest approximation to a dynamic TTRPG-style experience in a video game I've ever had.
I propose annual chocolate pumpkin muffin making
One note Tim - what Bethesda did in "Radiant AI" is not "complicated", it is very "simple"(!!)
much easier than it seems in the game.
Bruce Nesmith had a lecture where he revealed the essence))
"Presentation: Radiant Story - Bruce Nesmith" on
Vancouver Film School.
I recommend checking it out)
Wellthankyou🎉
Will check it out thanks. Also. Always fun when people tell Tim he’s wrong about how complex it is to code game systems
@ I wasn't talking about the Arcanum systems - I was talking about Skyrim and Bethesda)
Bethesda had really complex systems in the days of Daggerfall.
Skyrim is simple - it contains two checks: "Are there any ordinary locations that have not yet been visited" and "which location from the unvisited" to choose for the radiant quest.
@ In fact, there is a very big joke. I've seen a lot of cases where "emergent" systems were rated as "complex" - when in fact they were simple. Even programming professionals could be wrong. I personally witnessed as a producer of the studio where I worked - he said that thousands of people worked in Bethesda (in fact, there are only 90): "We don't have a thousand people like in Bethesda to make our own Skyrim"
Overestimating the difficulty of creating your own Skyrim is one of the reasons why no one is competing in this game.
@@YarGolubev The cost involved in making a Skyrim-like is not, by and large, in coding.
Hey Tim. Question about the value of short games and long games but not monitary value and instead the value of your time. I'm someone that has been trying for years to make games solo and while I have yet to finish one I feel I am making progress through mistakes but due to the time this takes to solo learn and make I rarely play games that often but I still love them. It feels like a lot of people put value on long games with a lot of content while I avoid almost anything these days longer than 10 hours. Curious how you feel about the benefits and drawbacks of long and short games. One of the things I do enjoy about short games is I feel like they tend to lack a lot of bloat and are able to be pretty straightforward (I loved that space marine 2 was basically a series of hallways with pretty landscapes and cool setpieces for example) so I'm spending less time trying to figure out what to do and where to go and it's much easier to not miss content since I'm the type of person that's not really interested in replaying the same game for content I missed however just as much they can be strengths for me someone else may find them to be weaknesses. Might also be cool to see a video on how weaknesses of some games can also be strengths depending on the player
Good question about game length. I have never done a video on the pros and cons. I will add it to the queue!
I think you might enjoy The Wayward Realms - (the spiritual successor to The Elder Scrolls II: Daggerfall). Trying to bring TTRPG's mechanics which were quite left away in videogames, and also through today's technologies (AIs, but carefully). Your last part feels similar to what they share online.
I hope it’ll come out and be fantastic. It has a writer from Morrowind on board, and that’s one of my favorites.
wow i just know/learn this dynamiq quest system, i thought it was manually created by dev, lol me haha.
Thanks a lot mr Tim!
but still have a question, did this system create unique scenario for each player? or it will be genrated before the game released on public?
This is a hypothetical system, but my expectation is that it would create unique scenarios for each player as needed.
Escort quests are fine in abstract theory, but tend to fail in practice just because game systems rarely give players the flexibility they would need to not feel awful. Too typically the escortee does not take care of themselves well enough (or even is just too slow and annoying when out of combat), and the mechanics don't give you enough to solve that yourself, like the ability to shove them around, carry them, push them down to avoid an attack, automatically intercept attacks against them, etc. I think in games, escort missions are only good if the game is specifically designed around that concept, with particularly skilled NPC AI, particularly complex player/NPC interaction options, etc. Or just make the escort "inoffensive" like Elizabeth in Bioshock, where she can't be harmed and does a great job staying out of the way, barely an inconvenience.
If the quest is generated on world creation you wouldn't need to regenerate areas. Could see that the item is a quest item before finding the quest giver
Skyrim bounties and F4 settlement help being the type of implementations to avoid, imo.
The moment I notice the pattern and realise the quest is next to completely randomized it tends to lose the value for me ...unless it's more of a grindy type of game where the randomness just helps with the variation. But slight location variation of quest items does not contribute enough to gameplay unless world is completely randomized as well and exploration is a decent fun part of the gameplay, I dunno.. 😅
The problem is that the variation in those systems is on variables that have little impact. What exactly you're fetching impacts nothing, where you're fetching it from at best prompts you to explore an area you haven't been to yet, and the quest's completion has no significant impact on the world. Maybe it's the programmer's curse, but this is immediately obvious to me and I simply can't enjoy these quests and they're actually off-putting to me as they break immersion. Fixing those issues in a tractable fashion is an open question though, there's a reason these things were implemented the way they were.
i’m curious if you have any ideas on how mechanics like this from open world rpg type games could be applied to games with a fairly linear story and level layout
You could dynamically generate all of the side quests on every level. The story wouldn’t change, but depending on the player build, different side quest and side quest solutions would be available.
The more you speak, Tim -- excitedly to boot -- about using programming in this way, the less hope I have personally for the future of genuine RPGs. The triple A industry, at least, obviously isn't all that interested in creating art -- especially choice and consequence RPGs -- at this point so much as building "infinite" (so-called) quest (actually distraction) machines that invite players to put their brains in park and do meaningless "stuff" for hundreds if not thousands of hours of their short and sweet lifetimes as I see it.
Have you anything to say that might assuage such concerns? Is there hope for striking a _healthy_ balance between mechanical and artistic, quantity and quality?
Like many things I talk about on this channel, dynamic quests are a tool that can be used in many ways, good and bad. Many people immediately jump to the worst ways it can be used, but there are many good ways as well. One commenter pointed out how dynamic quests could be chained into longer arcs, and the subsequent quests are created based on how the earlier ones were completed. That’s a brilliant idea, and it takes reactivity to a whole new level.
Another example is generated dialog in Arcanum. The game was praised for its reactivity, and generated dialog allowed that to happen in many cases. I would not replace all dialog with it. I wouldn’t even replace most dialog with it. But it can do things you can’t do otherwise.
My excitement is based on how these features can make games better.
@@CainOnGames Alrighty then. Have you anything to say to up and coming developers about striking a healthy balance between reactivity and responsiveness?
It's responsiveness, a quality, as opposed to reaction, emotional or otherwise, that I think sets yours and maybe a few others' games apart from the crowd -- the responsiveness involving the exchange of ideas between developers and players primarily in the worldbuilding, ambiance, story, characterization, dialogue, imagery, etc. That's largely overshadowed by tsunamis of meaningless stuff to do in modern games in my experience. Case in point: Fallout 76. The very nature and intent of the game makes it more or less inevitable that any ideas therein worth engaging with (and there are a few) are difficult to discover. "Fragments of creativity," they're called by video essayists such as Noah Caldwell-Gervais.
Much of the criticism of modern RPGs comes down to the feeling that they're "lifeless," "souless." Such criticisms are obviously not quantifiable, nor can they be "evidence-based." Therefore, it's difficult for players to get across to deveopers what they mean by that. It's something that has to be intuited via imagination -- empathy between developer and player.
@@lrinfi I am not sure I am qualified to talk about responsiveness. As you point out, it consists mostly of story, character, dialog, imagery, etc., things that I rely on team members who are strong in those areas. I do not participate much in the art of my games, and my approach to narrative (beyond group brainstorming on setting and story) is primarily a mechanics-driven approach. In other words, I think about and offer systemic solutions to my narrative designers. Those solutions can be relied on too heavily, like any other tool. But my games don't aspire to be simulations but instead to be player-driven stories. I provide the means by which the players can drive their stories.
@@CainOnGames Fair enough. So, you're the nuts and bolts guy who seamlessly rivets all that reactivity and responsiveness into a coherent whole? ;) I'd imagine that makes you immanently qualified to speak on striking a balance between them as opposed to emphasizing one over or even to the exclusion of the other. I'd imagine your experience as a game director, especially, gives you some unique insight into that having overseen both purely technical and artistic work, the question pertaining to the balance between them. I know you've spoken on some of the difficulties involved in communication at times between the art and programming departments. Must be like trying to get two different worlds speaking two different languages to cooperate :)
@@lrinfi I am more of the tools designer and programmer for the narrative designers. It's similar to creating the map tool and quest system for Arcanum after watching the trials and tribulations of Fallout level and quest designers. It wasn't like I didn't suggest several areas and quests for both games (I did), but someone else put them into the game, and I tried to make that process better.
The most fufilling quests Ive played are nuanced, subvert expectations, and have consequences and pay offs, while illuminating aspects of the world and systems. Procedural quests so often end up just feeling like repetitive chores. Will we get the same feelings completing a neural net's quest that we get from playing a human-authored quest? Its similar to the question of can we find meaning and fulfilment in AI art?
It's an established fact that humans can find order and meaning when there isn't any, cf Rohrshach tests. The key is giving them something they can do that with that's not going to break immersion. This is difficult. Even if you did that successfully, a big part of what makes any quest in an RPG engaging is impact on the game world, so you'd have to find a way for the game world to respond in a way that's consistent with the meaning a human player found. That's even more difficult.
What would be a situation where "mentioned" is used to decide something. Why is mentioned needed?
The two main uses of mentioned is (1) the player hears about a quest from someone other than the quest giver or (2) the quest giver offers the quest but the player doesn’t agree to do it (or decline either).
The mentioned state controls later dialog , e.g. “So have you decided the take care of those bandits?”
@CainOnGames ✌️
I love the idea of dynamic quests, they can be really useful in motivating people to explore parts of the world you wouldn't initially go to. But the biggest downside I've seen is how much they can clutter the user experience. It can make people falsely equate dynamic and regular questing. Fallout 3 and 4 have a really good example of 2 very different dynamic quests. Fallout 3 has the metal scraps for megaton and the old world tech for the Outcasts, you never get nagged to do these quests but you're rewarded for going the extra mile adjacent to your main questing. Fallout 4's Settlement system on the otherhand is very naggy which seems to be the main complaint, but unlike fallout 3 it constantly rewards you by changing the experiences you while exploring the world assuming you actually walk to places. The biggest flaw I've seen that fallout 4 and skyrim do is making dynamic quests populate the same space as main story quests falsely equating them in players minds.
Yea presentation to the player is an overlooked aspect of a lot of mechanics. They want to encourage the player to do the content, but to go so far as to bother them about it will have the opposite effect. I've seen the Bethesda team talk about it in interviews, and sadly I'm not impressed with their approach. I don't think they really understand the strengths of their own games, sometimes.
I do think Fallout 4's dynamic quests are really good overall. The quests themselves are extremely simple, but you have fun because of the open world and all the different things that can happen to you on the way. I wish they would've expanded on that and made better use of the system, instead of going backwards with Fo76 and Starfield.
@@Anubis1101 The questing in Starfield is really weird, but I don't think any one part of that game is at fault. The more I dove into the systems of the game the more I find really exceptional design and complexity, but everything is so segmented that what ever you do you always have a net loss experience. 76 is just a carnival, its fun but I don't want to live as a carnie.
@@mikeuniturtle3722 Starfield just copy-pasted a ton of stuff from other games, and stitched them together poorly. i dont know if it was intentional or not, but i can single out almost every mechanic in the game as being a wax museum ripoff of something else, mostly just their own Fallout 4.
Space combat? It's literally just Freelancer, but without the extra controls like turret mode, or the fast-paced difficulty of late-game missions. Somehow a game from 2001 is a much better space combat and exploration simulator than Bethesda's decade-in-development brainchild. And that's not even getting into its more recent spiritual successors, like Elite Dangerous.
Planet exploration? No Man's Sky, except NMS didn't feel nearly as tedious thanks to better implementation and overall better pacing of the gameplay to support it. By comparison, Starfield's planet surveying feels like dull busywork, especially on the more biodiverse worlds, where you'll be scanning dozens of plants and creatures that youve probably seen before. Then they had the audacity to have entire groups of perks to lessen that busywork, making it feel as if they knew it was bad and just liked to waste our time. I put almost 200 hours into my post-DLC playthrough, and never once felt like putting points into those was a good idea. As someone who loves exploration and space, this was one of the most disappointing things to me.
Ground combat? Fallout 4, but without all the dismemberment and VATS that made Fo4 combat interesting. If it weren't for the underlying feel giving it away, it'd just be the most generic shooter combat ever.
Settlement building? Again Fallout 4, but again they dumbed it down so much it's almost unrecognizable.
Story? Skyrim. I know that's half a joke, but it's also half not. They really thought a story fit for a medieval power fantasy would work in a game about scifi space exploration. Don't even get me started on how the temples are so much worse than the word walls, I will be here all day.
Likewise, the radiant quest system is also just Bethesda's standard implementation of it with very little expansion or improvement:
Fast travel here. Do one quick thing. Fast travel back. Unlike in Fallout 4, the world did not support it. Their idea of 'dungeon exploration' was reusing a short list of different points of interest, where nothing at all was randomized, not even the placement of items, and so they could not rely on that to keep the radiant quests refreshing. The most fun you can have with them is doing the pirate ones for the Crimson Fleet because it pays better and gets you in trouble with the law, adding some vague semblance of depth.
Starfield did some great stuff, too, like the shipbuilding. I really loved that! And while the actual basebuilding was very lackluster, being able to make your own base on an alien world IS pretty cool. Being able to board and commandeer ships is also such a nice touch, and something we couldn't have back in the days of Freelancer.
It's just a shame they couldn't make the rest of it work. I would understand if it went through development issues, as all games do, but interviews with the team after launch painted them as surprised of its poor reception, so there's gotta be something they're missing.
@Anubis1101 yeah i understand some of your points, its also weird how many mechanics are locked behind the skills. Like piloting with no skills invested or crewmember with those skills is a pain. There are entire UI elements hidden behind the higher tier skills.
I remember my friend was streaming him playing it in discord and I asked him what UI mod he was using. There was no mod he just invested in the stealth perks.
@@mikeuniturtle3722 Absolutely, I think they've always had a hard time making interesting perks, but they made it even worse in Starfield. Hiding the UI elements behind perks was wild. Their stealth system didn't get any upgrades either, which I've seen from more than a few reviews absolutely killed any interest in it for them.
The one that really got me is what Freelancer called "engine killing", where you cut your engines and keep your momentum. It made for some really cool maneuvers in combat like strafing or spinning around. But in Starfield, you need a perk in piloting to be able to do something that should be a mechanical function of the ship you're flying. It's inherently an advanced maneuver, but it's not something that would clutter the learning experience for new players, so they should've left it in.
I really wish I could sit down and talk game design with Todd or other members of the team; I really am legitimately curious about their thought process behind these things, because I cannot fathom it.
@cainongames “Hey Tim, I’m designing an AI-driven strategy RPG where the player’s real-world growth (cognition, fitness, and discipline) serves as the core progression system. Instead of XP or grinding, the player’s actual self-improvement triggers in-game challenges, forcing them to overcome Olympian trials that scale dynamically. The AI acts as a Dungeon Master, adapting the difficulty to match real-world skill-building. Would you classify this as a turn-based strategy RPG, a digital tabletop RPG, or something else?” I’ve made comments on your other videos but I’ve never been close enough to gamify my system so now I have questions lol
It sounds like you’ve already classified it as a strategy RPG. I’m not sure the exact RPG subgenre matters. Players will like a good game regardless.
@ the reason I’m asking is because I modeled it after the pipboy from fallout and reverse engineered the special framework from the game into real life application, I have a fallout question who wrote the RobCo lore for the game? Specifically the lore for the pipboy creation and its following successes? I’m sort of walking a similar line here
@@CainOnGames ruclips.net/video/uoIGG27_MJM/видео.htmlsi=8FGwxVL6t1DG_qss
@@SideQuestPress It's very hard to determine credit for something more than twenty-five years later. My notes don't indicate a creator for RobCo lore or the pipboy. The former was probably Chris Taylor, while the latter was likely Leonard Boyarsky, but I have no doubt that other designers and artists had input into both things.
@@CainOnGames Aye captain thanks for the heading! One more question before I fade into the background since you’ve worked on RPG progression systems for years, I’m curious-what’s your take on AI-driven adaptive progression replacing XP?
i sort of created an entirely new genre of game my AI Dungeon Master, instead of grinding, players improve real-world skills (cognition, fitness, etc.), and the game world adapts dynamically. Do you think an adaptive system like this could enhance RPG immersion, or do players generally prefer the predictability of XP?
Do you ever do conventions or mail-in autographs? I’d die to get your auto on my Fallout poster
There's something about the concept of dynamic quests that just turns me off. I don't know if it's the knowledge that a quest isn't hand crafted by a real person, or maybe the knowledge that the quests will be infinitely created for me and never end, but something about them just feels hollow and meaningless. I don't mind them as simple extra tasks in a gameplay loop as it's always nice stack more objectives/rewards into an area, but when I'm engaged in dialogue with an NPC and I realise the thing they are talking to me about is just a dynamic quest I'm completely taken out of the experience. I think I can tolerate them better if they kind of have a light shone on them saying "hey this is the dynamic quest system" in something like a bounty board.
Same here, they're not just uninteresting, they actively break immersion for me. One part of it is that it makes you watch as the sausage is made/shows you the mechanism that enables a magic trick/etc, iow it screams "you thought you were in this world? no, you're playing a game!" Another is that they have no impact on the game world, which makes them inconsequential and unengaging (and that happens with filler/trash quests for me even if they were made by humans).
As a player I've never really appreciated radiant (dynamic) quests. They've always felt like busy work and have never been particularly memorable.
A problem I have found with radiant quests is they never reflect changes in the game world that are a consequence of player choices/actions and this is really just a part of the greater problem that nothing ever really changes whatever the player does, despite what the marketing dept., promises. You know - “Shape the future of Tralfamadore etc etc”.
My question is, as we now live in a world of huge day one updates, would it be possible for a game to be updated mid play through to reflect choices made by the player. I realise that I have probably constructed a pipe dream during my sleepless nights and economics will make it a non starter but it would be interesting to hear what you would do with this. Perhaps it’s been tried before and I’ve missed it.
Helldivers 2 and Dwarf Fortress do that to a degree. Helldivers 2 is collectively driven by the players' achievements, coordinated by the dev team. It's a really interesting social experiment!
Dwarf Fortress records all major events, especially those of the player. When an artifact is crafted, or a named character slain, that information is immortalized in the associated cultures, and may be recounted by storytellers and artisans at any later point! Kruggsmash has a video called Legendary Stories where he explores just how deep the historical records in DF go, it's a really cool video!
You can also look at games like Kenshi that have dynamic world states- an almost "questless design", if you will, that respond dynamically to the player's actions and other incidental things. If you kill a faction leader, that area collapses and gets taken over by neighboring factions. It takes some mods to really expand on that system, but it's proof it can be done without a AAA budget or large team of people.
But, overall, I 100% agree, I want more games that change the world in response to the player's actions. It's one of the central pillars of my own games, and I hope that as AI proliferates, we'll see it used for actual cool stuff like this, instead of cheating artists out of work or hallucinating about the news.
You can create a dynamic world very easily in data and mechanics. The problem is graphics/sounds not being easily generatable. The second issue is that handcrafted, rather than pure procedural, content based worlds are fake. There's not enough characters to compare to real life, there's no economy or politics. There's nothing for your decisions to propagate through. You'd have to hand script any change the player made.
But you could easily make a map and menu rpg that allows real choice and consequence.
if i could i would programm a mix between shadowrun/cýberpunk2077 , fallout shelter stýle, endless version.. 😊👍
Alright, as a player, please don't do this in the end game, at least not infinitely. All of the engagement with any given location or NPC has been had, this just looks and feels like make-work.
Also, on the escort mission: walk speed or run speed, not in-between. That's around 60-80% of the frustration there. It's fine if the NPC stops if the player gets too far ahead, as long as nothing spawns on them when they're stopped.
That's all I had to say.
How do you make an item and inventory system not feel like one huge chore? I find the classic CRPG model exhausting: pick up everything irrespective of whether the item is interesting or valuable cuz why not maybe you’ll need it?, constantly have FOMO about whether your gear is best-in-slot cuz there are literally hundreds of pieces of gear for each slot, get encumbered and have to waste time dropping or selling useless stuff, have no idea how to find something you need within the hundreds of items in your inventory, etc. Do item and inventory systems need to be this bloated to make them fun and interesting?
You have identified the crux of the problem, and that is every solution has detractors.
Make a game with just a few items? Players complain that there are no good items for their build.
Restrict the builds? Players complain they cannot make the character build they want.
Add more items? Players will complain about all the useless items for their build, or finding the "good" items among the hundreds they are carrying (as you point out above). And when players try to sell the items they don't want, they will complain the economy is out of balance (if they can sell all the items and now have lots of gold) or that shopkeepers won't buy their items (if the shopkeepers have limited money or restrict the type of items they buy).
Add encumbrance to limit items carried? Players complain about not being able to pick up what they want, or having to waste time dropping or selling the items they don't want.
Even if you handle everything using system mechanics that many players like (with the caveat that you will never make all gamers happy), you may still not produce a UX that players like, and UX can be even more subjective than mechanics.
The solution is to decide on what is best for your game based on the design goals you set before making any setting, story, or system mechanics. There are MANY videos I made on these topics, including encumbrance, economy, and writing design specifications.
@@CainOnGames Thank you, sir, for such a thoughtful response!
So you would basicaly design and implement dynamic quests like they are already designed in yours or other games... radical i must say. You described one major flaw players have with such quests - rigidity. You seem to folow clear idea what and how should things be done. I think its time to re-define few things, like for example that you should be able to do such quest (lvl. req.) Why? Its random quest, if it suppose to feel "natural" you sometimes get screwed. This need to "not be too harsh" on players and in the same time being harsh in other areas (like inventory space) creates disconnect. I guess you are the game developer but there might be reason why gamers see it just as product, stale and predictible product.
Id rather a significantly smaller game with none of these. Theyre filler content for empty or boring locations that should just be cut entirely.
Rather than make the world 4x bigger or 16x more detailed.
Out of things to (rightfully!) object to with the current proliferation of generative software, I find it very odd for people to be conceptually opposed to more dynamic possibilities in RPG quest branching thanks to the ability to generate new VO more readily. Two of the biggest ethical issues with recent year's new generative software, source consent and labor undercutting, are easily addressed here by getting the consent of the original actor(s) and still paying them for the use of those original recordings. Does anyone in these comments have other ethical concerns around it that I'm simply missing?
I think those are the major ethical concerns, and I agree they could be addressed. Many people don’t like the sound of AI voice or don’t want generated dialog on NPCs, but those aren’t ethical considerations.
One of the issues is that every model has already trained using data without consent of the voices. There's just no "pure" system out there, it's already been poisoned with plagarism and theft
@@KynElwynn Is it true that all computer VO is trained on human actors? I used a system more than 8 years ago (probably closer to 10) that was purely synthetic. It exposed controls for pitch, speed, phoneme variation, acoustic variation, etc., so you could make a bass-voiced American man in an office or a soprano-voiced Australian woman in a cave, and there was no connection to human training sets. Have all of these VO models been abandoned?
@@CainOnGames Older purely-synthetic systems are really hard to find and considered obsolete in favor of the new shiny generative/LLM-based AI voices, sadly. If you know any good ones, though, I'm sure your reach would allow them to be brought back into the spotlight!
I know there are some AI projects that are trying to build "proper" databases with non-stolen training data, but I'm not sure how good they are for voices at this point.
You can do a lot with a singular voice too. I make electronic music as a hobby, or more specifically I really enjoy the process of sound design. I have taken my voice, an American male, and simply loaded it as a file into massive(big synth plugin) and have made myself sound as many other things, even a fairly convincing British woman.
Now this does take a while so definitely isn't on the industries radar, but ive always thought if I finally took the plunge and made a game that is what id do for any VO needed.
Do you think Avowed will flop?
The reason why many people don't like AI voice acting is not that it's "bad" and deprives them of salaries, but that it "cuts the ear with sound" - over time, you begin to recognize artifacts of sound generation - and AI sound becomes UNPLEASANT. It's just IMPOSSIBLE TO LISTEN TO him! It's the same with AI pictures)
please remove those quests from your games! thank you, they add nothing. Hand crafted quest is where it's at... Well at least if you want BG 3 sales...
Skyrim's Thieves Guild's repeatable "forge the numbers"-quests felt like a simple and fine way to gain reputation/money, until I started working for the national tax agency. Now I feel conflicted about doing them.
I rather would play a different game than play those "dynamic" infinite quests again and again. They are just so boring.
If it's "9-5" quests. If you wanna roleplay as a thief, so the game makes some rutine burglary quests for you, it could be fun, but otherwise I agree. I don't need simple filler quests if there is an actual crafted storyline to follow.
Yea it's heavily dependent on implementation. I think it's a great way to pad out gametime if you want to keep playing a game but have finished all the story and unique content, but the quests need to be involved enough to not just feel like busywork (*cough* Skyrim).
Cyberpunk 2077 is a very rare example of a game with tons and tons of unique content, to the point where you feel like you don't really need dynamic quests, but that of course took actual years of development, and all the financial costs associated with it. Still, I think there's a lot we can learn from how 2077 does its quests, and potentially how to make dynamic quests of a similar style.
Stuff like dynamic quests are basically the only reason AI should have anything to do with game development (not counting standard NPC AI, obviously). Like imagine the potential of functionally unlimited unique quests rather than just pulling from a pool of options. That'd be the dream.
How it all looks in practice: ruclips.net/video/5C3mvRm0aOY/видео.html (:
This whole channel is written in gold filigree 🤌
I've loved the videos on board game content. Would you please do a video about minigames or subsystems that board games might 'rip' from video games? Puzzles, dexterity challenges, etc.
Thanks so much 👌