Rogue Drones or, Adding More Game-Driven Content

[Recatek]

UPDATE: I posted some additional notes and thoughts/clarification later in the thread, you can find that post here!

Starbase is lacking in activity variety. For the most part, the bulk of the time an average player spends in the world can be divided into mining, building, and PvP (which is mostly searching, not fighting). Compare this with other games in the genre, and the breadth of practical activities is very limited. This, for the most part, hinges on the world and the minimal variety of stuff out in it -- it's all really just rocks and gas. I think this needs to change.

Rogue drones to the rescue!
Rogue drones are a faction of malfunctioning non-endo AI constructs that serve as NPC adversaries. Like resources in space, rogue drones are more frequent/active/dangerous in some regions than others. Fundamentally, they provide player objectives or points of interest (POIs), add risk to the world, and introduce an element of exploration, discovery, and rarity to an otherwise rock-and-gas universe. These occur in various different forms, including:

• Rogue Drone Derelict: Inert discoverable shipwrecks that spawn randomly in the world and can be encountered and salvaged. May or may not be detectable from far away by their radiation signature. May or may not have still-active autoturrets posing a threat to would-be salvagers. Can be salvaged for normal ship parts and potentially a minor reward.
• Rogue Drone Caravan: A collection of transport ships full of materials with autoturret (or autonomous ship, if possible) defenses. Can be discovered or detected via radiation signature. When attacked, the caravan broadcasts a chat notification and a trackable SOS signal to all players in the region and begins to flee in an arbitrary direction until destroyed or lost. May be carrying a minor reward.
• Rogue Drone Complex: A labyrinthine station-like complex with autoturret (or autonomous ship, if possible) defenders. These are telegraphed POIs (see below) that provide an incentive for multiple players to arrive at the same time and fight each other while making/blasting their way through the structure for a major reward deep within.

The possibilities expand from here certainly, but these three are intended to get the most reuse out of mechanics Starbase already has in-game.

Telegraphed POI? What's that?
For PvP to occur, at least two opposing players must be in the same spot at the same time. In a world as large as Starbase's, this is difficult to engineer. Looking to other games for inspiration, I believe the best way to make this happen on a consistent basis is by focusing on four key ingredients:

1. Players must be incentivized to go to a location with a tangible reward for success.
2. Players must know exactly (or very nearly) where that location is.
3. This incentive should have a limited time window, so that players need to arrive at the same time.
4. This incentive must be announced to players well before it happens, to give them time to prepare and travel to the destination.

Item 4 is where telegraphing comes in. The game must make it clear, well ahead of time, that a high-value POI like a rogue drone complex is going to enter the playing field. Travel in Starbase can take hours, and players aren't always in the right ship for the right activity. There are plenty of ways to contextualize the telegraphing portion of this -- imagine a news terminal in a given station that reports sightings of a rogue drone complex under construction and when its shields will be vulnerable to attack. But the advance notice must come hours or even days ahead of time. Ideally I would know, when I log off of the game tonight, that I should be on by tomorrow night at 7pm to do the rogue drone complex (or whatever else) with my guildmates. Lining up a variety of different objectives in a week can give players something to do every day they play the game, keeping them in the world and engaged with the economy and playerbase rather than sitting in the SSC or, as is currently the case, simply not playing Starbase.

Wait, go back, you said something about rewards?
I sure did, and this is critical. Rogue drones can, must, add flavor and rarity to the world. Right now, ship parts are ship parts. Some require rarer minerals, but for the most part, they aren't too special. Imagine that, at the core of a rogue drone complex, you have a chance to find a blueprint chip for a rare overclocked reactor -- good for 50 printed copies in a factory. Trade goods? You bet. Sell the chip, or run the blueprint through your factory and sell the reactors directly. Or risk hauling them to a more popular market to put them on the auction house there. There are plenty of opportunities for what an NPC adversary can add to the world. Rare dye colors for ships, player cosmetics, overcharged weapon ammunition, and so on. Things that make the world and the things out in it worth fighting and exploring for, with flavor and variety besides just rocks and gas.

This sounds gamey and artificial! What about a player-driven world?
There are two realities here that it's time to confront.

First, players can't be expected to provide all the content and activities for other players to do. Territory control and sieges are great, but sieges won't happen every day, and each one comes at a significant cost along with a demoralizing penalty for the loser. Sieges and territory conflict are not an activity you can expect to engage in every time you play Starbase; they should be one of many activities you should have available to you when you log in. Similarly, there are organizers who go out of their way to create content with staged battles, tournaments, and other events, but burnout sets in quickly and the job is thankless. Community events, be they player- or dev-driven, are also not a sustainable activity source. The game itself needs to do more work here.

Second, many fun encounters are asymmetrical. It's way more fun to stalk and blow up a trade ship than it is to have your trade ship stalked and blown up. Players don't always want to be on the butt end of someone else's fun. But do you know who doesn't care about whether or not they're having fun? NPCs don't care. NPCs are the perfect way to give players the fun of doing a thing that isn't fun to have done to you. They don't get bored sitting in a turret all day waiting for someone to show up. They don't mind driving a ship full of goodies from place to place all day just to be attacked. They don't rage-uninstall when someone blows up the station they spent days building. Is it as fun to blow up an NPC ship as it is to blow up a player's? Certainly not, but it can still be pretty fun, it's definitely better than nothing, and it means there will be more players out there in space to blow up.

As a game, Starbase needs to provide more things for its players to do. It especially needs to create activities that don't rely on players doing all the grunt work and absorbing all the costs and losses. Game-driven content creates more player-driven opportunities by keeping players around and directing them to locations where they're likely to encounter one another, or can still have fun even if they don't. It's a net positive, even if it comes at the cost of adding more "game" to the game. The best sandbox games aren't empty, system-less worlds where players provide all the content. They're deep, system-rich games that embrace and amplify all the different activities that the players want to engage in together. They provide content and goals for many different pillars of gameplay.

Okay, but why rogue drones?
Because they're weird and crazy! No, seriously. AI and NPCs are difficult to add to any game, let alone Starbase. Players are likely to be more forgiving of dumb AI if they're presented as malfunctioning AI in-world. Sure, yeah, that rogue drone ship just crashed into an asteroid. Oh well, I guess that's what you get when your circuitry is going haywire. Rogue drones are the perfect narrative conceit to get away with mistakes in the AI and NPC development process. They aren't endos, and don't need to act like we would. A rogue drone ship can just be a ship with no pilot -- the drone just is the ship, with some AI core box buried in the circuitry (valuable loot, perhaps). Same with turrets and other rogue drone structures. You don't need NPC endos at all to add AI adversaries to Starbase.

Another advantage of using malfunctioning AI as a narrative conceit is that they're strange, mysterious, and inscrutable. They can't be expected to act in rational ways, which gives designers more creative freedom in how to use them for gameplay. Imagine a multi-act game-wide story event where the drones are discovered to be building a temple of sorts to... something, and the whole playerbase needs to go stop it in time. Would that seem strange or out of place to you? Probably not -- they're crazy rogue drones, who knows what they're thinking or why they do things. You can really get away with a lot while still staying faithful to Starbase's robotic aesthetic and fantasy.

Now, does any of this require the antagonist to be rogue drones? Of course not, but I think they would fit well with the setting and allow for a lot of creative flexibility. I think they're far more thematic to Starbase than aliens or organic enemies would be, and trying to add enemy endo NPCs would bring far more expectations for polish and fidelity for limited payoff. Of these options, rogue drones seem to me like the best way to get the widest variety of NPC activities with the least risk or investment.

---

Anyway, that's what I think the game should focus on after its current major roadmap items are finished. I eagerly await the replies telling me to go back to WoW.

 

[Lukas04]

Definitly agree. The game just needs more Parts that arent designed by players alone.
The nice part about good POIs is that the game creates an event, but how it goes down is up for the players to decide.
Maybe you fly to a POI and find a faction of players guarding it, so that only their salvagers get any of the Loot.
You then get to decide how you deal with that Situation.

Drone encounters, even if relatively easy, would still make Space Travel more interesting and would add a feeling of being watched, which keeps you on your toes.

 

[Aha]

I absolutely love your idea, especially as I had the same idea for the NPC as being malfunctioning robots, even endos. My main idea about it and I think it would fit perfectly into your concept is that every ship that would be despawning from the world (clean up) would add to the armies of the NPC malfunction faction being restored by their blueprints and the restored missing parts could even have some unique colors to it. Therefore anything left out there gets "scavenged" by the NPC sending them back against the players! You could face your own ship if you are (un)lucky enough! The more players the game has, the more ships to be cleaned up, the bigger the NPC army is. It would need a lot of work, of course, it's just a concept.

 

[CalenLoki]

I like the idea of NPCs. Even if I was very much against for a long time.

I could see it expanded a bit to server more purposes:

Rouge drone patrol: numerous weak space Toyotas flying around in search for easy prey. Equipped with hand weapons, they wouldn't present much of a threat to anyone prepared. But if you decide to fly unplated, unarmed solo miner outside SZ they may do some harm. They would spawn very commonly, so it wouldn't be about "if" you meet them, but "when and how many". They'd be more common far from active player stations
Goals:
-encourage plating and arming miners and bring gunners/escort.
-Make miners better prepared for combat, so when they meet player pirates they aren't so helpless.
-Encourage building stations close to civilisation

Not all NPC encounters have to be opt-in. Some may come without asking.

 

[Recatek]

Yes, absolutely! There are tons of possibilities. It depends on how far the AI tech gets in terms of combat competency though. Right now I figure if players can make YOLOL autoturrets, then more accurate and dangerous game-driven ones aren't too far out of the picture. Caravans that fly from point to point and avoid asteroids don't seem too unreasonable either. Full-fledged moving and shooting enemy ships are probably a different matter entirely though, so I avoided getting too deep into that design space for starters.

 

[Busdriver]

Peer to peer is a problem if we want NPCs to do anything past 3km, but I do hope for something like this.

 

[Cavilier210]

No. Just outright no.

 

[LoPing]

Starbase is lacking in activity variety. For the most part, the bulk of the time an average player spends in the world can be divided into mining, building, and PvP (which is mostly searching, not fighting). Compare this with other games in the genre, and the breadth of practical activities is very limited. This, for the most part, hinges on the world and the minimal variety of stuff out in it -- it's all really just rocks and gas. I think this needs to change.

Rogue drones to the rescue!
Rogue drones are a faction of malfunctioning non-endo AI constructs that serve as NPC adversaries. Like resources in space, rogue drones are more frequent/active/dangerous in some regions than others. Fundamentally, they provide player objectives or points of interest (POIs), add risk to the world, and introduce an element of exploration, discovery, and rarity to an otherwise rock-and-gas universe. These occur in various different forms, including:

• Rogue Drone Derelict: Inert discoverable shipwrecks that spawn randomly in the world and can be encountered and salvaged. May or may not be detectable from far away by their radiation signature. May or may not have still-active autoturrets posing a threat to would-be salvagers. Can be salvaged for normal ship parts and potentially a minor reward.
• Rogue Drone Caravan: A collection of transport ships full of materials with autoturret (or autonomous ship, if possible) defenses. Can be discovered or detected via radiation signature. When attacked, the caravan broadcasts a chat notification and a trackable SOS signal to all players in the region and begins to flee in an arbitrary direction until destroyed or lost. May be carrying a minor reward.
• Rogue Drone Complex: A labyrinthine station-like complex with autoturret (or autonomous ship, if possible) defenders. These are telegraphed POIs (see below) that provide an incentive for multiple players to arrive at the same time and fight each other while making/blasting their way through the structure for a major reward deep within.

The possibilities expand from here certainly, but these three are intended to get the most reuse out of mechanics Starbase already has in-game.

Telegraphed POI? What's that?
For PvP to occur, at least two opposing players must be in the same spot at the same time. In a world as large as Starbase's, this is difficult to engineer. Looking to other games for inspiration, I believe the best way to make this happen on a consistent basis is by focusing on four key ingredients:

1. Players must be incentivized to go to a location with a tangible reward for success.
2. Players must know exactly (or very nearly) where that location is.
3. This incentive should have a limited time window, so that players need to arrive at the same time.
4. This incentive must be announced to players well before it happens, to give them time to prepare and travel to the destination.

Item 4 is where telegraphing comes in. The game must make it clear, well ahead of time, that a high-value POI like a rogue drone complex is going to enter the playing field. Travel in Starbase can take hours, and players aren't always in the right ship for the right activity. There are plenty of ways to contextualize the telegraphing portion of this -- imagine a news terminal in a given station that reports sightings of a rogue drone complex under construction and when its shields will be vulnerable to attack. But the advance notice must come hours or even days ahead of time. Ideally I would know, when I log off of the game tonight, that I should be on by tomorrow night at 7pm to do the rogue drone complex (or whatever else) with my guildmates. Lining up a variety of different objectives in a week can give players something to do every day they play the game, keeping them in the world and engaged with the economy and playerbase rather than sitting in the SSC or, as is currently the case, simply not playing Starbase.

Wait, go back, you said something about rewards?
I sure did, and this is critical. Rogue drones can, must, add flavor and rarity to the world. Right now, ship parts are ship parts. Some require rarer minerals, but for the most part, they aren't too special. Imagine that, at the core of a rogue drone complex, you have a chance to find a blueprint chip for a rare overclocked reactor -- good for 50 printed copies in a factory. Trade goods? You bet. Sell the chip, or run the blueprint through your factory and sell the reactors directly. Or risk hauling them to a more popular market to put them on the auction house there. There are plenty of opportunities for what an NPC adversary can add to the world. Rare dye colors for ships, player cosmetics, overcharged weapon ammunition, and so on. Things that make the world and the things out in it worth fighting and exploring for, with flavor and variety besides just rocks and gas.

This sounds gamey and artificial! What about a player-driven world?
There are two realities here that it's time to confront.

First, players can't be expected to provide all the content and activities for other players to do. Territory control and sieges are great, but sieges won't happen every day, and each one comes at a significant cost along with a demoralizing penalty for the loser. Sieges and territory conflict are not an activity you can expect to engage in every time you play Starbase; they should be one of many activities you should have available to you when you log in. Similarly, there are organizers who go out of their way to create content with staged battles, tournaments, and other events, but burnout sets in quickly and the job is thankless. Community events, be they player- or dev-driven, are also not a sustainable activity source. The game itself needs to do more work here.

Second, many fun encounters are asymmetrical. It's way more fun to stalk and blow up a trade ship than it is to have your trade ship stalked and blown up. Players don't always want to be on the butt end of someone else's fun. But do you know who doesn't care about whether or not they're having fun? NPCs don't care. NPCs are the perfect way to give players the fun of doing a thing that isn't fun to have done to you. They don't get bored sitting in a turret all day waiting for someone to show up. They don't mind driving a ship full of goodies from place to place all day just to be attacked. They don't rage-uninstall when someone blows up the station they spent days building. Is it as fun to blow up an NPC ship as it is to blow up a player's? Certainly not, but it can still be pretty fun, it's definitely better than nothing, and it means there will be more players out there in space to blow up.

As a game, Starbase needs to provide more things for its players to do. It especially needs to create activities that don't rely on players doing all the grunt work and absorbing all the costs and losses. Game-driven content creates more player-driven opportunities by keeping players around and directing them to locations where they're likely to encounter one another, or can still have fun even if they don't. It's a net positive, even if it comes at the cost of adding more "game" to the game. The best sandbox games aren't empty, system-less worlds where players provide all the content. They're deep, system-rich games that embrace and amplify all the different activities that the players want to engage in together. They provide content and goals for many different pillars of gameplay.

Okay, but why rogue drones?
Because they're weird and crazy! No, seriously. AI and NPCs are difficult to add to any game, let alone Starbase. Players are likely to be more forgiving of dumb AI if they're presented as malfunctioning AI in-world. Sure, yeah, that rogue drone ship just crashed into an asteroid. Oh well, I guess that's what you get when your circuitry is going haywire. Rogue drones are the perfect narrative conceit to get away with mistakes in the AI and NPC development process. They aren't endos, and don't need to act like we would. A rogue drone ship can just be a ship with no pilot -- the drone just is the ship, with some AI core box buried in the circuitry (valuable loot, perhaps). Same with turrets and other rogue drone structures. You don't need NPC endos at all to add AI adversaries to Starbase.

Another advantage of using malfunctioning AI as a narrative conceit is that they're strange, mysterious, and inscrutable. They can't be expected to act in rational ways, which gives designers more creative freedom in how to use them for gameplay. Imagine a multi-act game-wide story event where the drones are discovered to be building a temple of sorts to... something, and the whole playerbase needs to go stop it in time. Would that seem strange or out of place to you? Probably not -- they're crazy rogue drones, who knows what they're thinking or why they do things. You can really get away with a lot while still staying faithful to Starbase's robotic aesthetic and fantasy.

Now, does any of this require the antagonist to be rogue drones? Of course not, but I think they would fit well with the setting and allow for a lot of creative flexibility. I think they're far more thematic to Starbase than aliens or organic enemies would be, and trying to add enemy endo NPCs would bring far more expectations for polish and fidelity for limited payoff. Of these options, rogue drones seem to me like the best way to get the widest variety of NPC activities with the least risk or investment.

---

Anyway, that's what I think the game should focus on after its current major roadmap items are finished. I eagerly await the replies telling me to go back to WoW.

Rouge Drones? Play Eve much?? But on the whole I agree with you Starbase is all but Dead, I dont see how anyone can justify the cost of development with 280 -500 peek time players. Hope some stuff happens before they stop beating this dead horse.

 

[LemonGrab]

I like this idea, and I think it could even help with the tech tree, by giving you points for salvaging those rare parts the rogue drones carries

 

[CalenLoki]

Peer to peer is a problem if we want NPCs to do anything past 3km, but I do hope for something like this.

The question is: do we?
They could be run entirely client side, and exist only when someone sees them. Or at most be just few numbers when there's nobody around.
Just like asteroids, just trying to kill you. Actively.

 

[Cavilier210]

Now think... How does this in any manner mitigate the already existing push to only fly max speed? People aren't going to fight these. They'll bot farm, or fly past and away as quickly as possible. It doesn't actually add anything to the game, while costing considerable resources.

 

[Ex-Executioner]

The question is: do we?
They could be run entirely client side, and exist only when someone sees them. Or at most be just few numbers when there's nobody around.
Just like asteroids, just trying to kill you. Actively.

Exactly, no need to spend resources on it. They don't need to do anything while no player is remotely nearby.

 

[mikan]

Once NPCs are added, you may encounter NPCs that are okay to hit the asteroid, or NPCs that keep pushing the asteroid. That process can cause another bug

 

[ChaosRifle]

Similarly, there are organizers who go out of their way to create content with staged battles, tournaments, and other events, but burnout sets in quickly and the job is thankless. Community events, be they player- or dev-driven, are also not a sustainable activity source.

Been running weekly gamenights, endoboxes/corpses makes doing anything infantry as an event organizer extremely difficult, due to the nature of no inventory boxes to haul extra gear with us for when its lost to death, or just pause the whole match, let the one guy who died grab his gear, then move the station. its pretty rough. You nailed this post, and I can personally atest to this quote.


As for AI, Star Citizen AI is literally:
Always pitch/yaw to point at player
1: IF :distanceToTarget > 400 THEN :FcuForward=100 END
2: IF :distanceToTarget > 250 AND :distanceToTarget < 400 THEN :FcuForward=0 :FcuUpDown=100 END (technically it chooses 1 of 4 directions, up down left right and sticks with it for a set timer before changing)
3: IF :distanceToTarget < 250 THEN :FcuBackward=100 END

This is more of a pseudocode for yolol (and has states get set but never unset, like fcubackward). Done this way so you can see just how basic their AI is. this is genuinely the extent of starcitizen AI, and most people dont complain about it because they dont toy with it to see how basic it is, they attempt maneuvers and that makes the AI behave less obviously due to the changing ranges the player chooses. personally as an avid PVP'er its total trash AI, but hey, most people are okay with it and TBF, its infinitely better than no AI at all.

 

[Recatek]

Just to reiterate what I said above,

that's what I think the game should focus on after its current major roadmap items are finished.

Far from being a "shit" experience, I think all of these things would really be quite fun, especially when combined with other activities to do in the game world. I doubt I'm alone in that either. Considering all the support and other plausible activity ideas this thread has already generated, the playerbase desire for something like this is demonstrable. Starbase sorely needs mechanics that add life to its world, and that don't rely exclusively on player initiative to exist. I think adding game-driven activities and rewards to beckon players out into the game world is the best way to do that, and to create organic content from players bumping into one another in the process. I also think it would be especially nice if the catalysts for these activities were something other than the same old rocks and gas.

EDIT: The post this is in response to has since been deleted, so what I wrote probably doesn't make sense without that context.

 

[LauriFB]

I think there is a theoretical many people wins -spot in AI's: Source ships and possibly route/function design to players, and we would focus on the required AI functionalities and tech side. But like always, anything like this would require huge amounts of development from our side as well. From top of my head, there's the following things to tackle (at least):

• Risk vs. reward balance (are the ships just free loot, or would they use some sub-grade materials to cap the possible gains?)
• Combat balance (is it fun, or just one simple trick every time)
• Technical side (we have a lot of base work this, but a lot more would be needed)

Development-wise, player-generated ships with simple job AI's could be most likely easier to evolve into meaningful content instead of solely dev-created stuff. In best case the building and designing of these ships would provide continuous meaningful content for the ship designers as well. If the ship designers could be incentivized to create better content (like with a small reward every time a AI ship/convoy reaches their goal) it could drive into ever-evolving challenges for the attackers. Note that a playerless endo manning a tripod or handling simple navigation most likely isn't much of an issue, they both are likely pretty straightforward for development. It's all the other core tech which can turn into a lot of work easily.

There's also hybrid model possibility for additional content and challenges: the AI's will drive their ship/convoy happily, but once it's attacked the AI spots can be taken over by players who want their free five minutes of action. This would twist the attacking challenge not to be trivial, and the defending side would get some risk-free action. But like always, the balance can also tilt dramatically in these situations. Hybrid model would also help ship designers to get instant feedback on their convoy ships.

Note that all of this is purely speculative at this point.

 

[Ventus]

I'm onboard for NPC's now too, done right. I think they could add more immersion to the universe, but not in a lore sense.
Encountering players outside of the inner Eos belt was extremely rare even in the first few weeks of EA launch (Excluding the Moon Gate, but it also has the same SZ properties the Eos Pringle has). NPC's would of course easily fix that, making the universe seem much more inhabited or at the very least expanding.
It'd also feed into the roleplay-style a lot of players approach the game with. Player factions tend to be... 4th wall shattering on most encounters in-game, which is in their nature. Unless they join a faction that actively seeks to roleplay inside the universe, a Solo player would only be able to watch Rock Smacking: The Movie for so long before losing the "Pioneer" feeling of exploring and mining in the belt.
Now seeing a civilian miner pass them by, mining rocks with them? That opens up a few interactions. Trading, pirating, or obtaining information about the area from the NPC could go a long way as far as depth is concerned.
I could easily say I'd still be actively playing if I had a solid chance of encountering at least 1 ship outside of the Origin pringle, no capital ships needed.

 

[IronGremlin]

Just to reiterate what I said above,



Far from being a "shit" experience, I think all of these things would really be quite fun, especially when combined with other activities to do in the game world. I doubt I'm alone in that either. Considering all the support and other plausible activity ideas this thread has already generated, the playerbase desire for something like this is demonstrable. Starbase sorely needs mechanics that add life to its world, and that don't rely exclusively on player initiative to exist. I think adding game-driven activities and rewards to beckon players out into the game world is the best way to do that, and to create organic content from players bumping into one another in the process. I also think it would be especially nice if the catalysts for these activities were something other than the same old rocks and gas.

Yeah, ok, waiting until this roadmap is done solves like 1/5th the issue here, but that is admittedly somewhat less insane.


The core problem with this suggestion, and the dozens of others like it, is that you basically end up arriving at an idealized version of space engineers PvE, which is absolutely terrible from a reward curve perspective.



That's an issue with a largely single player game, it's a death knell for an MMO.


The solutions to this, such as they are, all violate core design elements - you make ships artificially cheap because they're AI controlled, introduce weird junk materials, etc, which means you've broken the consistency of the world - and now you've failed as a building sim instead of failing as an MMO.



Even the alternative/compromise presented here by Lauri doesn't completely overcome that core contrast - AI will always be exploitable, which means introducing automated combatants will always screw with your risk/reward curve.


Either you find a way to artificially limit what players extract from those encounters, which means you now have a near permanent dev time sink constantly tuning the PvE experience to stay relevant with shifting patterns of player behavior / feature content, or you treat those ships like any player ship and effectively turn the new primary method of acquiring resources from mining to rogue AI ship-jacking - which like, cool, I guess, but completely different game, and designing that would be pretty non trivial to do in a way that kept any of the current design goals re: resource distribution and concentration.


Tying the operation of drone ships to a CURRENTLY online and proximate player is basically the only way around the problem, as only a human will be able to prevent another human from exploiting the AI for EZ loot - and that still opens a massive can of worms that would permanently alter the landscape, because now all combat has to be rethought in terms of drone forces.


So it's not even just a matter of finishing the roadmap, it's like finishing the roadmap and letting it bake for 6 months or so in order to get some idea as to whether or not this is conceptually viable at it's most basic level.

 

[Askannon]

I think, instead of introducing AI convoys, why not create a reason for player convoys.
We have the Inbox already and the system behind market prices could be adjusted to allow temporary intervention, so calling for a day of special offers at trading terminals could work in incentivising flying to and from certain stations (if they have those) in a concentrated manner.

This would force gathering of materials in certain zones to have enough ready to profit from the events.
It would incentivise players to fly to that station, making their position predictable and increase density of players around that station... prime targets for PvP players (especially if they have their cargos full with valuables). And to balance their risks the monetary advantage should be considerable (enough to consider having people to defend the convoy).

Now, to avoid it becoming too much (be it invested time of devs or players growing bored) you just have to do what the oaks do: don't do it too often or the players won't want to do it anymore with enough force to make the event interesting (which would also reduce the needed dev time to manage the market).

Admittedly, this is an idea how the devs could promote interaction, when the players don't have the reason to do (yet).
It could even be explained as stations wanting to expand and needing resources to do so instead of magically growing ^^.

 

[Recatek]

I think there is a theoretical many people wins -spot in AI's: Source ships and possibly route/function design to players, and we would focus on the required AI functionalities and tech side. But like always, anything like this would require huge amounts of development from our side as well. From top of my head, there's the following things to tackle (at least):

• Risk vs. reward balance (are the ships just free loot, or would they use some sub-grade materials to cap the possible gains?)
• Combat balance (is it fun, or just one simple trick every time)
• Technical side (we have a lot of base work this, but a lot more would be needed)

Development-wise, player-generated ships with simple job AI's could be most likely easier to evolve into meaningful content instead of solely dev-created stuff. In best case the building and designing of these ships would provide continuous meaningful content for the ship designers as well. If the ship designers could be incentivized to create better content (like with a small reward every time a AI ship/convoy reaches their goal) it could drive into ever-evolving challenges for the attackers. Note that a playerless endo manning a tripod or handling simple navigation most likely isn't much of an issue, they both are likely pretty straightforward for development. It's all the other core tech which can turn into a lot of work easily.

There's also hybrid model possibility for additional content and challenges: the AI's will drive their ship/convoy happily, but once it's attacked the AI spots can be taken over by players who want their free five minutes of action. This would twist the attacking challenge not to be trivial, and the defending side would get some risk-free action. But like always, the balance can also tilt dramatically in these situations. Hybrid model would also help ship designers to get instant feedback on their convoy ships.

Note that all of this is purely speculative at this point.

Yeah, ok, waiting until this roadmap is done solves like 1/5th the issue here, but that is admittedly somewhat less insane.


The core problem with this suggestion, and the dozens of others like it, is that you basically end up arriving at an idealized version of space engineers PvE, which is absolutely terrible from a reward curve perspective.



That's an issue with a largely single player game, it's a death knell for an MMO.


The solutions to this, such as they are, all violate core design elements - you make ships artificially cheap because they're AI controlled, introduce weird junk materials, etc, which means you've broken the consistency of the world - and now you've failed as a building sim instead of failing as an MMO.



Even the alternative/compromise presented here by Lauri doesn't completely overcome that core contrast - AI will always be exploitable, which means introducing automated combatants will always screw with your risk/reward curve.


Either you find a way to artificially limit what players extract from those encounters, which means you now have a near permanent dev time sink constantly tuning the PvE experience to stay relevant with shifting patterns of player behavior / feature content, or you treat those ships like any player ship and effectively turn the new primary method of acquiring resources from mining to rogue AI ship-jacking - which like, cool, I guess, but completely different game, and designing that would be pretty non trivial to do in a way that kept any of the current design goals re: resource distribution and concentration.


Tying the operation of drone ships to a CURRENTLY online and proximate player is basically the only way around the problem, as only a human will be able to prevent another human from exploiting the AI for EZ loot - and that still opens a massive can of worms that would permanently alter the landscape, because now all combat has to be rethought in terms of drone forces.


So it's not even just a matter of finishing the roadmap, it's like finishing the roadmap and letting it bake for 6 months or so in order to get some idea as to whether or not this is conceptually viable at it's most basic level.

So I want to address a few things common to both of these posts, together:

Activities first, AI second
This suggestion thread is primarily about adding game-driven activities in the world in a cohesive and expandable way. Competent combat AI ships are secondary, and nonessential, to that. The three examples I gave intentionally do not require autonomous combat ships. At minimum they require implementation of an AI turret that can track and shoot players and of a ship hull that can fly in a direction and avoid asteroids in the process. Both of these exist to some degree in YOLOL form already, but I want to say here that rogue drones should absolutely, emphatically, not be programmed in YOLOL or limited by its constraints -- they don't have to be, and they would be completely hamstrung if done that way. These drones are game mechanics and really ought to be implemented in game code the way other game mechanics are, which leads into my next point that...

Rogue drones aren't endos
Rogue drones don't need to be bipedal robots walking around and using devices like players do. They don't need their ships to be fully wired up and driven by YOLOL. They don't need to play by the rules set for players. They are entirely different creatures and they aren't playing the game the way we do. If anything, they're more a form of hostile wildlife than a player-like NPC faction. A rogue drone autoturret doesn't need to look or act like an endo on a tripod; it can just be a self-contained turret that shoots at nearby players. A rogue drone ship can just fly on its own without any distinct pilot or accessible script. Rogue drones don't exist to populate the world with fake players, they exist to make the world more challenging and interesting for players. They don't need to use the same devices and logic that player ships do, nor do all their ship parts or weapons need to be player-usable or player-compatible (some should be, but not all). The advantage of using insane, malfunctioning AI with bizarre hardware as a narrative device is that it helps to explain away game balance considerations like making certain ship parts NPC-only ("this device is too unusual to connect to your ship's systems"). This helps to explain to the player why...

Rogue drone rewards are different
The developers can have complete control over how much or how little of a rogue drone ship or installation is salvageable or usable by players. They also have complete control over where and how often these things appear. Compared to the fields of near infinite asteroids full of resources that don't shoot at you, rogue drones wouldn't be a good source of raw resources, nor do they have to be a good source of normal manufactured ship parts. For one thing, rogue drone ships don't even have to have player-usable weapons or devices like MFC/FCU boxes on them, since they play by different rules. There may be some electronics on rogue drone constructions that can taken and used as-is, others that can't be used but instead refine into desirable materials, and some that are simply worthless once removed from their host drone. This still provides rewards for salvage-oriented players (who currently have precious little to salvage on a day to day basis) without hurting industry-oriented players or making rogue drones an easy ship part farm. Rewards-wise, rogue drones should mostly do the things asteroids can't, like provide unique cosmetics, decorative parts, rare limited-run ship part blueprints, and so on. Things with uniqueness and flavor. Maybe even little bits of lore, for people who are into that sort of thing. But most importantly, bulk resource gathering and ship part acquisition will already be well handled by asteroids, moons, and factories, rogue drones won't (and don't need to) compete with this.

---

I really want to underscore the point above about these drones being more like hostile wildlife than fake players. Their sole purpose is to create more fun in the game world. They don't need to worry about furthering their own goals or collecting resources or even having survival instincts. They aren't created by players for minmaxed personal gain on a traditional reward structure. They're created by game developers with a vision to provide interactive activities, to get people to log in to the game every day and get out into space to do things. I don't think player-sourced autoships would be as effective in accomplishing the same goals, especially if they're designed by players who are optimizing to a reward rather than to the attacker's fun. Not to say it's a bad idea on its own, but I think it would yield pretty separate results from what I'm trying to suggest here. Starbase really isn't lacking in things for ship designers to do. If anything, ship design is the game's strongest and most robust pillar. What I find lacking in Starbase is the variety of things for everyone to get out of the ship hangar and go do out in the wild, and I don't believe player-created content alone can sustainably fill that gap.