Draft: NPC Strongholds



  • I like what I see for the most part. Without hard numbers on the rewards for killing it's hard to say. I'm guessing a lot of these are going to get killed at stage one.

    Personally though, I'm happy to see more PvE content.



  • Looks fun, but I know I won't be letting any grow past tier 1 near me 🙂 But maybe in neighbouring sector 😛



  • I absolutly love the idea! This looks like a smart feature to engage most players and lure zombies back to the game.

    I can't wait to try to find the most efficient tactic to tear down their defenses!



  • I don't know whether I would let these grow for the boosts or kill them immediately for less interrupted mining. Seems like there could be two valid answers, depending on player preference, which is good.

    Also this will provide a lot of opportunities for younger AIs to test their attack code, without necessarily having to initiate a war to the death with a neighbour.

    Maybe after remaining at level 6 for a month unchallenged these strongholds could launch boosted attacks against every room in the sector, wipe them out, and create a new respawn zone...

    👌👎

  • Overlords

    I love it all except damage reflection. Seems unnecessary

    👍


  • I really like this! Just one question: If you nuke one of them will the containers be destroyed?



  • I am really looking forward to this. What has been presented looks very good. My concerns/questions right now are:

    1. Reward. This is the tricky bit. Killing level 0 cores is easy and no invaders in SK rooms is quite nice. If I consider the cost of killing the stronghold to be small compared to the reward, what's the reward/tick of tackling the largest stronghold, given it takes >400k ticks to get to level 5? It seems like the containers should hold at least 100k resources. Probably more.
    2. Predictability. It seems there are two ways this could go: the challenge could either be to tackle a fixed function opponent as cheaply as possible. This would involve finding weaknesses in the code and exploiting them. Alternatively, should the challenge be to develop a more general stronghold busting code the be able to target a moving target defense code. Convoys today are the former (though I've heard that their AI is to be upgraded).
    3. Actually being able to kill the level 5 ones. The information we have is too limited but with the description as-is I think it would be possible to have a literally unbreakable stronghold.


  • Am I understanding this correctly? If a level 0 stronghold appears in a sector and is immediately wiped out then that sector will be completely safe from invaders until another stronghold spawns?

    May I suggest a different idea then?

    Source Keeper Strongholds - Have the current Source Keeper rooms spawn a level 1 or 2 stronghold, maybe minus the tower. This Source Keeper Stronghold would be unkillable but also not level up. It will always be present in the room. It would spawn defender creeps to protect it's room from people remote mining it. For every energy or mineral mined in the sector, a little bit gets added to the Source Keeper and NPC Stronghold containers. Once the energy container is filled up the Source Keeper Stronghold sends out a sortie of attack creeps to some random room. The creep's energy cost should be deducted from the energy container, and their power dependent on the RCL of the target room. The more powerful the sortie the more energy is deducted. This would be basically like the current iteration of invader creeps.

    NPC Strongholds - Once the mineral container is filled up the Source Keeper Stronghold sends out a builder/defender sortie to establish the currently proposed NPC Strongholds, with some modifications. I would immediately have these NPC Strongholds spawn as level 1 with both an energy and mineral container. Once their energy container was filled they would also send an invader sortie out just like the Source Keeper Strongholds do and also have the strength of the attack based on the level of stronghold and the RCL of the target room. If it is only a reserved room the attackers should include something to attack the controller with to knock down that reservation. Once their mineral container is filled then a new piece of the next level of stronghold is built and an appropriate amount of minerals is removed. The cycle would continue slowly building one piece at a time until the maximum level is reached.

    Player activity and deleveling - Since both stronghold types are dependent on players actively mining resources then these NPC should provide a constant challenge to anyone in the sector. The more rooms or stronger rooms players are occupying then the stronger the NPCs become. Also the fewer rooms or weaker players then the weaker the NPC become. Since the NPC strongholds I outlined level up by player activity they should also delevel based on player activity. I have two ideas for that.

    Idea 1 - Each piece of their stronghold beyond the basic parts at level 1 will cost some sort of upkeep that is deducted from both the mineral container. Once the container reaches 0, or maybe 25% so some minerals are always available should someone attack, then the last added piece is removed and its mineral cost is added back into the container which will slowly tick down to an equilibrium with player activity. This would be just like the build up process but in reverse.

    Idea 2 - The stronghold spawns a creep with worker parts that actively repairs the ramparts using energy and rebuilds/refills the towers using minerals. Both resources would come from the self contained containers. No or little player activity means the stronghold can't repair and rearm. This could also be applied to the leveling up process to provide a more natural looking enemy that follows similar rules to the players.

    I feel idea 1 would be much easier to balance and provide a constant source of engagement with the players. Idea 2 would look more natural in the game world but it would probably be a lot easier to "game the system".

    Closing - I feel like the currently proposed system is just too basic and will become farmable content for some established players. Let a stronghold grow to a certain level, send in the appropriate amount of creeps, and collect your reward. A system that grows with the players and fights back at an appropriate level will in the end be much more engaging and rewarding. If you feel that this pushes the game too far into the PVE arena and away from PVP then you can just tune the NPC down to make them somewhere between a nuisance and an actual threat. I think I have provided plenty of tuning knobs to accomplish that goal.

    Hopefully someone read all of that 🙂

    KL

    👌


  • I like the idea of strongholds' development tied to player activity in a particular sector, but at the same time I'd like to see them grow in sectors with little of it.

    Say a sector has a few active rooms, then that sector would see the stronghold grow big. If a sector has a bunch of low activity/low level rooms, then the stronghold is not very active (like your newbie zones, etc.). and if a room is populated by big, active, high level rooms, then once again we see big and developed stronghold to challenge these rooms.

    Maybe something could be also affecting based on the amount of inactive rooms around, or something.



  • This is very cool. There are a lot of thoughts being posted here (because we're excited!), but I wanted to add one more.

    it sounds like this will supersede SK room invaders. In my opinion, SK rooms are one of the most genius features in the game right now. There is nothing else I know of that has such a diverse set of solutions. Some people use armed miners, some people use a melee+ranged patrol pair, some people use just a ranged creep, some people use a single boosted attack creep, some people use an attack creep for clearing the SK spawns and then a separate healer to just tank the damage of invaders, ect, ect. I haven't watched all that many people's rooms, but it seems like everyone has a slightly different approach.

    This is pretty much the holy grail of balancing for a strategy game. No cookie-cutter solution, and everyone weighing the tradeoffs differently. Obviously, NPC strongholds will attempt to also have a plethora of viable strategies, but I would lobby for preserving the existing feature while adding NPC strongholds. This has the added benefit of not invalidating significant coding effort by everyone who has implemented SK mining. Which probably doesn't resonate with the developers, but it's always something to consider when adding new features.


  • Dev Team

    @omnomwombat Could you please clarify what exactly would invalidate any coding efforts spent on SK mining?

    I totally agree with what you think about SK mining, just not sure if I understand how Strongholds could affect them.


  • Dev Team

    @kyralee Thank you for your suggestions. We'd prefer to not complicate things this way, at least at the start. Making it simple, then monitoring how it goes and then deciding if we want to improve it (and how) sounds like a good idea.


  • Dev Team

    @omnomwombat All Source Keeper mechanics will remain as they are, they will exist in parallel with NPC Strongholds, sometimes even in the same room simultaneously.



  • Generally I like the idea. I see some issues though:

    1. "Stronghold's structures do not decay over time" and "Those containers on pictures are not regular containers with 2000 capacity, they can have any needed capacity, up to millions units." After killing the stronghold, a few undecaying, large-capacity containers will be left? Is it intended so that winner can claim that room and use those for his purposes?
    2. Transferring energy is barely profitable, unless it is meant to be used on-site, again by claiming the room after destroying the outpost. Are there any plans to change that so that energy is a real reward from tier 1 outpost without claiming the room and using it to build the base?
    3. "a middle-level Stronghold is able to spawn additional cores in another room; the additional core can't grow into another Stronghold [...]" What will happen after destroying the outpost that spawned additional cores? Will they still not be able to create an outpost or will one of them be promoted to outpost in place of the destroyed one? (Possibly letting a player quickly populate the sector with bases equipped with extra-large storage in place, see previous points?)
    4. Will a player be able to claim the controller in a room with the Invader Core (eg as an immediate reaction to core appearance, assuming reserver creep is already present and next to the controller)? What will happen then if it will be possible? Will those cores attackController until such room is declaimed? Will it be possible to use safemode in that case? What would be effects of that?
    5. Will a player be allowed to claim controller in the outpost room itself, eg to start safe mode, possibly making the outpost destruction easier?

  • Dev Team

    Thank you for your questions!

    1. Those containers will not stay forever, they will disappear after some time similar to remainings of npc caravans (obviously, loot time will be longer than loot time of caravans). Plus, we are able to make those containers non-transferable by setting .storeCapacity to 0: players will still be able to .withdraw() but not store something there.
    2. We have no plans on changing that. Transferring energy is very profitable at short range (no remote mining could exist otherwise).
    3. In our current vision, these additional cores supposed to just disappear, but this is something yet to think about.
    4. No, controllers in a Stronghold room will be reserved by Invaders Core preventing players from capturing/reserving it.
    5. The same as previous


  • How would the reservation of rooms work? Would the stronghold magically reserve controllers in adjacent rooms? Or would claim creeps appear to do that?

    Not sure if I would like the magic way, since it seems to not allow locally defending my remotes. I have to attack the stronghold.

    Basically I would like Screeps complexity scale out using a small set of base mechanics (caravans do that well).



  • @o4kapuk ad 4: It's about additional cores in rooms reserved by player, not the stronghold room.

    In the original description: "it will attack controllers to remove player's reservation, then start to reserve the controller itself." So the reservation wasn't supposed to disappear immediately? What you just wrote contradicts that.

    According to original description, if core happens to appear when I have a reserver creep present in the room, I think I might have a few ticks to claim the controller after core's appearance. Not sure about attackController's timing and detailed effects. Will it allow me to claim in the first tick after core's appearance?


  • Dev Team

    @zyzyzyryxy This is correct, as attacking controller will take a time so theoretically you'll be able to claim the room, assuming you have GCL available. Thank you for pointing out to this case, details like this are yet to be worked on.



  • @artch @o4kapuk, thanks for the response and clarification. I saw something in slack that made me think spawn-at-edge-of-map invaders would no longer occur for source keeper rooms. Glad to hear that is incorrect!



  • @o4kapuk said in Draft: NPC Strongholds:

    No, controllers in a Stronghold room will be reserved by Invaders Core preventing players from capturing/reserving it.

    (it is actually my point 5)

    Will a player be able to attackController to remove reservation, then claimController? What will happen in that case?