test content
Logo
What is the Arc Client?
Install Arc
Proposal: Pet AI Scriptable
rhiwaow1
Member Posts: 40 Arc User
greetings,
as a fan of carriers, the pet ai is a topic that concerns me very much, especially as i prefer fighters over frigates. there's been a lot of discussion about improvements to the pet behaviour or other tricks for their survivability, which although good ideas are only local improvements without much additional benefit outside the specific fix they are supposed to deal with.
in the first stage, my idea is to enable the players to run scripts for the pet ai themselves as part of their game client: basically you get read access to the map and the objects there, a list of powers and their cooldown of the pet, callbacks on changes like object destruction (-> imminent warp core breach in that area) or player full impulse status change, and some customizable buttons to switch between scripts and script behaviours
for these buttons (and possible hangar-hangar interactions) i would envision a client-internal "pet" chat channel which would not be transmitted to the server; that channel would also be very helpful for script development/debugging
the script would be permitted to issue course/speed corrections and power activation for each pet in the hangar it's assigned to (special pets like the d'kyr support craft would need their own hangar, which is an enhancement that is geneally desirable)
the scripts might be shareable similar to the foundry system, with a "tip" option;
in the second stage, once the scripting system is established, it might be possible to allow "training" queues, with the same missions as the current pve queues but single player and reduced rewards; on the plus side, you get to select additional pre-defined ships as well as their scripts (which could be categorized as "generic" and "map specific" to allow for unique tasks as dedicating one ship to defend the kang on cure found).
the ships and their loadouts would be predefined by cryptic, allowing them to showcase new additions which in turn could raise interest in those ships.
cooldown on the training queues would be 0, as they're supposed to allow ppl to train up for unfamiliar missions/difficulty levels or try new loadouts/playstyles
in the third stage, once the training queues work, i would envision a "mixed mode" option for the regular queues: if enabled, you (or your team) will only wait a maximum amount of time till you enter a queue; if insufficient players are available, a random ship and a random script (from the cryptic approved list of scripts) will be chosen to fill each remaining slot, with the script identifying itself in the team chat on mission begin. players with the "mixed mode" disabled will of course not be drawn into those missions...
with this last stage i hope to revive some of the space queues which either have too low a payout or are found too difficult or are otherwise unaccepted by the majority of the playerbase, so that those who enjoy the mission can still play them.
additionally, once the third stage becomes an option, cryptic should have a decent base of ai scripts that are available for future episodes and upgrading existing ones
as a fan of carriers, the pet ai is a topic that concerns me very much, especially as i prefer fighters over frigates. there's been a lot of discussion about improvements to the pet behaviour or other tricks for their survivability, which although good ideas are only local improvements without much additional benefit outside the specific fix they are supposed to deal with.
in the first stage, my idea is to enable the players to run scripts for the pet ai themselves as part of their game client: basically you get read access to the map and the objects there, a list of powers and their cooldown of the pet, callbacks on changes like object destruction (-> imminent warp core breach in that area) or player full impulse status change, and some customizable buttons to switch between scripts and script behaviours
for these buttons (and possible hangar-hangar interactions) i would envision a client-internal "pet" chat channel which would not be transmitted to the server; that channel would also be very helpful for script development/debugging
the script would be permitted to issue course/speed corrections and power activation for each pet in the hangar it's assigned to (special pets like the d'kyr support craft would need their own hangar, which is an enhancement that is geneally desirable)
the scripts might be shareable similar to the foundry system, with a "tip" option;
in the second stage, once the scripting system is established, it might be possible to allow "training" queues, with the same missions as the current pve queues but single player and reduced rewards; on the plus side, you get to select additional pre-defined ships as well as their scripts (which could be categorized as "generic" and "map specific" to allow for unique tasks as dedicating one ship to defend the kang on cure found).
the ships and their loadouts would be predefined by cryptic, allowing them to showcase new additions which in turn could raise interest in those ships.
cooldown on the training queues would be 0, as they're supposed to allow ppl to train up for unfamiliar missions/difficulty levels or try new loadouts/playstyles
in the third stage, once the training queues work, i would envision a "mixed mode" option for the regular queues: if enabled, you (or your team) will only wait a maximum amount of time till you enter a queue; if insufficient players are available, a random ship and a random script (from the cryptic approved list of scripts) will be chosen to fill each remaining slot, with the script identifying itself in the team chat on mission begin. players with the "mixed mode" disabled will of course not be drawn into those missions...
with this last stage i hope to revive some of the space queues which either have too low a payout or are found too difficult or are otherwise unaccepted by the majority of the playerbase, so that those who enjoy the mission can still play them.
additionally, once the third stage becomes an option, cryptic should have a decent base of ai scripts that are available for future episodes and upgrading existing ones
Post edited by rhiwaow1 on
0
Comments
^ this
The devs here arent very receptive to suggestions. You think carrier pets just got released yesterday or something? you know how long people have been asking for improvements to pet AI or pet survivability? They cant even give us a way for carrier pets to be smart about avoiding warp core breaches...do you reaaaaaaally think they are going to go ahead with some 24 phase approach to giving us pet AI scripting? The only way i see your suggestion happening is if they already planned on doing exactly what you are saying anyways.
My thoughts as well.
STO Screenshot Archive
and i know that this isn't something that could happen soon, and if it happens then only when it fits into cryptic/pwe's plans.
however, pointing out synergies for other areas and possible features they might get out of it could spawn some ideas in the right minds that could lead to something like this becoming part of their plans. maybe they already have another idea which would work great with this, where adding such functionality might come in handy?
also: do you ppl really think that "they'll mess things up anyway" is a good way to get positive attention towards improving pets, regardless of the specific way of improvement?
Did you bump your head on the way to your computer before you posted all of this?
You can make Fighters super intelligent, it wont change the fact they're extremely fragile and I don't see the devs stepping up to change that any time soon since they recently decided to nerf all Carrier pets including all Fighters.
aoe can be partially avoided if the pets for example don't cluster together, which is something one could do when one has the option to program their ai themselves - as well as stuff like: do point defense, but if nothing falls into that category go after my target and not some other small fry nearby which you can't hurt alone with the massive hp on advanced. try to keep your front facing towards the target in case you have narrow-arc weapons, and try to get on the "weak" shield facing - on which you should focus fire. if you have healing abilities, check nearby allies as well as the mothership to heal them with. if you have torpedos and the target shields are about to fall, prepare for a torpedo run with the whole fighter wing as soon as they're down. maybe pop high yield torpedos when you start that run, too!
if you're a frigate with massive dual cannon armament, barrel your way to the enemy, if you're not under attack perhaps even stop to keep the enemy in your kill zone.
or, for mirror invasion and azure nebula, it'd be nice if the pets actively tried to draw aggro so that the mothership can do those interactions - point defense is a start, but occasionally they ignore some enemy which you then need to find yourself and issue an "attack this target" order for a little while.
this kind of ai is something that requires a lot of effort to develop, but enthusiasts will gladly invest that time - much more than it could be reasonably expected from any company
script-support for the foundry is a nice idea, once we have any kind of script support the cost of adding it to other areas of the game drops dramatically
Do you really think that suggesting a complex feature is a good idea when the devs have a record of releasing other buggy features, some of which have bugs that have been around for a long time?
Maybe if every feature they released was awesome and bug free, or at least cleared of all bugs after a relatively short time i would say yah sure go ahead and tell them what you want and how you want it done. But since thats not the way things are here then i think its best to just tell them what we want...and let them figure out on their own how they should do it. Or if you do tell them how you want it done then keep it simple. One thing you could do is to come up with a suggestion that uses current mechanics so that there is a low or no chance of failure/breaking something else. Which is why a lot of people have been suggesting something like simply adding a buff, aura or ability to all hangar pets that would improve their survivability.
proposing small changes to fix problems is something that's been tried for years with little to no effect. so let's try a different strategy.
hypothesis: the previous suggested fixes by the community are rejected internally by the developers/management
a) due to the change having no additional benefit outside the limited area they are supposed to fix and cost too much for their benefit (ai improvement)
or
b) due to the breaking of the design (static damage type immunity)
and regarding the "keep it simple": one of the worst things you can do is to toss in some "simple idea" with no information as to how you want this to look like; there are loads of websites hosting stories of designers about their customers from hell, with that being the most common theme.
the developers are grown people, they can disregard any part of my description as they see fit, but what they can't do is look into my mind to see what i meant. i actually have to come out and describe in some level of detail how i imagine my idea could work, and what benefits i imagine the idea can bring to the game
If i understand you correctly you want the ability to make your own pet AI scripts? MVAM and saucer seperation...current hangar pet behavior. Those are all things that run off of scripts are they not? When i hear reports about MVAMs flying 50km away from the player or the saucer doing weird stuff that is why your suggestion thread makes me uneasy. When i tell my Xindi frigates to recall and i go to 1/4 throttle but the pets just trail behind me and do not dock your thread makes me uneasy. I'm guessing that you know the whole story behind the space tray saving issue and the speed at which the problem is getting fixed? With the space tray issue still being the way it is, do you think its a good idea to ask for a ground tray loadout feature? it would be nice to see but will it end up having the same issues as the space tray loadout?
That is what i meant by "Maybe if every feature they released was awesome and bug free, or at least cleared of all bugs after a relatively short time i would say yah sure go ahead and tell them what you want and how you want it done" If the limited scripted behavior of MVAM, Saucer Sep and Hangar pets worked exactly as they do then i would say sure give rhiwaow what he wants. But those things dont work like they should. Should i also expect player created scripts to not work like they are supposed to? should i also expect it to take forever to fix if it has any kind of buggy behavior?