Dev Diary #16 - Automation


#1

#2

I love most of what is planned here, must admit im a little concerned about the “autobuild” feature and as such will probably never use this as some of us utilise planets for specific purposes and don’t wish to have them round themselves out evenly. ie: population hubs for loading of troops.
Just a basic set of guard/patrol/defend/aggressive modes and a simple waypoint system which can be used for both warships and freighters would be all that we need I believe…
Keep up the great work guys!
Vassago.


#3

I like most of what I see, but I agree with Vassago, the auto build feature is something that I probably wouldn’t use either as I build specific structures based on the planet’s purpose.

If we could set this to on/off by planet and by type of structure it would be best and then I might use it. Farms I’d like to always be automatic, if I run out of population room, on any planet I’d like to see a farm built to keep my population growing, however I might not want mines on certain planets to be automatic, because I have a different purpose in mind for that planet. Same for ECs, some planets I would upgrade because I’m saving population for a higher purpose, other’s I’d build new ones. So automated by structure type and planet would be best overall I think.

Love the feature for scouting.


#4

@mel this will certainly help some players. I have a concern that if I am offline and a mine gets built and the happiness becomes negative the program will take over and act. A lot of the time I may want that to happen. I know a lot of people increase their taxes in the short term to get some credits flowing. If they do that and then go offline we will neede the ability to turn this on and off - which I am sure you will include.

Another area of automation that would be great is in the fleet management. You already have ‘Attack & return’. Would be nice to have "Drop cargo and return’ as well. Scouts could be made to automagically scan every planet in a system when they arrive providing there are no other civs there.


#5

What’s the difference between Auto-Exploring a System and making it so that when you enter the system you are just automatically given all the info about the system? I know they aren’t Exactly the same thing, but I worry about automation essentially making key gameplay mechanics obsolete. Maybe in this case the tediousness makes the argument for automation overwhelming, or maybe it will mean players lose the feeling of exploring the galaxy. After all, once your fleet arrives in a system, you can explore it manually pretty quickly, and certainly in a single short gaming session, so I don’t really see this aiding players who can only log in once or twice per week.


#6

When the ship speeds in-system were dramatically faster, I agree, it wasn’t that bad.

Right now? They’re awful. And if you have 4-5 scouts (or more) running around, that’s an extremely tedious task.


#7

How about being able to toggle what building types get built? Then being able to sort their priority and set a maximum amount…


#8

I would like the buildings to not go offline while in queue. That is my main complaint. You can take it offline while upgrading, but while it is not upgrading and in queue, it should not be a thing. This greatly benefit those who has time to spend with planets. @mel @joe, please take this into consideration.


#9

@joe I would like to see some more automated combat. We have guard mode within planet orbit, but I would like to have fleets automatically engage an enemy in space as well.

As mentioned it the blog, we can’t be active 24/7. Having ships “patrol” my territory would help with incursions while I’m not online. With such a feature we should have the ability to mark other players as friendly, so we’re not accidentally taking out friendly fleets. Additionally, it should attack and return, so if I’m able to destroy an enemy my fleet should return to its original position. We should also only engage fleets we can catch (so we’re not getting drawn away by speedy decoys).

Second on the list would be waypoints. That could be a simple way to solve the planet scanning too. Just having the ability to set multiple waypoints would make scanning easier. If we then could set actions at a waypoint (pick up fuel, deposit minerals, reload ordnance, lay mines, etc) that would be awesome. This could also be something that was potentially controlled via research. Tier 1 allows a single queued action, up through tier 5 allowing five queued actions.

I like the building automation idea. Personally, I’d even be fine with it if it didn’t get any more advanced options. Automation should help with the mundane, but it shouldn’t be able to match the efficiency of an active player (otherwise why bother playing, just idle build everything you need). If we are going to have more granular controls I think they should require research/structures. This adds a gate mechanism so a player needs to be big enough to unlock the tech.


#10

Yeah i don’t want the ‘govna’ making too drastic decisions for me on building priorities. Something along the lines of the basic listed top to bottom check list with tweak of being able to toggle certain buildings NOT being buildable by automation which could simply be a on/off toggle OR simply being able to specify the most of a given building it can build. If it’s reached those limits of the buildings that it is allowed to build via the top down list then it just sits idle and has to wait for the supreme chancellor to return and authorize a new quota. I’d think certain buildings are just plain off limits to the ‘govna’ such as research stations, orbital defense, and shipyards. Those have to be specifically tasked for construction from high command…

I’d love for scout/freighter automation, or at the very least what archo suggested with a limited, exact limit to be determined, number of orders/waypoints that can be assigned to a fleet, but could be set on repeat. This would allow you to have a freighter make it’s rounds to a certain number of planets and at each stop you can designate what it is to drop off and pickup, to the best of it’s ability/resource limitations, and then move on to the next waypoint. This same scheme could be used for other fleets for patrolling fleets to allow them to move along a, as yet undetermined, limited number of waypoints and if non allied vessels are detected it will intercept them, but only to a point. Obviously we can’t have our fleets being baited into prancing across the cosmos, thus certain limits on how far from the patrol path that a fleet can give pursuit. Perhaps to deter silly tactics like sending in a useless bait ship to distract a patrolling fleet the higher priority target would take precedent provided it has a chance of even scratching the paint of the more dangerous target. Also might be worth considering having a patrolling fleet hold station if a non allied fleet is within sensor range but outside of it’s pursuit distance to prevent a live human from simple being able to idly watch the “stupid” patrolling fleet to sail on by simply because it’s out of pursuit distance despite it being perfectly aware there is a fleet just parked outside of it’s authorized engagement range.

Nothing will be perfect, and we can’t expect that, but we can at least try to give the offline player some modicrum of defense against the “LIVE” piranha’s that arn’t restricted by simple if/and/or statement automation. If made too easy to circumvent then there’d be little point in having any automation at all. It would be all to easy for a online player ordered strike force to decimate some poor offline schlups fleets/systems that they’ve spent weeks/months investing in only to lose because they can’t be online 24/7 365. Even if we could have something implimented that could text/email the player when certain criteria are met, enemy fleet detected etc etc that would still only help if they were actually capable of get online and respond. I can see it now, “sorry boss, but i have to head home to deal with an agressive kittie empire trying to gobble up my mining system.”

how much is too much, and how little is just plain pointless?? inquiring tin can’s want to know!!!


#11

Debatable, very controversial. Personally, I don’t think it’s very relevant. It would be better to see the opportunity to send a message to the player and understand what are the conditions of victory on the server.


#12
  1. If you have enough resources, power, population, happiness and there is a resource deposit on the planet, it will construct a new mine for this resource (only one mine per resource at a time, and no more will be built once the mining rate cap per hour is reached)

What is this mining rate cap you speak of?


#13

Second! I just started cranking up my mine count now that I’ve mostly settled into the planets I want to keep. Would had to find out that after queuing up 60 deep mines I top out at 20 or something silly like that.


#14

@Yarri which is debatalbe? which is very controversial and isn’t very relevant to you? Just curious if your talking about the dev’s blog or something I said, or perhaps someone else? It may be debatable, but kind of hard to have that debate when we don’t know what your debating exactly :slight_smile:

I would definitely be concerned if the only thing done was a message being sent to the player as the whole point is to be able to be at least competitive even if you can’t spend 24/7 online. Some folks are able to spend WAAAAAAY more time online then the average bear, er…person, and in the time a normal person, perhaps even a somewhat reasonably fast acting person takes to be able to get back home and online rather serious damage can be done by an online opponent. Granted we don’t know how things will be tweaked just yet but as things currently stand someone could sneak in and take a planet or simply burn it’s people/structures to the ground before you could even respond. So in general i don’t think it’s, in my opinion, irrelevant considering the scope and time involvement in this kind of game.

just sayin as everyone’s opinion matters, but i’m having no luck discerning if your refering to the topic in general or a specific item brought up in response to the blog.


#15

My take is that “too much” automation is when it can run nearly as effectively (if not moreso) than an active player. If upgrades are firing off as soon as a structure is complete that gives idle players too much of an advantage.

For me, the sweet spot of automation is when a player is at a point where they are too busy managing other segments of their empire that they can’t be bothered with the minute, and they are willing to accept some inefficiencies with lower levels of gameplay so they can focus their effort at higher levels.

@IkkeTM had an awesome post about APM here:

In that post Ikke uses Starcraft as an example, and I love the analogy. Early on in an RTS you have the time to make sure your basic units are operating at maximum efficiency. However, as the number of units increase, it becomes more difficult to micromanage everything. The AI runs the simple tasks so the player can focus on more important issues.

So, for a player managing 30+ planets it’s a pain to hop through all of them to build farms. I’d be OK if my pop growth dropped to 0 for 6 hours, and then the planet automatically built a farm. I would always have the option to proactively build farms, but since I have other time commitments it’s good to know that my empire would tick along without me for a day or two.

The flip side of your question about how little is too little… The big thing for me would be fleet automation here. As you mentioned, if my combat fleet can be baited, and goes off touring the galaxy because it saw a scout, that would suck. If it didn’t engage an enemy until the enemy was already past my fleet, that would also suck. I think fleet automation needs to be a lot closer to player-level AI, because that’s our main defense.

Which brings us to a bigger issue, the lack of defense. Right now I can send a scout at warp 9 past an opponents perimeter scouts. The fleet functions as a mine detector (if it blows up then there are mines). If we have auto-engage then the enemy fleet chases me, and like falls farther and farther behind. Eventually I get to a planet. I send a scout in to trigger the orbit defense, then I send in a couple other fleets to clear any planet-based fleets. After that I launch an invasion. I could enter an opponents space, decoy their perimeter, defuse orbit defense, destroy fleets, invade, and have the invasion won before my target even got a full night’s sleep. Meanwhile, my cost is a scout or two, and a fleet to counter their planet defense.

We need better planet defenses, or we need to slow down the entire invasion process.

Anywho, back on topic… I think automation should have a hockey-stick curve for adoption. For newer players automation should rarely be used. They have relatively few things to do, and therefore can spend more time making sure things are running at peak efficiency. After time a player should hit a tipping point where the increased efficiency of manually controlling everything becomes less important than other aspects of the game.

Another issue is that we currently lack enough end-game content to fully realize the benefits of automation. Since I’m at colony cap, most planets are at fleet cap, and I’m constantly out of credits I still have time for the simple tasks. I wish that I had more stuff that I could be doing that would make use of a 16/16 fleet of battleships and destroyers, but instead I chase one 11/14 pirate fleet every few days.


#16

@mel @joe Unfortunately a Syntis player will not use Autobuild. Credits are in too short supply and it would soak up necessary credits for when the player logged on… Auto build for organics gives them an unfair advantage on the scale of Pay to win…


#17

I mean, it’s not the update I want to see. The game still lacks the basic, primary functionality. We still can’t send a message to the player in the game. It is also unclear the meaning of the game, victory conditions Alliance on the server. And these are the basic things without which you do not have a holistic concept. Instead, in updates there are some minor and absolutely optional functions such as automation of researches and constructions. That’s why I write that is very controversial it looks. The game lacks dynamics, strongly lacks fighting and incentives to conduct military operations, there is also a holistic concept of the meaning of the game in principle. But instead, we see how the images of the race of people are redrawn and the research process is automated (which can be queued for a week or more)


#18

Wheee what a read!

Nothing like a Dev Blog to stirr the blood eh!

OK my 50c. Great idea, great INITIAL attempt at it but a ways to go yet. Highly amusing its coming out just as the early adopters of A3 are running out of things to actually do…

Right now I will be switching it OFF the moment the update hits, unless the AI’s list of things to do can be toggled so it doesn’t go bananas and build or deconstruct things I have a plan for! No I never switched on the Tech Auto either nor I suspect will I.

For example:

  • I do not want to see any mines AT ALL built on deposits much below 50% will this be screened I wonder? We shall see!
  • I do NOT want ANY T2 upgrades on any world under 10 Kkm in diameter ever. I wont be keeping those forever, period beyond T1 buildings will be a waste…

And so on there are plenty of cases where this kind of Automation algorithm will be a royal pain in the ass unless we can tailor it. And there are some great speculations above on how it might be “gamed” as well so cautiously cautiously methinks.

At the risk of reopening an old wound one small comment more then I am done. Automation limits would have been a wonderful way of implementing the corruption effect! The more corrupt your empire the less “Help” you get from Automation. This self limits itself and if taken to the extreme the Automation could work AGAINST you because of the “corruption” in your over large (inefficient) government…


#19

Most of the goals I have for planets are formulated in some 3-4 lines, which in fact put into the planet description for easy reference. An example would be:

Beron - 500/h (implicitly, I mean on deep mines here)
Zyril - 600/h (again deep mines)
T2 Fleet controller (I dont write this down, it’s implicit I need to max this)
T3 Shipyard
T2 Orbital defence (Again I dont write this down, it’s almost as implicit)
Fill - Science (fill means - if I have pop left after that, put them on science labs)

This is about the level of strategic decision making you have in planet development. I’m basically an AI working towards those goals when I check my planets. The AI should be able to take over this much, leaving me with just the strategic choices. Now one might argue this would leave the game very bare, but that’s a flaw in the strategic depth provided, not an excuse to maintain routine labor.


#20

Thanks for the excellent feedback everyone!

The planet management assistant will be disabled by default and
work on a per-planet basis, so if the ability to theme planet development e.g. defense, population growth etc. is introduced, it would then be possible to use it to achieve a strategic goal for that planet.

I like this idea, could be a first step in that direction, if it could be done as part of the wider task for freight automation/waypoints.