Dev Diary #16 - Automation

#21

This beautifully sums up my opinion on automation and the current game state. In A2, when you were managing 100+ planets automation like this was significantly more important because it was a painstaking task to micromanage. In A3 I still want the same level of automation, but with colony caps it seems less important. Adding automation at this point effectively removes most of my reason for logging in at all.

My struggle is that if we are going to have automation, the balance needs to be right. We, as players, need to be able to accept the trade off of micro efficiency for macro efficiency. I need to be able to spend less time on the details, and more time worrying about the bigger picture.

A simple example of what I mean by this. A new player should spend 75% of their time in planet/system view. A mid-level player should be spending about 40% of their time there, with 60% at galaxy view. An advanced player should spend nearly 90% of their time on galaxy view.

So, planetary automation should be built around this idea. We have a set number of planetary goals.

  • Mine (type)
  • Labs
  • Shipyard
  • Defense
  • Scanner
  • Population
  • Terraforming

The construction of power, farms, and ECs should be automatic in all cases, but we should have options for building levels. We should be able to toggle T2 and T3, and if it prefers upgrading existing, or adding new.

Now, the real thing, for me, is that the above automation should all be unlocked via research. Make it so people need to actually invest the time into being able to get that automation. Furthermore, I think there should be a structure on each planet for the local steward/governor. At T1 only T1 buildings can be selected, and you can only have one goal. T2 allows T2 construction, and you can add a secondary objective. T3 allows T3 buildings, and a 3rd objective. Also, at T1 it will only build/upgrade one structure every 12 hours. T2 will build/upgrade every 8 hours, and T3 every 4 hours.

This planetary office should also at least have a T1 version added into a new T4 colonization module.

To me, a system like this means that players need to be willing to invest in automation. That way it can somewhat justify an edge that it might provide.

This is also a concern I had. I am constantly fighting the credit battle, and if I were to find that I can’t build ships because my planets are busy building army base #127… I would suggest that in a 24-hour period a planet can construct nothing more than what that planet can afford based on the credit production for that period. Otherwise we will end up with planets racing to spend every credit that comes in.

I think getting shipment automation, or maybe simply automated cargo transport between systems should happen before we push for construction automation. If I have a planet with no beron, but my planet next door has plenty, it should automatically move resources between planets to carry out construction.

This also gets back to my earlier statement about managing from the galaxy view. When I am managing a system from galaxy view I should be able to see the aggregate resources of the system, and (considering every planet has a cargo transporter) use minerals as a giant pool. I should be able to set a ship build order for a system from galaxy view, without having to go down to planet view to do it.

Ok, I think I’ve said to much, I haven’t said enough.

1 Like

#22

@joe looking more closely at the algorytm, at step 1 & 2, I think you can run into a problem where tax levels cannt be reduced, but another EC can be build. Maybe another farm or power plant would need to be build first too, which is ok.

If I may propose a different algorytm, here goes:

Inputs:

  • Every build type aside from power T1 and entertainment T1 can be toggled.
  • Production goals: Tax rate as percentage, mining production per hour per resource, deep mining production per hour per resource, army size, science production, power surplus (to prepare for large buildings).
  • Unique buildings like shipyards and fleet controllers need to be build manually.
  • If constuction is meant to happen, but insufficient resources are available, abort and wait an hour.
  • A neat added feature would be to if the ability was given to shuffle blocks of buildings around, i.e. put science to the top of the priority queue.
  1. If happiness < 100%, adjust tax to +1 happiness, otherwise adjust tax levels to smallest positive change or 0%
  2. If abs (planet temperature – optimal temperature) > temperature tolerance, build climate station
  3. If power left > 120 && power left < 150, build T3 power planet (if toggled)
  4. If power left > 100 && power left < 130, build T2 power plant (if toggled)
  5. If power left < 100, build power plant
  6. If “farm free population” + total farm cap > planet max + 55k, dismantle farm lowest tier farm.
  7. If population > population cap - 5k && “farm free population” + total farm cap =< planet max + 25k, build farm (If toggled)
  8. If “farm free population” + total farm cap >= planet max + 25k, upgrade to farm T2 (if toggled)
  9. If “farm free population” + total farm cap >= planet max + 30k, upgrade to farm T3 (if toggled)
  10. If tax rate = 0% && happiness gain < 0, upgrade to T3 EC (if toggled)
  11. If tax rate = 0% && happiness gain < 0, upgrade to T2 EC (if toggled)
  12. If tax rate = 0% && happiness gain < 0, build EC
  13. If tax rate = 0% && happiness gain < 0 && happiness < 10, dismantle mne.
  14. If total temperature adjustment > abs (planet temperature – optimal temperature) +10, dismantle climate station
  15. Upgrade to T3 climate station (if toggled)
  16. Upgrade to T2 climate station (if toggled)
  17. If temperature =/= optimal temperature, build climate station(if toggled)
  18. Repeat the above 4 steps for hydro stations and land mass
  19. If power left > 120 && power left < power surplus, upgrade to T3 power plant (if toggled)
  20. If power left > 100 && power left < power surplus, upgrade to T2 power plant (if toggled)
  21. If power left < power surplus, build power plant
  22. If army size < army size goal, upgrade to T2 army base (If toggled)
  23. If army size < army size goal, build T1 army base
  24. If tax rate < tax rate goal, upgrade to T3 EC (if toggled)
  25. If tax rate < tax rate goal, upgrade to T2 EC (if toggled)
  26. If tax rate < tax rate goal, build EC
  27. If beron production < production goal, upgrade to T2 beron mine (if toggled)
  28. If beron production < production goal, build T1 beron mine
  29. Repeat the above 2 steps for all other resources
  30. If deep beron production < deep beron production goal, upgrade to deep beron mine.
  31. If deep beron production < deep beron production goal, upgrade to T2 beron mine
  32. If deep beron production < deep beron production goal, build T1 beron mine
  33. Repeat the above 3 steps for all other resources
  34. If science production < science production goal, upgrade to T3 lab (if toggled)
  35. If science production < science production goal, upgrade to T2 lab (if toggled)
  36. If science production < science production goal, build lab.

The above algo is less efficient than doing things manualy, it takes margins for simplicities sake, and those margins are a net loss, you’ll always be producing at least 100 more power than you need for example, also, it will also build too much of the buildings that have a cap (terraforming, farms) and then dismantle them. Manual is still better.

This algo can be ran every time the queue is empty, or at an interval if people feel that’s too efficient despite its inefficiencies.

Edits: some brushing up, didn’t get it right on the first time across the board. Should have probably let it brew a while longer before posting, ah well.

0 Likes

Colony automanager
#23

Good thorough work as always but as I never have run a planet at 100% happiness (except by accident) wouldn’t work for me as written.

But that’s the thing too broad an algorithm and play styles go out of the window unless were really careful. Would we need variations of the above for each race you think? Probably I woulda gueesed.

Also I don’t immediately see how the above allows for Maximum populations on planets? Can you elaborate?

0 Likes

#24

Why not? At 100% happiness you get bonusses, it seems like an obvious goal to achieve if at all possible…

1 Like

#25

Using labs as an example here, but really applies to anything. I wouldn’t presume to upgrade before building new structures. The upgrade costs are quite expensive, and in most cases it’s better to build a second T1 than to upgrade to T2.

Personally, I think we should limit auto-build to power, farms, and ECs. Those are what’s needed to keep a planet running. Construction of things that actually improve a planet (mines, terraforming, labs, etc) should all be manual. Perhaps adding an auto-upgrade option for mines (or the ability to build straight to T3 and upgrade from T1 directly to T3).

Once you start getting outside infrastructure it becomes a much more difficult thing to balance. A player that quits shouldn’t come back after a month to find all their planets fully developed.

If we are going to be adding additional build options through automation then it should be a much more directed process. You set a primary build, and it does infrastructure and whatever buildings you set. I also think additional automation should come at the expense of research and structures. You need to research a planetary manager, and construct the governor’s office. That way a player leaning heavily on automation has their overall research progression delayed while researching automation. They also have to build up the planetary management structure, so that requires some effort.

0 Likes

#26

Using labs as an example here, but really applies to anything. I wouldn’t presume to upgrade before building new structures. The upgrade costs are quite expensive, and in most cases it’s better to build a second T1 than to upgrade to T2.

Well, the idea there is that you toggle them off untill you want to start upgrading them. Then if you want to start upgrading them, but they are lower in priority than the T1, the algo wouldn’t reach them.

Personally, I think we should limit auto-build to power, farms, and ECs.

That’s certainly an option, but what about unemployment? Just build endless EC’s to compensate?

My opinion on the matter is that anything that you just build in bulk should be automated, which includes mines and such. Only unique buildings ought to be exempt. The rationale there is that building X mines isn’t that much less of a chore than building up the rest.

The things is, as I’ve explained over in the APM post, is that I feel in running a governor, I should have rather detailed control over the outcomes, for otherwise there isn’t much use in having it. For example, if I set a planet to mine farsu, I dont need it to mine more than my local shipyards can use. Just “mine farsu” wont do.

Another advantage of this system however, is that outscape can once again scale up to very large empires. Potentially, tying how often the algo runs to empire size / distance to capital for diminishing efficiency.

0 Likes

#27

PR +10% full bonus. Trust me its not that great and to achieve with my style of play not worth it.

You are more than welcome to disagree but I’ll let my Ranks in A2 (bearing in mind I joined several months after it started) be my back up.

What works for some don’t work for others. That’s all I am saying.

0 Likes

#28

Fair enough, tweaking in a setting for “desired happiness” in the production goals isn’t that hard.

1 Like

#29

@IkkeTM your thought processes seem good, the goal being optimisation… As a process that kicks off say every 6 hours… if build queue empty and has been empty for 2 hours then the algo can work out what to build based on the desired goals for the planet… be it happiness/research/mining/production… so building what is logical to maintain the desired goal could follow as you have specified.

If players were to order their desired goals… (happiness then mining)

Then if happiness is a goal and we are at 100% then secondary goal would attempt to upgrade a mine, or build a new mine…

I think upgrades are likely to be the objective as a new building can affect happiness and power… I’m going to suggest that the planet governor is pretty dumb.

I’m wondering if some of the complexity can be avoided by allowing a player to order buildings in advance… so you know your current builds and ordered builds and predict your final state, power, employment etc… create a larger queue… and let buildings be scheduled based on the state… so you can build a lv1 mine and schedule it to upgrade to 2 and then 3 the resources for builds can be taken in advance or could be taken by the planet governor when it fires up… scheduling one building per planet then moving to the next planet… and repeating until all resources/credits/personnel are used up… So every build is as desired but you can keep the queues high (maybe 10 slots) and builds start when resources are ready… Then there is no logic other than the queue itself… which is mostly there…

Additional governor actions could be to paused or even cancel a whole queue if the happiness dropped below 25% and an EC build injected in.

I can see an argument to building as per your post @IkkeTM as quit players’ planets could keep growing and maybe even if a player has turned off the governor… the game can turn it back of after 7 days of inactivity…

I think I would also like to receive an email or push notification that tells me exceptional things about my empire… so planet x has happiness of y, or no build from planet z in 24 hours… things like that…

1 Like

#30

I think what’s listed here is a great start in the right direction, but players should be able to construct their own “templates” of what type of planet they want to build. Limit it to 5 or 10 templates per player and then us come up with what we want ourselves.

1 Like

#31

I forgot to mention the need for rally points, or some other mechanism for newly built ships to automatically join fleets.

I’m sitting at 62/72 fleets right now, with 5 shipyards producing. If I have them all producing destroyers I run into an issue. When a batch completes my fleet count goes to 67/72 and the next batch starts cooking right away. Some hours later I log in and send my new destroyers to join up with their fleets. For sake of example, it’s a long trip from shipyard to fleet. While the fleet is in transit my next batch comes out of the oven. This puts me at 72/72 fleets. Now we see the issue. My fleet limit is reached even though I have 10 destroyer fleets (each with a flagship level of 10, but with only one ship per fleet).

So, first off, when a ship is finished being built I would like an option to have it automatically move to join a fleet (of course this option would need to be toggle-able, and you would get to select the target fleet). Additionally, this new ship (a fleet unto itself) should not actually count against my cap because it’s going to join an existing fleet.

Now, to prevent players from abusing this, I would implement mechanism that kicks in when a player exceeds their fleet capacity for a period of time. If a player exceeds fleet capacity they get an over-limit warning. Then, after 6(?) hours a fleet is destroyed at random. The process would repeat until the fleet limit was no longer exceeded. Basically, rioting for fleets.

If we really wanted to have fun instead of the fleet being destroyed it could go rouge and be converted to a pirate fleet.

0 Likes