Revised Colonization feedback

Describe your experience with the latest version of FreeOrion to help us improve it.

Moderator: Oberlus

Forum rules
Always mention the exact version of FreeOrion you are testing.

When reporting an issue regarding the AI, if possible provide the relevant AI log file and a save game file that demonstrates the issue.
Post Reply
Message
Author
User avatar
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Revised Colonization feedback

#1 Post by Dilvish »

Playing a bit has confirmed, for me at least, that the distance dependent timing on colony buildings is working out well. It just feels right that an outpost off in the distance would take a while to colonize, and the amount of variance in the time seems about right. I can't claim to be close to adapting the AI well for all this, but I'm liking the feel of the play.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

Magnate
Space Dragon
Posts: 425
Joined: Sat Nov 08, 2014 3:44 pm

Re: Revised Colonization feedback

#2 Post by Magnate »

Hmmm. I've just upgraded from 8017 or so to 8050 and had my first experience of the new colony building via outposts.

In terms of numbers, the costs and timings and distance-dependent variations feel about right. I don't have a problem with that.

But I do have a problem with not being able to blow enemy colony ships out of the sky, and having to conquer every outpost before they turn into colonies. I used to ignore outposts quite happily (pace the occasional strategic gem like orbiting a black hole), but now they're much more of a pain. It just feels wrong to create a colony without a ship to take people there.

Does this new method require any tech to do? I didn't notice it until I tried my first exobot colony.

Is it more expensive and slower than using colony ships? If it isn't, why would we build them? Just spam outposts everywhere, and turn them into colonies when you get the right race and tech to make one.

Just a few thoughts - I have only played one game, and that was without realising this had changed. I'll try to offer some more considered feedback when I've played a full outpost-spamming game.

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Revised Colonization feedback

#3 Post by MatGB »

It's, deliberately, slightly cheaper than building a colony ship, as the main thrust of the design was to reduce the annoyance of colonising by reducing micromanagement, that's definitely acheived.

The reason you might build colony ships, especially in the early game, is that you can't convert an outpost into a colony unless they're supply linked, and outposts have reduced supply compared to colonies. And there might be that lovely looking Large/Huge Good Environment world with a nifty special that the outpost method simply won't work on.

But the costings and mechanics are deliberate not just to encourage people to use this over standard colonising, but as ell to encourage aggressive expansionist players like, well, me, to actually do some colonising after the early game. That objective is very much acheived, at least for me.

Regarding
It just feels wrong to create a colony without a ship to take people there
But you are sending ships there, that's what Supply Lines are, merchant vessels, traders, colonists, etc moving safely within the area you've secured. You can't use this method to colonise outside of Supply, deliberately, and you need to get an Outpost there (normally by ship) anyway.

I am, however, completely confused by this...
I do have a problem with not being able to blow enemy colony ships out of the sky, and having to conquer every outpost before they turn into colonies
One of the things I love about FO is the vast number of different strategies and playstyles it supports. The idea that someone else runs a strategy that destroys free colonies build by others is just strange to me, I like to let AI colony ships through into nice planets then grab them once they're established, once the AI has been adapted to use the ne method that'll be a lot harder and improving my species mix will require more work.

For what it's worth, the AI currently only uses the new mechanic for Exobots, so you don't need to worry too much about it for a bit, but once the AI team have adapted behaviour to use the new mechanics we'll both need to adjust a bit, I won't get as many free people, you'll need to be paranoid about potential outposts you're ignoring. Me, I like to grab the outposts, now you can just turn them into colonies at will.

Feedback is always good. A reminder that there are many different preferred playstyles is also good.

Curious, why try to shoot colonies down? It seems a strategic mistake to me.
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Magnate
Space Dragon
Posts: 425
Joined: Sat Nov 08, 2014 3:44 pm

Re: Revised Colonization feedback

#4 Post by Magnate »

So I've now played a full game (finally won as Egassem!) so here is some more feedback:
MatGB wrote:It's, deliberately, slightly cheaper than building a colony ship, as the main thrust of the design was to reduce the annoyance of colonising by reducing micromanagement, that's definitely acheived.

The reason you might build colony ships, especially in the early game, is that you can't convert an outpost into a colony unless they're supply linked, and outposts have reduced supply compared to colonies. And there might be that lovely looking Large/Huge Good Environment world with a nifty special that the outpost method simply won't work on.

But the costings and mechanics are deliberate not just to encourage people to use this over standard colonising, but as ell to encourage aggressive expansionist players like, well, me, to actually do some colonising after the early game. That objective is very much acheived, at least for me.
Yes, it's certainly easier in terms of micromanagement to use this method. It's also hugely flexible to be able to decide which species to colonise with after the outpost is in place rather than before. But I think the ease and flexibility should come at a cost - I think it should take longer and cost more to do it this way, not take less time and cost less. I think the min turns for a new colony should be six rather than five, so that it's a turn longer than building a colony ship (when you add three for the outpost ship). The PP cost needs to go up too, so that when you add the cost of an outpost ship the total is more than the cost of a colony ship - at the moment it's less. Currently this method is a no-brainer when we surely want a strategic choice.
Regarding
It just feels wrong to create a colony without a ship to take people there
But you are sending ships there, that's what Supply Lines are, merchant vessels, traders, colonists, etc moving safely within the area you've secured. You can't use this method to colonise outside of Supply, deliberately, and you need to get an Outpost there (normally by ship) anyway.
This is fine logic for basic colonisation, but falls apart when we consider capturing enemy outposts. Previously you would have had to build a colony ship of the desired species and send it to the outpost - now you just build whatever colony you like in vastly less time. Perhaps there should be an additional time and cost for using this method on captured outposts?
Curious, why try to shoot colonies down? It seems a strategic mistake to me.
I'm always behind the AI in defence tech, and usually also have fewer warships - so taking his colonies is non-trivial and I'd rather colonise them myself than allow him the toehold of extra detection and supply. Too many times I've lost invasion fleets to defences which sprang up way faster than I expected (it occurred to me that the AI maybe prioritises Defence Network Regen!).

I'm not saying I don't like this method - and for big galaxies it works pretty well when there are vast distances involved because interruptions to supply pause the building (whereas a colony ship can continue on its way). I also like the challenge of no longer being able to ignore enemy outposts. But I think being able to whack up a new colony in five or six turns on a newly-captured frontier outpost is too easy when a colony ship would take over twice as long to build and send.

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Revised Colonization feedback

#5 Post by Vezzra »

Magnate wrote:Yes, it's certainly easier in terms of micromanagement to use this method. It's also hugely flexible to be able to decide which species to colonise with after the outpost is in place rather than before. But I think the ease and flexibility should come at a cost - I think it should take longer and cost more to do it this way, not take less time and cost less.
This is something we specifically try to avoid, and goes against our fundamental design philosophy. Game mechanics/elements that help you reduce micromanagement must not ever come at some kind of in-game costs (be it research, production or whatever resources). The reason for that is, if we allow for/do something like you propose here, then the player is faced with the decision to either put up with more micromanagement or play a less optimal game. Which results in players who do more micromanagement gaining an advantage, and by that more or less force the other players to take the more micromanagement heavy route too.

That totally contradicts our design goal to avoid micromanagement as much as we possibly can. If we discover that a certain game mechanic/element can used in a way that micromanaging them gives you an advantage, we consider that a design flaw and will redesign the game mechanic/element in question accordingly.

That's the reason why we don't have techs, buildings, etc. that serve the purpose of reducing micromanagement. And because of that, it's specifically intended that the upgrade outpost approach is clearly and always the superior one to the colony ship, so the player does not have to choose between higher costs or more micromanagement.

See also this post.
Currently this method is a no-brainer when we surely want a strategic choice.
I think this is the fundamental misunderstanding here: These two colonization methods are most certainly not intended to give you a strategic choice between two different approaches, each with it's own advantages and disadvantages.

They are supposed to be two different methods for two different cases, and each one is designed to be the obvious best choice for one of these two cases:

* The upgrade outpost to colony method is the "normal", "standard" way how colonization works in FO. It applies to all planets within your supply range, and for these planets it is designed to be the clearly superior method.

* The colony ship method is the exception, the "special case", intended for planets beyond your supply range. Initially, I even intended to make that dependent on a tech, so to gain the ability to colonize planets beyond your supply range would have required the additional investment of research. But the others (rightly) objected that the case where you don't have any planets suitable for colonization within supply range in the early game happens so frequently that this would often result in a tedious early game, and it would also create a huge imbalance between players that actually do get suitable planets within supply range at game start and those that aren't so lucky.
This is fine logic for basic colonisation, but falls apart when we consider capturing enemy outposts. Previously you would have had to build a colony ship of the desired species and send it to the outpost - now you just build whatever colony you like in vastly less time.
Why does the logic fall apart in case of capturing enemy outposts? So it's much more rewarding to conquer enemy outposts - or much more important to prevent your outposts froom falling into enemy hands. But as that possibility is open for everyone, no one gains an advantage by that.
Perhaps there should be an additional time and cost for using this method on captured outposts?
Keep in mind that the build time is already dependent on the distance of the nearest colony with the species you want on the colony you build on the conquered outpost. Conquered outposts tend to be farther away from your main worlds, so they should take longer to build anyway.
But I think being able to whack up a new colony in five or six turns on a newly-captured frontier outpost is too easy when a colony ship would take over twice as long to build and send.
If it turns out to be the general preference, this could easily addressed by reducing the build costs and time for the outpost ship and increasing that of the colony building. That way the total build cost and time can remain the same, while putting more emphasis on the colony building. That was my original intend anyway, but the others objected to making the outpost ship too cheap.

Magnate
Space Dragon
Posts: 425
Joined: Sat Nov 08, 2014 3:44 pm

Re: Revised Colonization feedback

#6 Post by Magnate »

Vezzra wrote:This is something we specifically try to avoid, and goes against our fundamental design philosophy. Game mechanics/elements that help you reduce micromanagement must not ever come at some kind of in-game costs (be it research, production or whatever resources). The reason for that is, if we allow for/do something like you propose here, then the player is faced with the decision to either put up with more micromanagement or play a less optimal game. Which results in players who do more micromanagement gaining an advantage, and by that more or less force the other players to take the more micromanagement heavy route too.

That totally contradicts our design goal to avoid micromanagement as much as we possibly can. If we discover that a certain game mechanic/element can used in a way that micromanaging them gives you an advantage, we consider that a design flaw and will redesign the game mechanic/element in question accordingly.

That's the reason why we don't have techs, buildings, etc. that serve the purpose of reducing micromanagement. And because of that, it's specifically intended that the upgrade outpost approach is clearly and always the superior one to the colony ship, so the player does not have to choose between higher costs or more micromanagement.
Thank you - that's very clear.
I think this is the fundamental misunderstanding here: These two colonization methods are most certainly not intended to give you a strategic choice between two different approaches, each with it's own advantages and disadvantages.

They are supposed to be two different methods for two different cases, and each one is designed to be the obvious best choice for one of these two cases:

* The upgrade outpost to colony method is the "normal", "standard" way how colonization works in FO. It applies to all planets within your supply range, and for these planets it is designed to be the clearly superior method.
In which case my one piece of feedback is that for very long distances this isn't quite working. The five-turn minimum is fine, but the colony build quite quickly scales up to 12+ turns and I have seen as high as 23. I am finding that from about 15 or 16 turns it's quicker to build a fast colony ship and send it across the gap. Don't forget the delay when you conquer your first world of a new race while their happiness goes up - this delays the new method but does not delay building a colony ship (if the place already has a shipyard).
* The colony ship method is the exception, the "special case", intended for planets beyond your supply range. Initially, I even intended to make that dependent on a tech, so to gain the ability to colonize planets beyond your supply range would have required the additional investment of research. But the others (rightly) objected that the case where you don't have any planets suitable for colonization within supply range in the early game happens so frequently that this would often result in a tedious early game, and it would also create a huge imbalance between players that actually do get suitable planets within supply range at game start and those that aren't so lucky.
Again, thank you for setting this out - very clear.

CC

User avatar
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Re: Revised Colonization feedback

#7 Post by Dilvish »

Magnate wrote:for very long distances this isn't quite working. The five-turn minimum is fine, but the colony build quite quickly scales up to 12+ turns and I have seen as high as 23. I am finding that from about 15 or 16 turns it's quicker to build a fast colony ship and send it across the gap.
Hmm, yes, the timing was a first pass meant to correspond to the amount of time a ship would take, but it doesn't take into account the development of faster ships, and it also makes an approximation by using number of jumps instead of total jump distance. We should probably use jump distance, and then we could also peg it to a faster ship; by the time that an empire is big enough for the speed matter much it is probably reasonable to assume they'll have faster ships.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Revised Colonization feedback

#8 Post by MatGB »

Magnate wrote: I am finding that from about 15 or 16 turns it's quicker to build a fast colony ship and send it across the gap. Don't forget the delay when you conquer your first world of a new race while their happiness goes up - this delays the new method but does not delay building a colony ship (if the place already has a shipyard).
For what it's worth, this is the first implementation of Happiness as an actual mechanic that does something, part of Geoff's stated objective with it was to reduce the 'logrolling' effect of being able to build immediately on conquest, and I'd like to see, fairly quickly, a restriction on shipbuilding of various types that key off the same basic 5 happiness requirement. It's just that we've not discussed doing it properly, and both are stopgaps until we've time to think through how we want it to work overall.
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Revised Colonization feedback

#9 Post by Vezzra »

Magnate wrote:In which case my one piece of feedback is that for very long distances this isn't quite working. The five-turn minimum is fine, but the colony build quite quickly scales up to 12+ turns and I have seen as high as 23. I am finding that from about 15 or 16 turns it's quicker to build a fast colony ship and send it across the gap.
Thanks, that's a very important observation. The revised colonization mechanics are quite new, and the numbers a first rough estimate which still need polishing.

I guess Dilvish suggestion:
Dilvish wrote:the timing was a first pass meant to correspond to the amount of time a ship would take, but it doesn't take into account the development of faster ships, and it also makes an approximation by using number of jumps instead of total jump distance. We should probably use jump distance, and then we could also peg it to a faster ship; by the time that an empire is big enough for the speed matter much it is probably reasonable to assume they'll have faster ships.
makes very much sense. However, I can't find a corresponding "DistanceBetween" FOCS function (which would yield the jump distance instead of number of jumps). Did I miss something, or do we need to implement that first on the C++ side?
Magnate wrote:Don't forget the delay when you conquer your first world of a new race while their happiness goes up - this delays the new method but does not delay building a colony ship (if the place already has a shipyard).
Oops, that's actually a bug I think. You're right, only the colony buildings require the happiness >= 5 condition, but not colony ships. That needs to be fixed.

My proposal: Add happiness >= 5 to the location condition of the colony pod ship parts. Anyone any objections?

While we are at it, I wonder if we should extend that to all ships in general - you only can build ships on worlds with happiness exceeding a certain treshold. After all, the guys from that world are going to be the crew of the ship, and if they are too angry, that shouldn't work... comments?

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: Revised Colonization feedback

#10 Post by MatGB »

Vezzra wrote: My proposal: Add happiness >= 5 to the location condition of the colony pod ship parts. Anyone any objections?

While we are at it, I wonder if we should extend that to all ships in general - you only can build ships on worlds with happiness exceeding a certain treshold. After all, the guys from that world are going to be the crew of the ship, and if they are too angry, that shouldn't work... comments?
If you want to do it it'll save me the bother of finding the time, part of the idea of Happiness was to reduce steamrollering conquests, so this would be a good plan.

But...

Can we have a sitrep for colonies that reach the threshold of 5 in the same way we get one for population 3? That'd good for colonisation and shipbuilding purposes.
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

User avatar
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Re: Revised Colonization feedback

#11 Post by Dilvish »

Vezzra wrote:However, I can't find a corresponding "DistanceBetween" FOCS function (which would yield the jump distance instead of number of jumps).
You're right; I had recalled a distance function being added but it was just a direct distance. I've got it partly done now (the parser part anyway) and should be able to finish today sometime. There is a universe method ShortestPath which gives the right value between systems, and I just need to make a more general version of it akin to Universe::JumpDistanceBetweenObjects() (which gives number of jumps). Speaking of which, I wonder if we should rename that to just be Universe::JumpsBetweenObjects() to make it a bit more clear. **edit -- Ah, I am now realizing that the current name is probably to help make it more clear it just returns number of jumps rather than also the actual jumps path.
While we are at it, I wonder if we should extend that to all ships in general - you only can build ships on worlds with happiness exceeding a certain treshold. After all, the guys from that world are going to be the crew of the ship, and if they are too angry, that shouldn't work... comments?
Sounds ok in general. Keeping in mind the discussion to someday add starbase type immobile hulls (akin to the current colony_base hull but with externals slots), I would think that this requirement should only apply to mobile ships; even angry citizens are likely to be willing to defend their planet.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

User avatar
Dilvish
AI Lead and Programmer Emeritus
Posts: 4768
Joined: Sat Sep 22, 2012 6:25 pm

Re: Revised Colonization feedback

#12 Post by Dilvish »

ok, I added a ShortestPath ValueRef and FOCS handling for it. One proposal for how to use it here is in a Pull Request I opened on my fork to facilitate any discussion of the actual content changes: https://github.com/Dilvish-fo/freeorion/pull/1

Note, you'll need newly compiled FO excutables to use this content.
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Revised Colonization feedback

#13 Post by Vezzra »

I've split off the discussion about requiring happiness >= 5 for building colony ships/ships in general to a new thread. Please continue the discussion there.

User avatar
Vezzra
Release Manager, Design
Posts: 6095
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: Revised Colonization feedback

#14 Post by Vezzra »

I've moved Dilvish' last post concerning the implementation of using ShortestPath to calculate build time costs for the colony buildings to the Revised colonization mechanics thread. I think that's the more appropriate place to continue this discussion.

Post Reply