FreeOrion

Forums for the FreeOrion project
It is currently Wed Dec 13, 2017 2:57 pm

All times are UTC


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 new topic Reply to topic  [ 38 posts ]  Go to page Previous  1, 2, 3  Next
Author Message
PostPosted: Mon May 02, 2016 5:28 am 
Offline
Space Floater
User avatar

Joined: Wed Apr 08, 2015 4:21 am
Posts: 15
Would it be possible in the future to have "advanced" galaxy generation settings? i.e. where you can type in percents for values, adjust individual A.I. difficulty (to make them all maniacal), etc.

I've noticed that on some seeds most of the closest systems are all filled with a certain "class" of planets (Inferno, Radiated & Barren). I assume this skew is due to the abnormal seed that was inputted, and it would be nice to know how the random universe generator works so that such occurrences could be prevented. (I also want to know for curiosity's sake)

As for strategy, I don't think FO is balanced (as in recommended settings)around having a research centric focus for the early/mid-game as it would always be better to expand. If one really thinks about it, having Adaptive Automation earlier is good, and RP is roughly at around 0.5x the PP in the first 100 turns (So it's more valuable right?). But the problem is that although you may have 2-3 colonies, you have no warships or really ships at all by the time you get it. If a focus on industry had been the goal from the beginning, then you would have had the same amount of colonies as well as warships/troop ships. I think that the only time research works is if you're against the A.I. in a large system (30+ per player) with little to no monsters. Laefna could theoretically start off with research, but they have a penalty. Otherwise, Chato in a young galaxy is quite frankly broken. Gysche could try, but the -weapon dmg and -troops is a steep price for the setback in early game fighting power.

With monsters and specials, this is especially noticeable as your path or a good system is blocked by a maintenance ship or dyson forest. Sometimes with monsters set to "high", you can't even play the game without queuing up a couple outposts asap to stop the kraken nests/drones from taking over the galaxy.

This is why races with an industry penalty (i.e. Cray) aren't very good comparatively as they're too dependent on not being harassed in the early game. In a normal game with med to no monsters, I would focus on research for the early turns. The moment I get Zortium (usually after Nascent, or maybe I go to production after Nascent) is the time to crank up production on robocruisers. Lasers, Symbiotic, and Radar are a lower priority when you compare it with a stack of Mass 4 robocruisers & destroyed monsters/conquered worlds/specials. As the game progresses (mid-game, turns 50-100), you will have built up the largest fleet because the A.I. doesn't know how/where to stack their ships, letting you pick them off without losing any. It wouldn't matter if you only have 25 RP by then; a stack of laser 4 ships is good enough against anything they can whip up by then (still robos w/o shields). Once you have around 10 ships, it basically turns into a waiting game for troop ships. Because you conquered the A.I. so early, you will be able to focus on research with the additional colonies/home world. Robocruisers are good until plasma & diamond before they need to be replaced, as the A.I. now has either asteroid hulls or self gravitating ones. All you need to do is catch up on research, and then continue to outproduce them. With each victory, the A.I. has less of a chance and your footing becomes firmer (They still don't stack their entire fleet, except on homeworlds). This is assuming that you are within easy reach of one of the A.I., else a mid-game focus on research can be considered if there is no combat. Regardless of what you choose, many techs aren't particularly useful/necessary as they a) can take effect after hundreds or thousands of RP/and or PP have been dumped in over 50 turns aka. late game or b) Only useful against A.I. One such example are the stealth techs; they require you to spend double the RP to get it (sensor + cloaking) and a human player can counter it easily if they paid attention to certain things (i.e. they have radar while you already have scanners, so they may expect cloaking soon). (Detection past neutron scanners is usually not required unless playing against experimenters) Of course, this is coming from someone who likes to optimize for the best efficiency, so this might not be your playstyle.

If one wanted to play a long game or fight the "boss", then research is the way to go. But having 30-ish systems means a quicker game, and thus an industry focus is the best. (Can basically win without micrograv/GGG, no asymptotic architecture, nanotech medicine, only robocruiser).

_________________
Image


Top
 Profile  
 
PostPosted: Mon May 02, 2016 4:15 pm 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4288
Location: Sol III
UnifiedStars wrote:
I've noticed that on some seeds most of the closest systems are all filled with a certain "class" of planets (Inferno, Radiated & Barren). I assume this skew is due to the abnormal seed that was inputted, and it would be nice to know how the random universe generator works so that such occurrences could be prevented. (I also want to know for curiosity's sake)
The universe generation scripts are implemented in Python and use Pythons standard RNG (Mersenne-Twister IIRC). The relevant Python modules for the creation of the star systems and planets are:



Top
 Profile  
 
PostPosted: Sat May 07, 2016 7:33 pm 
Offline
Space Floater

Joined: Thu Oct 10, 2013 6:48 pm
Posts: 30
Taking another look after a long break. I'm pretty rusty, so take everything with a grain of salt. That said as someone who likes to rush research early I thought I'd make a few comments.

UnifiedStars wrote:
As for strategy, I don't think FO is balanced (as in recommended settings)around having a research centric focus for the early/mid-game as it would always be better to expand. If one really thinks about it, having Adaptive Automation earlier is good, and RP is roughly at around 0.5x the PP in the first 100 turns (So it's more valuable right?). But the problem is that although you may have 2-3 colonies, you have no warships or really ships at all by the time you get it. If a focus on industry had been the goal from the beginning, then you would have had the same amount of colonies as well as warships/troop ships. I think that the only time research works is if you're against the A.I. in a large system (30+ per player) with little to no monsters. Laefna could theoretically start off with research, but they have a penalty. Otherwise, Chato in a young galaxy is quite frankly broken. Gysche could try, but the -weapon dmg and -troops is a steep price for the setback in early game fighting power.

With monsters and specials, this is especially noticeable as your path or a good system is blocked by a maintenance ship or dyson forest. Sometimes with monsters set to "high", you can't even play the game without queuing up a couple outposts asap to stop the kraken nests/drones from taking over the galaxy.

So just played a bit with Cray on default settings, random seed, young galaxy, high monsters, high specials. I found the monsters more a help then a hindrance for a research start. Fact is they're shielding you from worse enemies & slowing down races that need to invade early to build up steam.

UnifiedStars wrote:
This is why races with an industry penalty (i.e. Cray) aren't very good comparatively as they're too dependent on not being harassed in the early game. In a normal game with med to no monsters, I would focus on research for the early turns. The moment I get Zortium (usually after Nascent, or maybe I go to production after Nascent) is the time to crank up production on robocruisers.

I think you're underestimating the Cray here. If you want to rush robocruisers then take a look how long it takes humans to research Zortium + robot hulls. Now take a look how long it takes the cray to research an industrial centre + Zortium + robot hulls. It's around the same time & with an industrial centre down the cray will out produce humans. In my own game I found a suitable gas giant quite early too & had a gas giant generator down pretty early, fast production isn't necessarily a problem for them.

UnifiedStars wrote:
One such example are the stealth techs; they require you to spend double the RP to get it (sensor + cloaking) and a human player can counter it easily if they paid attention to certain things (i.e. they have radar while you already have scanners, so they may expect cloaking soon). (Detection past neutron scanners is usually not required unless playing against experimenters) Of course, this is coming from someone who likes to optimize for the best efficiency, so this might not be your playstyle.


I wouldn't be so sure on this one. How do you know they have the detection tech above yours if they can see your scouts from further away? Unless you're right on there doorstep they can continually move their scouts out of your detection range. Particularly if they're going for stealth there not exactly going to leave their ships where you can see them, not unless they have to. Not sure how easy/well it works in practice, I've never really tried stealth in FO.


Top
 Profile  
 
PostPosted: Sun May 08, 2016 4:09 am 
Offline
Dyson Forest
User avatar

Joined: Wed Jul 16, 2014 6:14 pm
Posts: 212
Location: Black Forest
wobbly wrote:
I wouldn't be so sure on this one. How do you know they have the detection tech above yours if they can see your scouts from further away? Unless you're right on there doorstep they can continually move their scouts out of your detection range. Particularly if they're going for stealth there not exactly going to leave their ships where you can see them, not unless they have to. Not sure how easy/well it works in practice, I've never really tried stealth in FO.


This doesn't describe the way the AI works in my games. AFAIK the AI enqueues a large number of techs from a pool out of 3 or 4 which distinguishes only a few galactic mapitems, such as presence of Asteroid-belt etc. Stealth & Sensors might be therein, but not at a high priority.
It's fairly easy to go stealth by yourself and be successful with it.
The AI ship-placement routines are also nowhere near sophisticated in order to allow for them to fool a player by ever-positioning outside your vision. And esp. scout ships are usually the first thing you see from them as they're coded to send them into unknown distances.


Top
 Profile  
 
PostPosted: Sun May 08, 2016 4:30 am 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3291
Observation: the AI doesn't use stealth deliberately,a nd doesn't prioritise detection. Stealth as a strategy currently needs a lot of work to be fun/usable, and that's an ongoing project of mine that isn't a high priority, I just make tweaks as and when I have the willpower/time.

I have never played a multiplayer game against a human opponent, I suspect stealth then might be viable without being overpowering, but currently it's too blunt an instrument that can fail easily, so I don't consider it balanced that way at all. I hope it will be, but it's not there yet.

Basically, don't factor stealth into generic strategy guides as a) it's not really playable yet and b) it's complex and the UI isn't up to it properly.

(having said all that, stealthed bioadaptive hulls in large numbers can be great fun to mess around with, but agains the AI they are far too powerful. At the moment, once you've learnt how to use them, you've won)

_________________
Mat Bowles

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


Top
 Profile  
 
PostPosted: Tue Aug 23, 2016 9:09 am 
Offline
Space Krill

Joined: Tue Aug 23, 2016 9:04 am
Posts: 1
Very good guide as I just started playing this game and I can get a good starting empire but fall flat on what to get to capture planets any way you can add this to your guide ?


Top
 Profile  
 
PostPosted: Tue Aug 23, 2016 9:26 am 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3291
Martinpford wrote:
Very good guide as I just started playing this game and I can get a good starting empire but fall flat on what to get to capture planets any way you can add this to your guide ?

Are you stuck on how to use troop ships, or on the best strategy to attack an enemy held system?

The former is a game documentation problem that we might want to expand on within the game itself, the latter is a good thing for a guide to cover.

_________________
Mat Bowles

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


Top
 Profile  
 
PostPosted: Tue Aug 23, 2016 9:34 am 
Offline
Release Manager, Design
User avatar

Joined: Wed Nov 16, 2011 12:56 pm
Posts: 4288
Location: Sol III
Or maybe he just wants to know how to colonize new planets...?


Top
 Profile  
 
PostPosted: Tue Aug 23, 2016 5:39 pm 
Offline
Vacuum Dragon

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 506
One of things that new people don't grasp right away is that you can't invade a planet that has shields up. So worlds controlled by AIs will have at least minimal shields and prevent invasion by unescorted troop ships. You have to have some sort of warship to knock the shields down.

I would encourage anyone having difficultly to put together a specific case that is giving them trouble, not just a vague "having trouble doing X". Fully describe the situation, what you tried, your expectations, and how your results varied from that. That's similar to what we used to tell people to do on the comp.lang.c Usenet group when faced with programming problems.


Top
 Profile  
 
PostPosted: Wed Aug 24, 2016 9:30 am 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3291
defaultuser wrote:
I would encourage anyone having difficultly to put together a specific case that is giving them trouble, not just a vague "having trouble doing X". Fully describe the situation, what you tried, your expectations, and how your results varied from that. That's similar to what we used to tell people to do on the comp.lang.c Usenet group when faced with programming problems.

Agree completely here: expectations are particularly important, those of us writing in game documentation generally know exactly how it's supposed to work, so writing instructions for someone expecting something completely different can be really hard, if we know what players expect/think ought to happen then we can explain what's actually meant to happen a lot easier.

_________________
Mat Bowles

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


Top
 Profile  
 
PostPosted: Sat Sep 24, 2016 11:04 pm 
Offline
Krill Swarm

Joined: Sat Sep 24, 2016 9:58 pm
Posts: 10
As many have already said: Great post!

As a long-time "Stars!" fan, I find old habits hard to break, and learning the new game system has been a combination of un-learning old things as well as learning new things.

8)

O


Top
 Profile  
 
PostPosted: Thu Dec 08, 2016 11:18 pm 
Offline
Space Squid

Joined: Thu Dec 11, 2014 3:22 am
Posts: 53
Hi,

Quote:
On to the rather long list of research projects:

Algorithmic Elegance
Planetary Ecology
Mass Driver 2
Subterranean Habitation
Mass Driver 3
Robotic Production, Military Robotic Control (just pick the latter; the game will automatically add the former)
Mass Driver 4
Nascent Artificial Intelligence
Nanotech Production, Adaptive Automation
Orbital Construction
Fusion Generation, Orbital Generation
Microgravity Industry

This is much of the early game!

But I think that the order depends on species (yes, human is specified) and the map.

For most species, Adequate planets are inadequate without at least one tech. So if there are no Good planets around to colonize, but there are Adequate planets, the first two techs should usually be Ecology and Subterranean. A research-focused strategy also prioritizes these 2 techs, since increasing population on the home planet increases both research and industry, while providing other benefits. Conversely, retaining production focus makes Nascent AI a critical priority.

To the extent that Good planets abound and Adequate planets do not, Eco+Sub can wait. Narrow species have a hard time with anything other than Good until much later in the game, so are less likely to benefit.

So two clusters so far:

AE+AI (usually second, sometimes first) and EcoSub (usually first, sometimes deferred). Mix n match can be reasonable.

Unless there is extreme pressure, researching better MD can wait. In the best of all worlds, ignoring MD entirely is great, though this is usually not possible for most species on many maps. But research-heavy species can usually do this.

MD2-4 is a third cluster (start when first new warship with MD begins to be built, which may never happen.)

What next? Production focus likes Robotic Production. Everyone likes AA. A total focus on production suggests Fusion right after RP, and deferring AA which is very expensive and less important relative to existing production.

But it's necessary to balance lots of production versus useful things to produce.

An optimistic early game might look like:

Eco
Sub
AE
AI
RP
NP
AA
Laser (unless early MD is preferred, especially for Good Pilots)
Force Fields
MRC
Zortrium
Laser2-4 (as first ship rolls off production line)
***mid-game***

But Orbital Construction might be needed for supply lines. Earlier ships might be needed to deal with a nearby AI. NP+AA might need to go after the four (maybe 3) improved military techs, when it is better to build a military than build lots of civilian ships, such as when there's not much to colonize without use of force.

Microgravity is an expensive tech that also requires an Outpost on an Asteroid in a system with production-focused colonies, and an expensive prerequisite. I might be missing something, but I don't see this as an early-game tech.

Robotic Hull is more expensive than Asteroid Hull... but a much better value: +15 speed, damage control, +1 Fuel vs slightly better Structure and +1 internal slot. In a head-to-head battle, an Asteroid fleet of equal PP value at full health will win. But all things are not equal: The faster Robofleet is more likely to be where and when you need it to be. The more resilient roboships are more likely to be at full health when you need them to be, after monster killing duties. You also get them much sooner than Asteroid Hulls, which require OC+Micro as prerequisites, and then an Asteroid Processor (and an outpost that needs exobots and lots of growth tech to become a colony.)

All in all, I think I'd rather spend early RP on AA+robohulls than OC+Micro+AH.

Shields are nice for clearing Maintenance Ships, which there seem to be lots of.

[/quote]
Scout as much as possible. Knowledge of your surroundings is key to putting your Colony Ship to good use. Note that with that colony ship you can colonize one more jump out of your supply lines compared to an outpost ship. (In theory you can colonize anywhere with your Colony Ship, but try to keep your supply lines connected.) Try to colonize before the Nascent AI tech arrives. But apart from that take your time and find the best planet available. Sure, we would of course like to settle as soon as possible, but there are opportunity costs at work here. So, do not rush, but also do not hesitate.
[quote]

I think that on balance, it is best to colonize a good world ASAP. The wasted colony ship benefits are more than compensated by earlier growth and one less hull. If an Adequate or Good world is visible, it's fine to outpost a good world in the home system and move the colony ship. But most species won't know for a few turns whether there's a good medium+ world just out of supply range (outposting a small or tiny world just out of range brings it into range).

Anyway,

Ken


Top
 Profile  
 
PostPosted: Fri Dec 09, 2016 9:48 pm 
Offline
Vacuum Dragon

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 506
I always start research with Adaptive Automation, then scoot Nascent AI up one spot. I want to get to the two flat-rate adds as quickly as possible. After that it's situational to an extent. As noted, if you have Adequate planets early, then at least Planetary Ecology. If there are early warship needs, slot Military Robotic behind Robotic Production.

These days there's no need to research the Mass Driver upgrades early. Build the ships first and research the weapon upgrades when needed.


Top
 Profile  
 
PostPosted: Wed Feb 01, 2017 6:21 pm 
Offline
Space Krill

Joined: Wed Feb 01, 2017 5:26 pm
Posts: 3
Should say up front that I appreciate the heck out of the work that's been done on this game. I played a fair bit of Space Empires IV back in the day...obviously I loved it, but you get bogged down in the logistics. FreeOrion is a huge improvement on that (FWIW, I'm playing 0.4.5 in the Ubuntu package repository). I do miss the tactical combat and the variegated weapon trees in SE IV (null-space projector!!), but I've found to my surprise that another thing I appreciate about FO is the delightfully batshittedly tangled web of techs, rather than SE IV's more linear and plodding (aside from weapons) progressions. Also I appreciate that the AI is smart enough to beat me, Ph.D. or no :roll:...at least for now.

I've only played two games, a quickstart game with humans and a game as George (picked it solely because the race name is so hilarious) with, I don't remember exactly, but a largish cluster map with 15-20 systems per AI, fair number of monsters & neutrals, aggressive max AI, oh and low starlane density. I myself have a perhaps perverse liking for deliberately setting out my economy and getting a ways up the tech trees before going to war with some quirky slant to my ship designs compared to the AI's (that was how I generally played SE IV). I'm a bit embarrassed to say that in both games I've gotten beaten back by one or more of the AIs. The human game I can say I was lost in the tech tree and had no clear direction and made lots of mistakes in not using resources, etc. The George game I am still making mistakes, but perhaps my biggest problem was deliberately deciding to avoid robo and asteroid hulls and go full bio. I suspect you can make it work even against an aggressive AI if you've got your plans in place and think through the implications of population techs, plan out research vs. industry systems better than I have with only one game under my belt, and be aggressive enough. I didn't pay attention to how weak George was at research, nor to the fact that my other neighbor in a choke-pointed southeast corner of the map was Cray and that really I should have gone to war and dropped troops on that bad boy pronto and have the Cray start doing my research. Still, the bio hulls are rough. Very little innate structure even after 30 turns, and my warships ain't getting 30 turns to grow. I am cranking out ravenous battleships with death rays and xentronium armor as fast as I can and am still getting rolled over by those cursed self-gravitating hulls with plasma and diamond. At this point I may have to abandon the northwest corner of my empire (where my Xeno and Cellular facilities currently are), fall back to the homeworld, and protect it with that starlane deviator tech while yellow and red (red is the alpha AI) go at each other for a while, I research the self-healing bio flagship hull, and then we'll see.

The weirdness of the tech tree does mean that you pick up all those population increase techs on the way to your endosymbiotic and ravenous hulls. I think leveraging that could still be a viable strategy against an aggressive AI. The bio hulls are fast, and I do sometimes catch an AI fleet on both sides of a starlane jump that takes me one turn but them two.

But the next game I start, I think I'm going to see how quick I can get fractal hulls...!

Oh, the other thing on which I've wasted some amount of direct research effort AND mis-selection of hulls to research is stealth. It's too easy even for the AI to catch my stealth scout. It's a real pain that you have to be so far into the game even on the settings I chose to get scouts that could pass monsters. By that point, your (successful) enemy empire is backed up on the other side and will see you even if the maintenance ship doesn't.


Top
 Profile  
 
PostPosted: Thu Feb 02, 2017 6:54 pm 
Offline
Programmer

Joined: Mon Feb 29, 2016 8:37 pm
Posts: 204
Welcome def-mornahan.

You seem to be having fun so I won't provide any unasked for help. However, if you have any questions, there are a few how to play descriptions on the forum and people will answer questions here.

One thing. The latest version is 0.4.6. There are a few changes changes. For Ubuntu, you will either have to compile from source, or ask whomever does the FreeOrion packaging for Ubuntu for help.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 38 posts ]  Go to page Previous  1, 2, 3  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group