FreeOrion

Forums for the FreeOrion project
It is currently Wed Dec 13, 2017 1:03 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  [ 91 posts ]  Go to page Previous  1 ... 3, 4, 5, 6, 7  Next
Author Message
 Post subject: Re: My Feedback
PostPosted: Fri Jun 10, 2016 9:41 pm 
Offline
Vacuum Dragon

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 506
If the colonies starts are of different species, I believe that you can start two at the same outpost. The rules won't allow two of the same type to start, but different ones I think you can do. That would be another way. Again, why the AI logic would do that is another question.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Sat Jun 11, 2016 12:21 am 
Offline
Space Floater

Joined: Sat May 21, 2016 9:11 pm
Posts: 33
Why exactly does the AI add another race colony to an existing populated planet?

Many times I will capture a planet with an existing race, only to find a another race colony in the build cue.

Not that I can do that. I have depopulate with a concentration camp before I can add a new racial colony.


I also finally used that super bomb thing for the first time. I likely won't try that again very often. Way to clumsy to use. Since you have to build it from your side and trigger it from your colony, and your ship has to pass through your lines you are just as apt to blow yourself up if you aren't careful. It also kills the star and planets. Empty systems just plain suck and there is already void and disruption generators that achieve mildly similar effects. You can move planets too, if your late game goal is to create a long path with no planets to get support from. The bomb being late game itself means you have access to these planet moving techs at that point. With the late game tech point research steamroll all of these options are hitting about the same time.

I realize it may be hard to program that way you like but the bomb trigger really should be on ship option panel command. Split fleet, merge fleet, trigger weapon, rename, scrap, etc. And lesser strength options should be possible. Perhaps something that just scours life away with hard radiation leaving a system full of radiated planets, and then perhaps something that just does "X amount" of damage to system ships. If you worry about their game effect then let the technology be accessed at the same time. Surely once you have the power to turn an entire system to dust you can be given a fleet or life killer bomb. It's just late game stuff for players wanting a perfect victory to satisfy their game OCD compulsions.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Sat Jun 11, 2016 12:09 pm 
Offline
Dyson Forest
User avatar

Joined: Wed Jul 16, 2014 6:14 pm
Posts: 212
Location: Black Forest
defaultuser wrote:
If the colonies starts are of different species, I believe that you can start two at the same outpost.


Another possible explanation could be that an outpost was captured by a race, which couldn't go on and finish an already begun colony-building. This then gets stuck greyed out in the queue and most likely, you or the AI just enqueued another suitable species.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Sat Jun 11, 2016 6:40 pm 
Offline
Vacuum Dragon

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 506
Kassiopeija wrote:
defaultuser wrote:
If the colonies starts are of different species, I believe that you can start two at the same outpost.


Another possible explanation could be that an outpost was captured by a race, which couldn't go on and finish an already begun colony-building. This then gets stuck greyed out in the queue and most likely, you or the AI just enqueued another suitable species.

That's another possibility, although I often see the "never build" one as exobots.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Sat Jun 11, 2016 6:43 pm 
Offline
Vacuum Dragon

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 506
IamZeke wrote:
Why exactly does the AI add another race colony to an existing populated planet?

Many times I will capture a planet with an existing race, only to find a another race colony in the build cue.

Not that I can do that. I have depopulate with a concentration camp before I can add a new racial colony.

That's what we've been discussing. You can do that with a few of the techniques mentioned, like starting a new colony before a previous one finishes or colonizing with a colony ship or base at an outpost that is currently building a colony. You will notice that the colony that is added to your queue can never finish building.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Sun Jun 12, 2016 12:58 pm 
Offline
Dyson Forest
User avatar

Joined: Wed Jul 16, 2014 6:14 pm
Posts: 212
Location: Black Forest
It even can be of an advantage, eg. if a system is reclaimed then the race can continue their colony building. Or if a colony is destroyed by an attack depopulating it etc pp then you could also decide to finish the building that's already been started.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 4:28 am 
Offline
Space Floater

Joined: Sat May 21, 2016 9:11 pm
Posts: 33
It's been a couple weeks and a dozen games since my last post. Still trying new things.

I've definitely come to prefer cluster galaxies to reduce the number of empty systems. Big thanks on that tip earlier.

I've noted 3 things worth taking about.

First is the somewhat irritating nature of new colonies and teraforms where it requires another turn delay to actually implement. Sure, it's easy to figure out when it actually happens because it's just one turn later, but getting rid of the delay or adding another turn would be a good "polishing" improvement. Not something that needs to be rushed to solve, but a worthy tweak at some point to reflect what the time to build really is.

Second is a derivative of the comment frequently made that the AI has no cheats. But I do believe there is one. Unless an opponent ship is traveling a long or multiple jumps where you see it between systems the player doesn't know until the turn it happens when an AI moves. One turn you see the ship in one system and the next turn it can be in another next door. The player doesn't know in advance about the move. But the computer can know about your movement in the half turn when you set the destination and before you actually arrive. It has that moment when you hit the turn button to respond before turn resolution. At least it really appears so.

I can defend a system with a ship stack and let it set there for many turns. Then should I move it out in another direction the AI seems to know to jump in the very turn the player leaves, whereas if the AI does it, the player has to see the move happen first and can only respond the next turn. By rights, if the AI can't see until after the move is made then the player could move away one turn and then come back. The AI would never be able to land on the undefended system if you came back the next turn because you would be moving back as the AI sees the opportunity to enter unopposed.

Maybe the game can't see and I'm guessing wrong, but when I've seen derivatives of my belief while chasing their fleets in whack-a-mole fashion more times than I can count it goes beyond the pale of simple randomness. The AI has the opportunity to know this, even if you say it doesn't do it. You enter your command to move and then hit the turn button. The AI has the in-between turn moment to see the command and move. Am I suggestion it be fixed? Not necessarily. But the player should know the AI can see your command to move in advance. Leave it you choose but be open about it, or find a way for the AI not to get the data of the move until turn resolution starts.

Third thing to mention is that I agree that research is too fast, but it is also too slow in the beginning too. I can barely get the basic early boosts of growth, research, production, supply, sensors, and the first decent robotic ship with a few upgrades by around turn 100. But by turn 200 I'm closing out so many researches I'm not really making use of any of them. While nice the sentient and titanic hulls almost never get used. I flat out never bother with all those various organic hulls. The fact is that by the time I'm building a decent sized stack of nanorobotic hulls I'm closing in on completing my solar hull research. In truth I'm running most of the buildup game on basic robotic hulls and shifting to solar hulls to win the game. What happens in between blows by so fast as to not be useful. I doubt I've ever used any shield type other than black or multispectral. Only weapons I use are laser 4's and deathray 4's.

I'd much rather get some of the early techs faster. Say one decently armed and armored basic robot hull, plus the first couple basic boosts to production/research/supply/sensors/growth by turn 50 instead of 100, but then really increase the research point cost so that by turn 200 I'm only about halfway through the tech tree instead of being most of the way through. If I'm not in the middle researching transcendence by turn 300 then I'm having an off day. Changing the rate higher in the later research would dovetrail great with my larger hull size and ship upgrade suggestions earlier.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 6:01 am 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12041
Location: Munich
IamZeke wrote:
But the computer can know about your movement in the half turn when you set the destination and before you actually arrive. It has that moment when you hit the turn button to respond before turn resolution.
Yes, but there is no chance to order fleets (or anything else) between that mid-turn update and fleet movement, because that update comes after fleets have moved for a turn. Any move orders issued then would only be acted on on the following turn. The human and AI players get that update at the same time, which is after all orders for a turn are received and executed.
Quote:
But by turn 200 I'm closing out so many researches I'm not really making use of any of them.
Can you post a few screenshots of mid-to-late game tooltips of research on some of your bigger planets? I want to know how much various things (techs, buildings, etc.) are contributing to the total for a well-developed world / empire.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 9:47 am 
Offline
Creative Contributor
User avatar

Joined: Fri Jun 28, 2013 11:45 pm
Posts: 3291
The big problem with research is that there are almost no effective boosters until about turn 100ish (in most games for me), then 3 come along all at once, with another three following soon afterwards.

Nascent AI and Algo Elegence give you a boost at the beginning but it's not massive, it's balanced, then there's nothing (unless you've a telepath) until you can afford Quantum Networking, Stellar Tomography and Distributed Thought which, normally for me, come at roughly the same time one paying for the other. If you've a telepath it changes and you get distributed thought significantly earlier.

Almost immediately after that you get Enclave of the Void and Collective Thought, then there's Pure Energy Metabolism, so by turn 150 you have huge boosts and, unless you're going for a tech win, you can turn most of your worlds to production and still have huge research numbers.

I also tend to complete Quantum Networking at the same time as Force Energy Structures, which means the massive target value boost the former gives is acheivable very quickly due to the latter, I'm fairly sure F/E structures is the most individually powerful/overpowered tech in the game at the moment.

Rebalancing this has been something we've discussed a LOT already, but always in vague terms because it'll, by necessity, require an AI rewrite that'll take time to bed in.

My not-fully-worked-out plans are to move Stellar Tomography earlier in the tech line (possibly simply just to after Gravitonics rather than the expensive-at-that-point Theory of Everything) and the same for Enclave of the Void, but make that work in a similar way to Industrial Centre, with late game refinement techs that improve it rather than the single huge boost it currently gives.

But at the same time as doing that we need to recost almost every other tech, especially the late game techs, and while making tweaks here and there is viable for the sake of the AI team it might be better to do most of it at roughly the same time we rebalance all the bonuses when Influence is ready to go (which will massively change the balance of the game as it's effectively another focus setting you need to pay for stuff).

Having said all that, if the AI team say that their plans to redo researching (which I know is happening but I don't understand) could take into account changes relatively easily in the same way that the AI now takes account of ship hull tweaks dynamically, then we could make some tweaks to the tree very quickly.

_________________
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  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 1:16 pm 
Offline
Dyson Forest
User avatar

Joined: Wed Jul 16, 2014 6:14 pm
Posts: 212
Location: Black Forest
I usually ignore most research-enhancing techs (except the early, easytoget ones) because I find they rather distract you from getting the stuff you really want. In the beginning to get them just takes too much time.

Still, research in continously growing because of a constant increase in planet count + population increase. At some time around midgame then there will be this situation when you get so much excess research that you don't know where to put it, every tech that seems valuably is already selected but you've still got overflow techprod...

@IamZeke if you're having to chase the AI around you either need faster ships, or ones with better stealth. The AI definitely cannot react to your movement if you'll hit him in a single turn. (if he jumps away then this decision has been made the turn before - the only thing I wonder is why this sometimes shows and sometimes not--- I mean the parking place of a ship is displayed right while a leaving ship left... but sometimes even ships that are parked jump away)


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 5:40 pm 
Offline
Programmer

Joined: Sun Feb 14, 2016 12:08 am
Posts: 359
Before recosting the techs, maybe introduce an efficiency malus based on the number of contributing pop centers, e.g.:
Code:
priority = 200
effects = SetTargetResearch value = Value / max(1, [[POPCENTER_COUNT]] ^ 0.05)

POPCENTER_COUNT
'''Count condition = And [ PopulationCenter OwnedBy empire = Source.Owner Research low = 0.1 ]'''


While the same could be done for industry, it should probably be grouped by supply line.

_________________
Any content posted should be considered licensed GNU GPL 2.0 and/or CC-BY-SA 3.0 as appropriate.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 8:07 pm 
Offline
Creative Contributor
User avatar

Joined: Sat Sep 13, 2003 6:52 pm
Posts: 1396
Another option is increase tech costs based on existing techs

(so that you can go for techs A, B, or C in the same amount of research... but getting A+B costs more than 2x as much as just A)


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 9:43 pm 
Offline
Space Floater

Joined: Sat May 21, 2016 9:11 pm
Posts: 33
Moving another booster to early game will help, or will giving more boost to an existing early one would do the same. So would simply reducing some of the early research point costs for some of the basics.

But then you get the planet numbers steamroller going and the researches start flying through. Mind you, this helps if you get trapped on a map with severe roadblocks early on limiting your planet count and I'm not sure how to solve that problem if you raise costs on later research.

But if you can get a breakout to a reasonable number of planets then research becomes such a blur after around turn 150 that you rarely bother with all the hull and weapon types. All you need is a sufficient blocking and protection force made from 1stgen robot ships to guard your area and in no time you can start cranking out massive solar and titanic hulls that win the game with sheer brute force. Like moving from a squirrel rifle to a howitzer.

The more I think about it the more cloudy a solution becomes. The intensely variable setup nature of this game can leave the player with wildly different early opportunities to colonize. So easy research fixes that come to mind each become problematic depending on how the star map layout unfolds.

Perhaps the solution lies in planet development. Making the player choose between quantity or quality of planets. Perhaps adding research centers that boost a planet's research capacity, but limit their building the way ships are done. Right now a player can be building a dozen shipyards or lighthouses and still be building ships from the combined leftover pool of PP, but the next group of ships stays blacked out awaiting the PP pool to be freed up.

So if you make planetary research booster buildings build like ships then they can only build them one at a time. This would be an issue for players trying to boost a lot of planets at once, but for the player either forced or choosing to bulk up a few select planets then it isn't a big problem. Then you couple that with increase research costs across the board. Then the player can't steamroll the research tree.

I realize that's just a conceptual idea without the difficulties of playtesting to get the numbers right, but it's about the only way I see making a system that works for both large and small planet groupings and allows for increasing research costs across the board.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Mon Jun 27, 2016 10:04 pm 
Offline
Programming, Design, Admin
User avatar

Joined: Wed Oct 08, 2003 1:33 am
Posts: 12041
Location: Munich
Geoff the Medio wrote:
Can you post a few screenshots of mid-to-late game tooltips of research on some of your bigger planets? I want to know how much various things (techs, buildings, etc.) are contributing to the total for a well-developed world / empire.
Various modifications to the resource-output-boosting techs are possible, including things like switching
Code:
value = Value + [[SCALE_FACTOR]]*Target.Population
to
Code:
value = Value + min(5, [[SCALE_FACTOR]]*Target.Population)
or making techs act on the 5 most-population planets, rather than all planets, or having one (or 5) per-empire buildings that give a fixed bonus where they are produced, rather than a direct tech effect. The suggestion of being limited to producing one at a time is also interesting, without the need to limit how many can be controlled once produced. As long as such a building isn't too expensive to produce for a small empire, these sorts of changes limit the relative benefit of having a big empire, and make research output (or other resource output) more linear, rather than quadratic or faster growing with time or empire size.


Top
 Profile  
 
 Post subject: Re: My Feedback
PostPosted: Tue Jun 28, 2016 3:48 pm 
Offline
Vacuum Dragon

Joined: Wed Aug 26, 2015 6:15 pm
Posts: 506
Yeah, I haven't seen any evidence that the AIs get mid-move information. In fact, a technique I use is to lure enemy fleets off heavily defended systems to a neutral location or my system by vacating one turn then moving back the next. Sometimes the AI sees me leave and then enters a move to that same system location where I'm returning. That way I can get a match-up that doesn't include their planetary defenses.

I would worry more about later-game research if that were a stumbling block. In the size games I play (I prefer 150 systems for time + lag considerations), generally I have the tech I really need at a reasonable point. After that research really isn't important, especially if ahead of the AIs.

Sometimes in a game I get off to slow start and an AI lands in a rich area, so I can find myself behind at mid-game. Then I have worry about getting Pure Energy Metabolism and possibly Artificial Black Hole (if no colonizable ones present) and developing tougher ship tech than I normally need to employ.

It would be more interesting if there were some earlier research boosts combining tech and buildings similar to gas giant generators or solar generators. Something along the lines of research stations (single system boost) or university systems (empire-wide). Then it becomes a decision using those precious earlier RP and PP amounts on boosting research, increasing production, or devote to military.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 91 posts ]  Go to page Previous  1 ... 3, 4, 5, 6, 7  Next

All times are UTC


Who is online

Users browsing this forum: Google [Bot] 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:  
Powered by phpBB® Forum Software © phpBB Group