Foraven wrote:
If you can play those cards with impunity, it means there is a flaw in the AI's logic. There shouldn't be any "safe" or "Jack of all trades" cards to play.

If all cards were about equally likely to be countered, I would simply pick no card as not to make things worse. I do not need the boost, but giving the opponent one could cost me ships. Why risk that?

Having an Hero that spend all it's points in buffs would always be much stronger than one that spent a few points in it and then "wasted" it's points on non combat/support ones. The way it is right now, you can safely pick non combat skills without gimping your hero, under your system everyone would be forced to pick all the combat skills if they want to remain competitive (remember i also think Multiplayer, won't make a different version for them).

The opportunity costs for choosing a non-direct-power-skill would be massively lower when now since one power-skill would give much less of a buff (in my out of the blue example, 6 to 9 percent instead of the current 12 to 20). Your cross-class approach could add to that if done right (i.e. massively buffed effects from wit/labor for fleet utility skills), making non-perfect (i.e. not commander/pilot) heroes closer to the perfect ones by giving significant buffs for those otherwise wasted economic stat points.

That's because you don't understand how much code is involved for the AI to make those "simple" decisions. You forget all the instances where you have to chose between building a new improvement, some more ships, convert to food or science or switch to a different exploitation because your Empire is short on something. It may seem no brainer to you, but it's quite complicated (especially when it's done with weights instead of if-then-else code). The governors are a huge part of the AI and they work exactly the same for the players (exept they don't queue up ships).

Oh, I understand that perfectly. It wasn't a criticism directed at you, but the actual game designers of those games. Compared to the decisionmaking processes when writing an entire empire AI, a build queue is pretty easy stuff. The governors themselves could work like I described above and would always be better than what all those games currently have. There would just have to be a second layer of routines above that which would overwrite their programming to insert ship building or otherwise empire wide strategic things into appropriate worlds. That latter part is a bit more complicated, but the systems governor is not. Heck, I usually build ships in about two, at most three different systems because that more than suffices - other systems basically build improvements and sit on conversion the rest of the time.

stormfox wrote:
If all cards were about equally likely to be countered, I would simply pick no card as not to make things worse. I do not need the boost, but giving the opponent one could cost me ships. Why risk that?

Here i don't understand your logic; not playing cards IS making things worst. I guess i do fight (and win) lots of battles you just don't engage in unless you have no choice at all. Be fortunate the AI isn't as risk adverse, chasing fleets all over the map wouln't be fun (Sins of a Solar Empire had AI like that initially, it was so risk adverse that it did not even bother to defend it's colonies unless it had a massive fleet to do so, thus made it easy to beat.).

The opportunity costs for choosing a non-direct-power-skill would be massively lower when now since one power-skill would give much less of a buff (in my out of the blue example, 6 to 9 percent instead of the current 12 to 20). Your cross-class approach could add to that if done right (i.e. massively buffed effects from wit/labor for fleet utility skills), making non-perfect (i.e. not commander/pilot) heroes closer to the perfect ones by giving significant buffs for those otherwise wasted economic stat points.

I just don't feel like adding tons of new skills. It's easy to add the code, but adding the descriptions in multiple language and providing icons is a chore. That's why i'm never to add completely new skills; it's just too much hassle.

Oh, I understand that perfectly. It wasn't a criticism directed at you, but the actual game designers of those games. Compared to the decisionmaking processes when writing an entire empire AI, a build queue is pretty easy stuff. The governors themselves could work like I described above and would always be better than what all those games currently have. There would just have to be a second layer of routines above that which would overwrite their programming to insert ship building or otherwise empire wide strategic things into appropriate worlds. That latter part is a bit more complicated, but the systems governor is not. Heck, I usually build ships in about two, at most three different systems because that more than suffices - other systems basically build improvements and sit on conversion the rest of the time.

The AI already work that way; It runs on governors but with a routine that override build queue to add ships or to switch governor when needed. Making it do it isn't the hardest part, it's making it do it when it makes sense that is. Human players are always better at min-maxing than programmers are at making the AI do the same.

Foraven wrote:
Here i don't understand your logic; not playing cards IS making things worst. I guess i do fight (and win) lots of battles you just don't engage in unless you have no choice at all. Be fortunate the AI isn't as risk adverse, chasing fleets all over the map wouln't be fun (Sins of a Solar Empire had AI like that initially, it was so risk adverse that it did not even bother to defend it's colonies unless it had a massive fleet to do so, thus made it easy to beat.).

Sins still has that problem sometimes, but I don't. But on topic - playing NO card would be better whenever I would not have a very big chance of getting my cards through - because getting nothing and him getting his cards basic benefit for free is better than me getting nothing and him getting the counter bonus in addition. It's absolutely not me not fighting (in my most recent game, I killed an AIs homeworld around turn 30), it's that the cards are only relevant in close battles (which rarely ever happen) and even then its more important that I lose as few ships as possible than wether I kill him five or ten times. AI fleets that are not superior to mine die on a confrontation, or get destroyed to at least 3/4, even if I pick defensive bonuses that actually lower my damage output.



Example: The one battle said destroyed AI put up was with his hero fleet against my hero fleet. Same number of ships (5 each), similarly built, mine were just a tad better (I had already built in the next kinetics, he seemingly did not and I had a power module). Our heroes were similar, too (mine was level 3, his level 2). He was about 90% of my strength, and the battle was relatively important because I knew if I won this, he is dead and I can start invading his homeworld. I played engineering-accuracy, engineering-repair, engineering-repair and won with no more than slight hull damage. I could see my ships getting low after each salvo, but the repair cards later swung them back up and I did not lose a single ship. He lost all.

I just don't feel like adding tons of new skills. It's easy to add the code, but adding the descriptions in multiple language and providing icons is a chore. That's why i'm never to add completely new skills; it's just too much hassle.

I had a look at the xmls today - I see your point, it's pretty tedious. Perhaps I will try my variant of your mod out if I find the time - it would likely not need much new stuff, just recycling of known things and some number changes. For example, making 4 offense upgrades with 2/2/2/2/2 out of the current 3 with 3/3/4 is easily done with some copy-pasting. I wait for your next update, though - I really like your mod as a baseline and would hate to have to change things every few days when you update :-)

The AI already work that way; It runs on governors but with a routine that override build queue to add ships or to switch governor when needed. Making it do it isn't the hardest part, it's making it do it when it makes sense that is. Human players are always better at min-maxing than programmers are at making the AI do the same.

The latter makes no sense for me. Its no sorcery to do that, and those programmers do that for a living. Can't they take one day to formulate a perfect (or almost perfect) standard build order? It's the same with obvious balance issues in games - every player could make the necessary changes to the big offenders in five minutes, why can't the designers?

stormfox wrote:
Example: The one battle said destroyed AI put up was with his hero fleet against my hero fleet. Same number of ships (5 each), similarly built, mine were just a tad better (I had already built in the next kinetics, he seemingly did not and I had a power module). Our heroes were similar, too (mine was level 3, his level 2). He was about 90% of my strength, and the battle was relatively important because I knew if I won this, he is dead and I can start invading his homeworld. I played engineering-accuracy, engineering-repair, engineering-repair and won with no more than slight hull damage. I could see my ships getting low after each salvo, but the repair cards later swung them back up and I did not lose a single ship. He lost all.

Don't forget your racial traits; the AI may have similar designs but if you have bonuses he doesn't have, your ships are effectively stronger (even if the MP is the same). I should code something about that, maybe the AI would weight it's chances better if it had more accuracte numbers to play with.

I had a look at the xmls today - I see your point, it's pretty tedious. Perhaps I will try my variant of your mod out if I find the time - it would likely not need much new stuff, just recycling of known things and some number changes. For example, making 4 offense upgrades with 2/2/2/2/2 out of the current 3 with 3/3/4 is easily done with some copy-pasting. I wait for your next update, though - I really like your mod as a baseline and would hate to have to change things every few days when you update :-)

Feel free to mod that in if that makes you happier .

The latter makes no sense for me. Its no sorcery to do that, and those programmers do that for a living. Can't they take one day to formulate a perfect (or almost perfect) standard build order? It's the same with obvious balance issues in games - every player could make the necessary changes to the big offenders in five minutes, why can't the designers?

Designers (and modders) generally don't play the game the same way competitive players do. Depending on what traits you pick, the optimal build order change. Also, depending on what systems you have to colonize, the build order can change; even more so if you are short on Dust and have to chose what is more urgent to buyout. Making the AI blindly follow build orders without any thoughs for issues and priorities is just making it crash it's economy.

I just wanted to chime in that stormfox's approach to combat cards is very similar to the one that I endorse, and that he is not just making stuff up Although I take a more defense neutralization approach based on the ineffective defenses of the earlier versions.) The cards provide the illusion of an interesting system, but really just boil down to minimizing your chance of disaster.

thuvian wrote:
I just wanted to chime in that stormfox's approach to combat cards is very similar to the one that I endorse, and that he is not just making stuff up Although I take a more defense neutralization approach based on the ineffective defenses of the earlier versions.) The cards provide the illusion of an interesting system, but really just boil down to minimizing your chance of disaster.

Well, i never felt that way when playing the cards; i try to get the edge or diminish the one of my opponents, i'm not overly concerned about the minor extra bonuses he can get from countering my cards (i know the code, it's not like it's +50% or +100%). I do enjoy the cardfight even if it's not my first choice to spice the combat (combat tactics would have been better in my opinion).





On a side note, i'm currently running into a strange bug in my current test version; i colonize a planet from the build queue and i don't have a colony when it's built. As far as i can tell it never happened in my earlier builds, though i would like to know if other players have had that problem with my mod. Second bug is i can't change the monopoly value; for some unfantomable reason it remain at +20 no matter what i set it to. I'm running mad with my strange bugs, i have to take a break before i start to smash my keyboard...

thuvian wrote:
I just wanted to chime in that stormfox's approach to combat cards is very similar to the one that I endorse, and that he is not just making stuff up Although I take a more defense neutralization approach based on the ineffective defenses of the earlier versions.) The cards provide the illusion of an interesting system, but really just boil down to minimizing your chance of disaster.

Thanks. Very well phrased and sums it up good. I will have to steal that phrase (...minimizing your chance of desaster) sometime :-)

Foraven wrote:
Don't forget your racial traits; the AI may have similar designs but if you have bonuses he doesn't have, your ships are effectively stronger (even if the MP is the same). I should code something about that, maybe the AI would weight it's chances better if it had more accuracte numbers to play with.

The one thing I have is aggressive (or whatever the +dmg racial is called) on level two I think. For one, I would have thought it would be part of the strength calculation, seeing as adding a hero or making an existing admiral better influences the displayed mp, too. Besides this, I do not see a great difference between fighting sophons or hissho, for example.



Said recent game where I crushed the hissho before turn 50 turned out to be one of those "gang up on the human" games - both my neighbours bombarded me with (bad) fleets nonstop, fleets they of course pulled out of their asses and paid no upkeep for or they would be broke. I was never in danger (and if I had been, I could have spent a few turns to whip up another 3 fleets or so instead of sitting on science conversion on 90% of my systems), I even had the time to sometimes juggle around my two admirals between my 3 fleets so they would get xp from as many fights as possible. I had a few losses when the AI started building missiles instead of just kinetic ships, but those went away with my next fleet upgrade. I think the only other losses I had where against some pirates due to their ridiculous strengths and the fact that I had no hero to spare for my 4th fleet (the pirate hunters, which were also freshly built and not level over 9000 ships like the rest of my fleet).



It was basically like playing eternal war besides me actively wanting peace with one of them so I could abuse the trade routes. I guess trade routes boosters and corporate heroes in general have another major problem...

Designers (and modders) generally don't play the game the same way competitive players do. Depending on what traits you pick, the optimal build order change. Also, depending on what systems you have to colonize, the build order can change; even more so if you are short on Dust and have to chose what is more urgent to buyout. Making the AI blindly follow build orders without any thoughs for issues and priorities is just making it crash it's economy.

You usually do not buyout when not swimming in cash, the exception being that first refinery or institute if you can afford it. Everything else gets bought when my economy makes hundreds of dust per turn and is otherwise just built.



Question: Do you destroy structures when invading a system?



If not, there are very serious flaws in the AI since I have had multiple examples of AI systems (including the homeworld of that unlucky sophon) where I did not even find a refinery, which is inexcusable since it should always be the first thing to be built on any colony under any circumstances if you do not have institute yet (and then its the second thing to build).



The AI could never crash its economy because of rampant cheating. It does not have to worry about unhappiness (at least not expansion dissaproval, which is by far the heaviest contributor) or upkeep cost. If it followed my build order (and of course only built money upgrades when the start paying for themselves, i.e. careful sweeping only when at least 2 people are on a planet with moon and so on), it would be impossible to go broke and if getting close to even balances could aways switch one system to conversion to dust.

Foraven wrote:
Well, i never felt that way when playing the cards; i try to get the edge or diminish the one of my opponents, i'm not overly concerned about the minor extra bonuses he can get from countering my cards (i know the code, it's not like it's +50% or +100%). I do enjoy the cardfight even if it's not my first choice to spice the combat (combat tactics would have been better in my opinion).

The cards ARE combat tactics. That is exactly what they represent. You give three consecutive orders to your fleet the opponent does the same. I actually like the simplicity of the system, I just do not like the way it works out.



But really, when your 1544 MP fleet battles the 47th stupid small 848 MP fleet the AI throws at you, why should you care about anything but coming out of the fight in 100% shape so you can keep that constant slaughter up without having to retreat because of damage (or god beware, needlessly lose ships with experienced crews) and without having to build up additional fleets (for which you don't have admirals and that might even have some initial losses because of that and every battle they fought would be a battle your admirals could have gotten xp for)?



I am not above nuking a strong fleet with a throwaway fleet I scratchbuilt for that purpose and then sealing the deal with my hero fleet in the same turn if I cannot kill them reliably with just the hero (usually only happens against pirates because they are so ludicrously strong for the time they appear sometimes), but other then that, there are no reasons to every put an effort in "more firepower" or to risk any ships.

On a side note, i'm currently running into a strange bug in my current test version; i colonize a planet from the build queue and i don't have a colony when it's built. As far as i can tell it never happened in my earlier builds, though i would like to know if other players have had that problem with my mod. Second bug is i can't change the monopoly value; for some unfantomable reason it remain at +20 no matter what i set it to. I'm running mad with my strange bugs, i have to take a break before i start to smash my keyboard...

I have not experienced that, sorry (or luckily, because that means that at least 1.35 did not have this).

stormfox wrote:


PS: You should perhaps make a comment about the no music thingy in your main post here, I only saw how to fix that because I checked the readme for a contact adress of yours or something :-)

I've wondered about that too - it just seems to *stop* after using the mod - question is, how does one fix it???

fredzeppelin wrote:
I've wondered about that too - it just seems to *stop* after using the mod - question is, how does one fix it???

Nevermind... RTFM applies here.... =)

stormfox wrote:
The one thing I have is aggressive (or whatever the +dmg racial is called) on level two I think. For one, I would have thought it would be part of the strength calculation, seeing as adding a hero or making an existing admiral better influences the displayed mp, too. Besides this, I do not see a great difference between fighting sophons or hissho, for example.

Military power is tied to the modules, not calculated from a formula guestimating the strength of the ships. Heroes do improve the MP, but that's just a percent added. Race traits did not boost the MP (added the feature for my next version).

Said recent game where I crushed the hissho before turn 50 turned out to be one of those "gang up on the human" games - both my neighbours bombarded me with (bad) fleets nonstop, fleets they of course pulled out of their asses and paid no upkeep for or they would be broke. I was never in danger (and if I had been, I could have spent a few turns to whip up another 3 fleets or so instead of sitting on science conversion on 90% of my systems), I even had the time to sometimes juggle around my two admirals between my 3 fleets so they would get xp from as many fights as possible. I had a few losses when the AI started building missiles instead of just kinetic ships, but those went away with my next fleet upgrade. I think the only other losses I had where against some pirates due to their ridiculous strengths and the fact that I had no hero to spare for my 4th fleet (the pirate hunters, which were also freshly built and not level over 9000 ships like the rest of my fleet).

You know i spent a lot of time tweaking those "bad" designs you speak of and trying to force the AI to use more variety of designs than the handful of design templates the devs had included. I'm saddened you did not find any improvements over what the stock game provide.



Did you combine my mod with anything else from other mods? I'm sure something must be wrong, not working or broken because my AI used to provide some challenge a few versions back.

Question: Do you destroy structures when invading a system?

No. Either the AI is broken and did not build them, or you made it run broke and it started to sell them off (the AI has code for that, though Ail knows how that works better than i do).

The AI could never crash its economy because of rampant cheating. It does not have to worry about unhappiness (at least not expansion dissaproval, which is by far the heaviest contributor) or upkeep cost. If it followed my build order (and of course only built money upgrades when the start paying for themselves, i.e. careful sweeping only when at least 2 people are on a planet with moon and so on), it would be impossible to go broke and if getting close to even balances could aways switch one system to conversion to dust.

Not sure what you base your opinion on, but ES AI does not cheat; difficulty settings does give it bonuses or maluses, but it doesn't get any unfair advantages outside of that. If it has ships, it has built them with the tech it has and pay the upkeep.

The cards ARE combat tactics. That is exactly what they represent. You give three consecutive orders to your fleet the opponent does the same. I actually like the simplicity of the system, I just do not like the way it works out.

Do you actually wait to see how the battle unfold before picking the next card? The real advantage of playing manual is you can actually see how the first round turn out and decide what the next round should be, if you don't that may be why you feel your cards always get countered unless you play it extra safe...

But really, when your 1544 MP fleet battles the 47th stupid small 848 MP fleet the AI throws at you, why should you care about anything but coming out of the fight in 100% shape so you can keep that constant slaughter up without having to retreat because of damage (or god beware, needlessly lose ships with experienced crews) and without having to build up additional fleets (for which you don't have admirals and that might even have some initial losses because of that and every battle they fought would be a battle your admirals could have gotten xp for)?

Last time i played a long game with my mod, the AI i was trying to invade had fleets comparable to mine and was capable of giving me a run for my money (i actually lost fleets trying to invade his territories). Most of the AIs were getting ahead of me and i had to rush better weapons because my fleet were not good enough...



Btw, i had instance in earlier version where the AI was spamming early game designs without ever retrofitting them... That made the AI rather weak and also broke as it was having way more ships than it was sane to have; maybe that's what happened in your game (even though i believe i crushed that bug).

I have not experienced that, sorry (or luckily, because that means that at least 1.35 did not have this).

That's a good news... I hope i can figure out how the heck my game can use earlier version of my files than the one i actually run the game with.

Foraven wrote:
Military power is tied to the modules, not calculated from a formula guestimating the strength of the ships. Heroes do improve the MP, but that's just a percent added. Race traits did not boost the MP (added the feature for my next version).

I saw that some other mod (Nova?) redid those values, and I appreciate that you will let the AI actually use all multipliers now (what a glaring oversight on the original AI scripts!). I rarely look at the opponents, though (at best I check which weapon they predominatly have), just the rough numbers and that is enough to know wether to fight or retreat. If I can do that, why can't an AI that could actually calculate its odds?

You know i spent a lot of time tweaking those "bad" designs you speak of and trying to force the AI to use more variety of designs than the handful of design templates the devs had included. I'm saddened you did not find any improvements over what the stock game provide.

Sorry to rain on your parade, really. The main problem isn't even the ship designs imho. Its the fleets. If the AI would build 5 fleets of 9 CP instead of 9 fleets of 5 CP it would fare better in combat and not be as irritating to play against at the same time. Very often I outtech the AI in addition to everything else though (and I usually get the weapons techs last during each "circle"), so lateron they would be the underdogs anyways. Perhaps the jumps in performance from one tech tier to the next are too drastic, too? I have not really looked at the numbers, ever (needlessly complicated), but perhaps that direction is worth it to take a look at.

Did you combine my mod with anything else from other mods? I'm sure something must be wrong, not working or broken because my AI used to provide some challenge a few versions back.

Nope. Your mod is the first and so far only one. Don't take this the wrong way, but perhaps you just build and tech up to slowly and that resulted in your AI giving you a challenge. Neither yours nor the original one usually does that barring an extremely bad start for me.

No. Either the AI is broken and did not build them, or you made it run broke and it started to sell them off (the AI has code for that, though Ail knows how that works better than i do).

It was their home system. I declared war around turn 30 and sieged them for 9 turns or so, then it was mine. Would that really have them sell something off, and then the building that is the most important one until you get institute? Makes no sense. I saw refineries missing in some colonies of the other players, too, btw.

Not sure what you base your opinion on, but ES AI does not cheat; difficulty settings does give it bonuses or maluses, but it doesn't get any unfair advantages outside of that. If it has ships, it has built them with the tech it has and pay the upkeep.

And that is exactly what cheating is. The AI even on normal has an about 20% boost to everything and a ridiculous expansion boost since it does not really have to worry about happiness (which is the most limiting factor by far in the first 100 turns of the game). In addition to that I am pretty much sure that it gets even MORE discount on upkeep costs or the masses of halfassed fleets it throws around would make it bancrupt every time, especially considering how bad it builds up.

Do you actually wait to see how the battle unfold before picking the next card? The real advantage of playing manual is you can actually see how the first round turn out and decide what the next round should be, if you don't that may be why you feel your cards always get countered unless you play it extra safe...

I never play manual (I actually deactivated the option for it because I accidentally clicked it sometimes). I cannot be bothered to take five minutes for a combat that otherwise takes five seconds to do, no matter how pretty it my look. And even if I did, what would I gain from knowing what the first enemy card was? It's not that I can derive any hints for the second card play from that - at best I could react to the situation by playing a more defensive or more aggressive card than I had originally planned because of the preliminary results (and even that would make no sense if the opponent has kinetics, for example). Nope, I can just as well pick my three risk-minimizing cards up front and click "yeah yeah, just tell me how much I won and lets get on with the game".

last time i played a long game with my mod, the AI i was trying to invade had fleets comparable to mine and was capable of giving me a run for my money (i actually lost fleets trying to invade his territories). Most of the AIs were getting ahead of me and i had to rush better weapons because my fleet were not good enough...

If that is the case, your buildup was too slow. A few times in larger games it can happen that some sophon on the other end of the galaxy becomes large and starts being slightly ahead in tech - that usually means you have to crush one or two AIs quickly, seize their territory, and thus add their science output to yours. If he borders you, even better, just kill him.

Btw, i had instance in earlier version where the AI was spamming early game designs without ever retrofitting them... That made the AI rather weak and also broke as it was having way more ships than it was sane to have; maybe that's what happened in your game (even though i believe i crushed that bug).

Nope, they later built missile ships, it's just that they still stood no chance. The first missile ships as I said actually managed to destroy a ship or two every now and then until I retrofitted some flak in, but that's it. Again, I wasn't even using anything besides the basic hull type in my fleets (the frigate?), not even the next best one (destroyer I guess) with 20% more room and cheaper combat equip.



Try it sometime. Build refineries, explot, scout, two colony ships, perhaps another scout. Research all tier 1 stuff beginning with planet types, then economy, and finish with the two military upgrades and then get the T2 thingy which gives the +xp building, too. Build that on your homeworld (and perhaps your second colony), upgrade your basic defense ship to power module, 3-5 anti-kinetics and rest kinetics or missiles, your choice, perhaps add a +HP module. Build 2x5 of those and chose an AI to die.

stormfox wrote:
I saw that some other mod (Nova?) redid those values, and I appreciate that you will let the AI actually use all multipliers now (what a glaring oversight on the original AI scripts!). I rarely look at the opponents, though (at best I check which weapon they predominatly have), just the rough numbers and that is enough to know wether to fight or retreat. If I can do that, why can't an AI that could actually calculate its odds?

As far as i can tell, the AI weight it's odds according to the MP it has.

Sorry to rain on your parade, really. The main problem isn't even the ship designs imho. Its the fleets. If the AI would build 5 fleets of 9 CP instead of 9 fleets of 5 CP it would fare better in combat and not be as irritating to play against at the same time. Very often I outtech the AI in addition to everything else though (and I usually get the weapons techs last during each "circle"), so lateron they would be the underdogs anyways. Perhaps the jumps in performance from one tech tier to the next are too drastic, too? I have not really looked at the numbers, ever (needlessly complicated), but perhaps that direction is worth it to take a look at.

I dunno there why it would not fill it's fleets, in my tests it did. Will have to check that out.

Nope. Your mod is the first and so far only one. Don't take this the wrong way, but perhaps you just build and tech up to slowly and that resulted in your AI giving you a challenge. Neither yours nor the original one usually does that barring an extremely bad start for me.

But i did tweak it's build orders and what tech to rush and all. I guess it's that hard to make the AI perform as well as a human player, so many things we take for granted the AI can't do.

It was their home system. I declared war around turn 30 and sieged them for 9 turns or so, then it was mine. Would that really have them sell something off, and then the building that is the most important one until you get institute? Makes no sense. I saw refineries missing in some colonies of the other players, too, btw.

I will have to do some tests and see what is going on.

And that is exactly what cheating is. The AI even on normal has an about 20% boost to everything and a ridiculous expansion boost since it does not really have to worry about happiness (which is the most limiting factor by far in the first 100 turns of the game). In addition to that I am pretty much sure that it gets even MORE discount on upkeep costs or the masses of halfassed fleets it throws around would make it bancrupt every time, especially considering how bad it builds up.

Having an handicap isn't cheating per see, that doesn't break the rules of the game. It doesn't conjure ships out of tin air, get techs without following the tech tree or things like that. The AI on normal doesn't have any bonuses btw, it plays with what it has and does suffer from expansion maluses. It just doesn't focus on the same things we do, that's why the results differ.

I never play manual (I actually deactivated the option for it because I accidentally clicked it sometimes). I cannot be bothered to take five minutes for a combat that otherwise takes five seconds to do, no matter how pretty it my look. And even if I did, what would I gain from knowing what the first enemy card was? It's not that I can derive any hints for the second card play from that - at best I could react to the situation by playing a more defensive or more aggressive card than I had originally planned because of the preliminary results (and even that would make no sense if the opponent has kinetics, for example). Nope, I can just as well pick my three risk-minimizing cards up front and click "yeah yeah, just tell me how much I won and lets get on with the game".

You are missing half the game then. That's the part i enjoy the most because i'm really good at judging what the AI is likely to play next when watching the battle. When you play automatic, you gimp the AI because it preselect it's cards at the begining of battle rather than pick them as the battle unfold (the AI weight what cards to chose depending on it's fleet and enemy's fleet is, what it will chose may differ massively depending who lost ships or how damaged they are).

If that is the case, your buildup was too slow. A few times in larger games it can happen that some sophon on the other end of the galaxy becomes large and starts being slightly ahead in tech - that usually means you have to crush one or two AIs quickly, seize their territory, and thus add their science output to yours. If he borders you, even better, just kill him.

I don't pretend i'm a great player. Am also not the "go for the kill" type, i do enjoy roleplay my games (thus most likely gimp myself doing so). Also the AI is coded to roleplay as well, it's not designed to be merciless as human players are likely to be.

Nope, they later built missile ships, it's just that they still stood no chance. The first missile ships as I said actually managed to destroy a ship or two every now and then until I retrofitted some flak in, but that's it. Again, I wasn't even using anything besides the basic hull type in my fleets (the frigate?), not even the next best one (destroyer I guess) with 20% more room and cheaper combat equip.

Early rushes will cripple the AI greatly; in the first 50 turns it's in the expansion phase and usually go for the empire buildup techs rather than combat ones. If you want a real challenge, just let it build up it's first few colonies, then it will be more challenging.

Try it sometime. Build refineries, explot, scout, two colony ships, perhaps another scout. Research all tier 1 stuff beginning with planet types, then economy, and finish with the two military upgrades and then get the T2 thingy which gives the +xp building, too. Build that on your homeworld (and perhaps your second colony), upgrade your basic defense ship to power module, 3-5 anti-kinetics and rest kinetics or missiles, your choice, perhaps add a +HP module. Build 2x5 of those and chose an AI to die.

I know there are rush techniques, but i rather not use them. My first concern is getting my enjoyment from the game; killing off the AI when they are at their weakest isn't really fun to me.

BUG



I just tried to build a new race for a "eternal war" setting this time, and noticed that all the diplomatic traits cost 0 and "eternal war" cannot be taken. What gives?

Foraven wrote:
As far as i can tell, the AI weight it's odds according to the MP it has.

If it did that, it would not throw fleet after fleet against a massively superior one.

I dunno there why it would not fill it's fleets, in my tests it did. Will have to check that out.

In the early game it could be that it fills, just does not research the CP buffs fast enough. After a while though, it runs around with a ridiculous amount of total ships (which a human player could never support) but no punch in all those fleets.

But i did tweak it's build orders and what tech to rush and all. I guess it's that hard to make the AI perform as well as a human player, so many things we take for granted the AI can't do.

The point is, again there is no real trick to my research "strategy". The way research costs are set up, you basically research in "rings", sometimes beelining one ring further (often just for one tech) before doing the rest of the round. The only real question is which of the four sides to start at and which to keep for last. If I can build strong fleets from just researching the basic combat techs and then doing no ship upgrades for the next 50-80 turns (which is exactly what I do), I would expect the AI ships to actually become hard to beat at some point because they should be technically similar or even more advanced than mine even IF they neglected military tech too (which should lead to them building up their planets better).



Besides the basic priority problems the AI seems to have, perhaps it would be helped if it built less fleets so more production turns are used to actually do something useful?

Having an handicap isn't cheating per see, that doesn't break the rules of the game. It doesn't conjure ships out of tin air, get techs without following the tech tree or things like that. The AI on normal doesn't have any bonuses btw, it plays with what it has and does suffer from expansion maluses. It just doesn't focus on the same things we do, that's why the results differ.

Its the very definition of cheating. Besides, yes, the normal AI cheats:



http://endlessspace.wikia.com/wiki/Game_settings



I am actually amazed that the endless difficulty AI is actually beatable, given that it buyouts for 10% of the original costs, which means it can buy almost anything it wants in almost limitless numbers. Thats such a ridiculous buff (together with the almost as ridiculous overall buffs to everything else and the fact that their ships are double as strong) that it says something about how bad their build orders and ship designs have to be if they can lose with that.

You are missing half the game then. That's the part i enjoy the most because i'm really good at judging what the AI is likely to play next when watching the battle. When you play automatic, you gimp the AI because it preselect it's cards at the begining of battle rather than pick them as the battle unfold (the AI weight what cards to chose depending on it's fleet and enemy's fleet is, what it will chose may differ massively depending who lost ships or how damaged they are).

It will make as much difference for the AI as for me - basically none. The obvious choices are still the obvious choices, the ships do not suddenly change in their stats mid-batte. If the game would simulate in three chunks, I would likely play three cards, but sorry, I won't watch minutelong battles.

I don't pretend i'm a great player. Am also not the "go for the kill" type, i do enjoy roleplay my games (thus most likely gimp myself doing so). Also the AI is coded to roleplay as well, it's not designed to be merciless as human players are likely to be.

It doesn't have to be min-maxed. It just needs to do something useful for once. The suboptimal faction traits and rpg preferences will take it down a level or two anyways.

Early rushes will cripple the AI greatly; in the first 50 turns it's in the expansion phase and usually go for the empire buildup techs rather than combat ones. If you want a real challenge, just let it build up it's first few colonies, then it will be more challenging.I know there are rush techniques, but i rather not use them. My first concern is getting my enjoyment from the game; killing off the AI when they are at their weakest isn't really fun to me.

Believe me, I usually do that. It was just an example of why I seem to build up that much faster than they (and you) seem to do. Of course I can play passively for another 50 turns and wait for the inevitable and tedious and immersion-breaking declarations of spite (because it sure isn't war they are bringing to me) that all those 4x games sadly have. If I do that, the game is even more boring because I will never have a battle against an enemy I do not outtech by 300%, then.

New version 1.36 up: Lots of tweaks and improvements that hopefully translate into better AI (and gameplay). Let me know if anything doesn't work as intended. Enjoy!

Just tried your newest version. The bug with the diplomatic traits still exists - I guess you made some error when editing them and killed some dependancies or something?



The AIs at least seem to build full fleets now. They were still not really dangerous, but at least at the very start (I was basically skirmishing with 2 AIs and some pirates from turn 5 on) I lost two or three ships and had to retreat once. While I like the fact that you downgraded the invasion power of normal weapons, I think you might have to lower the invasion module in the tech tree then - I had to siege for ~20 turns per colony and wait for the invasion module upgrade before I could realistically attack the remaining homeworld, which was a bit tedious.



AI did strange things again btw, the amoeba that invaded me (and in the end got crushed for that) had seemingly endless ressources to build fleet after fleet from just two remaining systems under siege and somehow managed to get those out of the sieged systems in spite of them being blockaded. At the same time, the homeworld was missing stealth colonies when I finally got it a few turns ago, which should have been the #1 priority for them since they had been sieged for over 30 turns.



I do not really like the dependancies for system improvements. So far the only thing they did was complicate my build queue because I had to stop putting stuff in there until the refinery was done and the add the rest instead of being able to set new colonies into fire and forget mode for the next 50 turns. Not really a problem, but slightly annoying and so far I did not see any real benefit of it. If it helps the AIs choose better build paths, I am all for it, but the opportunity cost makes it a hassle.



I played on hard this time and by the current turn (still in the double digits) the game is basically won again. I have ~10 systems (including the four I captured from that amoeba) and another 6 or so "free" within my area of influence to be colonized as soon as I get the next less expansion dissaproval research and/or colonial rights. No pirates can spawn since I have scoutships in all systems I do not live in yet. Of the two that attacked me, one is dead and the other about to be (of course the dead one threatened me as usual and demanded systems for a peace treaty five turns before losing his homeworld... sigh...). One sophon is my friend and minor trade partner, one of the others I just contacted by scouting a wormhole, the rest is nowhere to be seen but has nothing judging from their scores (especially the brown AI is basically dead, it has 2 systems and that's it, I guess one of the others is killing it off right now). My trade income from just four or five routes is again ridiculously high (200 dust and 100 science, roughly) and makes it so I can run at 15% taxes and still make hundreds of dust per turn. If I had not just explored all moons and removed all anomalies and built institutes, almost all my planets would sit on convert to science out of boredom.



I did not write this to brag - I am pretty much sure that many, many players are much better and more focussed in this game than me or know the mechanics better. It just shows that its basically impossible to not win within 100 turns or so unless multiple cheating AIs brutally gang up on the player, which is no fun either but just frustrating. I even weakened my racials by removing all negative attributes (which were basically free points because my heroes never die and my systems never get invaded even with the penalties) and exchanging the attack bonus with a defensive one for kicks - I even removed my 10% science bonus for something that was so unimportant that I actually forgot what it was. So it's definitely not some racial trait magic that makes me win every battle, it's the fact that the AI does not build their systems and fleets properly.

stormfox wrote:
Just tried your newest version. The bug with the diplomatic traits still exists - I guess you made some error when editing them and killed some dependancies or something?

Can you give me more details?

The AIs at least seem to build full fleets now. They were still not really dangerous, but at least at the very start (I was basically skirmishing with 2 AIs and some pirates from turn 5 on) I lost two or three ships and had to retreat once. While I like the fact that you downgraded the invasion power of normal weapons, I think you might have to lower the invasion module in the tech tree then - I had to siege for ~20 turns per colony and wait for the invasion module upgrade before I could realistically attack the remaining homeworld, which was a bit tedious.

The invasion parameters are hard to balance, maybe i was too quick to make that change.

AI did strange things again btw, the amoeba that invaded me (and in the end got crushed for that) had seemingly endless ressources to build fleet after fleet from just two remaining systems under siege and somehow managed to get those out of the sieged systems in spite of them being blockaded. At the same time, the homeworld was missing stealth colonies when I finally got it a few turns ago, which should have been the #1 priority for them since they had been sieged for over 30 turns.

If they retreat from battle, they can get out of system (unless you own/control everywhere they could go, then they just die). As for the sieging, that doesn't prevent them from building ships or researching (FF sieging rules are different, they get less fids but not zero at anything). Now, i don't know if the AI prioritize defense tech when blockaded; if it dooesn't, i will try to add the feature.

I do not really like the dependancies for system improvements. So far the only thing they did was complicate my build queue because I had to stop putting stuff in there until the refinery was done and the add the rest instead of being able to set new colonies into fire and forget mode for the next 50 turns. Not really a problem, but slightly annoying and so far I did not see any real benefit of it. If it helps the AIs choose better build paths, I am all for it, but the opportunity cost makes it a hassle.

Well, the point was to force the player to build up it's colonies first instead of just queuing (and buying off) the improvements.

I played on hard this time and by the current turn (still in the double digits) the game is basically won again. I have ~10 systems (including the four I captured from that amoeba) and another 6 or so "free" within my area of influence to be colonized as soon as I get the next less expansion dissaproval research and/or colonial rights. No pirates can spawn since I have scoutships in all systems I do not live in yet. Of the two that attacked me, one is dead and the other about to be (of course the dead one threatened me as usual and demanded systems for a peace treaty five turns before losing his homeworld... sigh...). One sophon is my friend and minor trade partner, one of the others I just contacted by scouting a wormhole, the rest is nowhere to be seen but has nothing judging from their scores (especially the brown AI is basically dead, it has 2 systems and that's it, I guess one of the others is killing it off right now). My trade income from just four or five routes is again ridiculously high (200 dust and 100 science, roughly) and makes it so I can run at 15% taxes and still make hundreds of dust per turn. If I had not just explored all moons and removed all anomalies and built institutes, almost all my planets would sit on convert to science out of boredom.

I'm wondering how you can get such trade, is it because of a Hero or the traits you have picked? It shouldn't be that high, and i double checked all the traits and none have any absurd values or typos.

I did not write this to brag - I am pretty much sure that many, many players are much better and more focussed in this game than me or know the mechanics better. It just shows that its basically impossible to not win within 100 turns or so unless multiple cheating AIs brutally gang up on the player, which is no fun either but just frustrating. I even weakened my racials by removing all negative attributes (which were basically free points because my heroes never die and my systems never get invaded even with the penalties) and exchanging the attack bonus with a defensive one for kicks - I even removed my 10% science bonus for something that was so unimportant that I actually forgot what it was. So it's definitely not some racial trait magic that makes me win every battle, it's the fact that the AI does not build their systems and fleets properly.

Well, if you can pinpoint where they flaws are i could probably improve it. When i play the AI seem fine, some values may be off when it's at war or something or it doesn't rush certain improvements when it should (i did a lot of work on it but some weights might be off). I guess i should war my AI to death and try to figure out what it's missing when under pressure.

Foraven wrote:
Can you give me more details?

The details were one post up - all diplomatic traits cost 0 besides eternal war, which cannot be picked at all.

The invasion parameters are hard to balance, maybe i was too quick to make that change.

I think just moving the invasion module down to one of the first two techs would solve the problem. How about at least swapping the invasion module and stealth colonies so that the name of the corresponding tech still somewhat fits? That way the invasion module would be available much sooner and the counter to invading only happens after one can effectively invade.

]If they retreat from battle, they can get out of system (unless you own/control everywhere they could go, then they just die). As for the sieging, that doesn't prevent them from building ships or researching (FF sieging rules are different, they get less fids but not zero at anything). Now, i don't know if the AI prioritize defense tech when blockaded; if it dooesn't, i will try to add the feature.

I know that they can build a fleet on the planet and then plop it out. I am talking about fleets 3-4 systems away while their only two systems were under siege. I am talking about them not building ground defense when their homeworld is sieged. I am talking about their one colony which has basically nothing in it and 5 or 6 population crapping out ships like no tomorrow. Sorry if I start sounding frustrated - it's because I am. I am soooo tired of 4x games creating artificial difficulty through tediousness and irrititating AI behaviour instead of just having them build stuff well. I would not even have a problem with a slight boost to their economy or something, but the way they are usually handled (and are handled in ES) is just something that makes me wish programmer hell is having to play against cheating annoying AIs for all eternity.

Well, the point was to force the player to build up it's colonies first instead of just queuing (and buying off) the improvements.

But it does not actually affect the colony buildup in any way. I still build the same things in the same order - now I just have to create my queue over 2-3 turns instead of being able to set it once and forget about it. It's effectively the same, just more tedious. If some super strong improvement should not be available from colony founding on and is still available, it is underpriced, because if it was expensive, I could not realistically buy it out. If I am rich enough to be able to buyout anything, I can still do that now, it just takes me two or three turns instead of one - no relevant difference in outcome. I do not think dependancies solve any problems, but bring with them a problem of tediousness and a problem of missing clarity (why can't I build x here but over at the other system?).



Lets go one step back. Which improvements do you think are problematic when bought out early and why?

I'm wondering how you can get such trade, is it because of a Hero or the traits you have picked? It shouldn't be that high, and i double checked all the traits and none have any absurd values or typos.

I have no traits that influence trade or money whatsoever. For some reason, the game chose my border colony for the two big trade routes instead of my home system with the hero in it, so no boosts through any improvements or heros either. Just baseline. The two major trade routes there give something like 65/45 each, and have been doing so from turn one of the peace treaty (at around turn 50). At this point in the game, that is some serious money. Btw, two other trade routes to some other systems are total crap with 4/3 or something, which is equally stupid since they could as well not be there. Why are there such huge discrepancies between trade route values? In the current situation I would expect all routes to be around 10-20 dust and science each, with the smaller systems giving perhaps half of what the major systems produce. That way, they would still be valuable but not make or break the economy. Oh, btw, I got 8 unused trade routes on my homesystem alone, twiddling thumbs.

Well, if you can pinpoint where they flaws are i could probably improve it. When i play the AI seem fine, some values may be off when it's at war or something or it doesn't rush certain improvements when it should (i did a lot of work on it but some weights might be off). I guess i should war my AI to death and try to figure out what it's missing when under pressure.

I would start with the way they pick fights. Hint: they always pick on goliaths when they are dave and left their sling at home. They never build a strong economic base in their colonies and never build defensive stuff when attacked. In the early game, they stay with kinetics for too long (remember how in the last game I said the only battles I lost ships in where those they had missiles while my ships had neither flak nor something else than kinetics themselves?) instead of at least mixing some missiles.



Tangent: Going missiles right at the start almost wins the game by itself because they decide battles during the long range segment, and since everyone is a glass cannon, most ships do not even arrive at medium or melee range to effectively counter with kinetics. Kinetics by themselves only make sense when you are tanky enough to actually arrive there - which with the current damage:mitigation values means only when you are superior technologically and/or in ship size.



Back to AI warfare: They send fleet after fleet to be crushed by the very same fleet again and again instead of recognizing they need to change something up. They go on your nerves by sneaking little inconsequential fleets all over your empire and sieging here and there without any real consequence than you having to build one anti-annoyance fleet and send it on a permanent cleanup circle running mission. It costs player ressources, but not really empire ressources, but they again lose ship after ship.



Fighting the AI is a bit like going up on a hill and standing there until their hordes ran dry. After that you just walz into their territory and take over all of their planets, one or two at a time, without any resistance anymore. The AI is so focussed on always getting on your nerves (because you can sure as hell not call that attacking) that it is completely spent by the time you actually have breathing room to attack yourself.

stormfox wrote:
The details were one post up - all diplomatic traits cost 0 besides eternal war, which cannot be picked at all.

Not a bug, i want them to be free (you can only pick one anyway). As for eternal war, it's tied to militarism, you need it first.

I think just moving the invasion module down to one of the first two techs would solve the problem. How about at least swapping the invasion module and stealth colonies so that the name of the corresponding tech still somewhat fits? That way the invasion module would be available much sooner and the counter to invading only happens after one can effectively invade.

Will look into that.

I know that they can build a fleet on the planet and then plop it out. I am talking about fleets 3-4 systems away while their only two systems were under siege. I am talking about them not building ground defense when their homeworld is sieged. I am talking about their one colony which has basically nothing in it and 5 or 6 population crapping out ships like no tomorrow. Sorry if I start sounding frustrated - it's because I am. I am soooo tired of 4x games creating artificial difficulty through tediousness and irrititating AI behaviour instead of just having them build stuff well. I would not even have a problem with a slight boost to their economy or something, but the way they are usually handled (and are handled in ES) is just something that makes me wish programmer hell is having to play against cheating annoying AIs for all eternity.

Are you sure they didn't have a colony elsewhere or simply another AI sending ships? The AI can't conjure fleets, it can only build them at a colony. As for building ships, there are traits that can make it cheaper and cost less upkeep. And last, you could jailbreak the savegame and see what they are really doing, maybe you'll learn how that AI pull it off.

But it does not actually affect the colony buildup in any way. I still build the same things in the same order - now I just have to create my queue over 2-3 turns instead of being able to set it once and forget about it.

The idea isn't to change the build order, it's to delay how soon you get them. In that regard it works.

It's effectively the same, just more tedious. If some super strong improvement should not be available from colony founding on and is still available, it is underpriced, because if it was expensive, I could not realistically buy it out. If I am rich enough to be able to buyout anything, I can still do that now, it just takes me two or three turns instead of one - no relevant difference in outcome. I do not think dependancies solve any problems, but bring with them a problem of tediousness and a problem of missing clarity (why can't I build x here but over at the other system?).

It's dependant on industry output; late game improvements can requires as much as 600 industry before you can consider buying them out. So i doubt it would just take you 2-3 extra turns to get them. Of course you can get around that by moving population with your colony ships, that would speed up the process.

Lets go one step back. Which improvements do you think are problematic when bought out early and why?

I want colony development to take more time; In larger/high tech game you can build up your colonies in a few turns and fill up the build queue before you actually have any population there. The idea behind my change is that you need the population to get the industry to be able to queue up the better improvements... Capturing enemy colonies becomes quicker then.

I have no traits that influence trade or money whatsoever. For some reason, the game chose my border colony for the two big trade routes instead of my home system with the hero in it, so no boosts through any improvements or heros either. Just baseline. The two major trade routes there give something like 65/45 each, and have been doing so from turn one of the peace treaty (at around turn 50). At this point in the game, that is some serious money. Btw, two other trade routes to some other systems are total crap with 4/3 or something, which is equally stupid since they could as well not be there. Why are there such huge discrepancies between trade route values? In the current situation I would expect all routes to be around 10-20 dust and science each, with the smaller systems giving perhaps half of what the major systems produce. That way, they would still be valuable but not make or break the economy. Oh, btw, I got 8 unused trade routes on my homesystem alone, twiddling thumbs.

It's a bug, though i don't know how to fix it. Does who it get the traderoute to matter?

I would start with the way they pick fights. Hint: they always pick on goliaths when they are dave and left their sling at home. They never build a strong economic base in their colonies and never build defensive stuff when attacked. In the early game, they stay with kinetics for too long (remember how in the last game I said the only battles I lost ships in where those they had missiles while my ships had neither flak nor something else than kinetics themselves?) instead of at least mixing some missiles.

They don't pick on goliath per see, they pick on who is the most threatening (who they hate the most) first (and they hate the most the one with the biggest score). The way it is set, most of the AI focus on improvements when not at war, then rush weapons stuff when it happen. The AI doesn't seem to change it's pace according to how large the map is or how many opponents are there (nor can i see any way to change that).

Tangent: Going missiles right at the start almost wins the game by itself because they decide battles during the long range segment, and since everyone is a glass cannon, most ships do not even arrive at medium or melee range to effectively counter with kinetics. Kinetics by themselves only make sense when you are tanky enough to actually arrive there - which with the current damage:mitigation values means only when you are superior technologically and/or in ship size.

How can you tell if they reach medium or close range when you actually play in automatic? In my games they do often reach close range, even when missiles are present. Giving more health to ships would just empower the tank approach; ie ships with lots of health and repair modules. The problem i see is you get ahead in tech way too easily; either i broke something in a previous version or you abuse some flaw in the gameplay i am not yet aware. Will have to look into this.

Back to AI warfare: They send fleet after fleet to be crushed by the very same fleet again and again instead of recognizing they need to change something up. They go on your nerves by sneaking little inconsequential fleets all over your empire and sieging here and there without any real consequence than you having to build one anti-annoyance fleet and send it on a permanent cleanup circle running mission. It costs player ressources, but not really empire ressources, but they again lose ship after ship.

Alright, that i probably know the cause; i will fix that part right now.

Foraven wrote:
Not a bug, i want them to be free (you can only pick one anyway). As for eternal war, it's tied to militarism, you need it first.

Okay, I will look into that.

Are you sure they didn't have a colony elsewhere or simply another AI sending ships? The AI can't conjure fleets, it can only build them at a colony. As for building ships, there are traits that can make it cheaper and cost less upkeep. And last, you could jailbreak the savegame and see what they are really doing, maybe you'll learn how that AI pull it off.

They were wiped out after I got those colonies, so no, they had no others. And no, it wasn't someone else. Amoeba do not have any bonuses in that regard and since their systems were blockaded, their many trade buffs did not work. Basically, it should have been impossible to build that many ships.

The idea isn't to change the build order, it's to delay how soon you get them. In that regard it works.....

It's dependant on industry output; late game improvements can requires as much as 600 industry before you can consider buying them out. So i doubt it would just take you 2-3 extra turns to get them. Of course you can get around that by moving population with your colony ships, that would speed up the process.

How should it take me longer? As far as I know buyout costs are simply a factor of the industry points costs of the improvement in question. Usually I buyout refineries, institute and perhaps supermarkets or something and the rest gets built by itself within 30 turns or so. Before I could just throw everything in there and forget about it, now I have to do it in 2-3 steps. That is the only difference.

I want colony development to take more time; In larger/high tech game you can build up your colonies in a few turns and fill up the build queue before you actually have any population there. The idea behind my change is that you need the population to get the industry to be able to queue up the better improvements... Capturing enemy colonies becomes quicker then.

How does this work? I have not seen anything to support this.

It's a bug, though i don't know how to fix it. Does who it get the traderoute to matter?

I doubt it, but I have not looked into it enough to make an educated comment on that.

They don't pick on goliath per see, they pick on who is the most threatening (who they hate the most) first (and they hate the most the one with the biggest score). The way it is set, most of the AI focus on improvements when not at war, then rush weapons stuff when it happen. The AI doesn't seem to change it's pace according to how large the map is or how many opponents are there (nor can i see any way to change that).

As I said, this isn't just a problem with ES or your mod, its a general problem with 4x games. It is especially aggravating when longterm allies suddenly start declaring war on you just because you are winning, forcing you to crush them. Basically this behaviour is killing any immersion, internal logic and is bad for those guys, too. Oh, and it transforms any game into a "kill them all" game.

How can you tell if they reach medium or close range when you actually play in automatic? In my games they do often reach close range, even when missiles are present. Giving more health to ships would just empower the tank approach; ie ships with lots of health and repair modules. The problem i see is you get ahead in tech way too easily; either i broke something in a previous version or you abuse some flaw in the gameplay i am not yet aware. Will have to look into this.

You can see the three combat phases during automatic - look at the fleet health bar while combat resolves.



I am not "abusing" something and you did not suddenly break something. It was the same way before I used your mod, likely even worse. I just do useful buildups and use just enough military development (research as well as fleet building) to get and hold a good territory for me, colonize just enough planets so my happiness does not completely tank and then build up those systems as fast as possible. As I said before, most of my systems are sitting on science conversion most of the time anyways. Its not hard to keep getting ahead that way.

Btw guys, new version 1.37 Up! A few minor fixes while i try to figure out how to improve the AI. Enjoy!

stormfox wrote:
They were wiped out after I got those colonies, so no, they had no others. And no, it wasn't someone else. Amoeba do not have any bonuses in that regard and since their systems were blockaded, their many trade buffs did not work. Basically, it should have been impossible to build that many ships.

Strange, it never happened to me. A savegame would have been helpful. If it happen again, please send me the save.

How should it take me longer? As far as I know buyout costs are simply a factor of the industry points costs of the improvement in question. Usually I buyout refineries, institute and perhaps supermarkets or something and the rest gets built by itself within 30 turns or so. Before I could just throw everything in there and forget about it, now I have to do it in 2-3 steps. That is the only difference.

When you swin in dust (something that usually happen before the endgame), you can buyout anything no matter how your industry is (it's just more expensive). Anyway, it's a feature i can easily remove if it's unpopular.

As I said, this isn't just a problem with ES or your mod, its a general problem with 4x games. It is especially aggravating when longterm allies suddenly start declaring war on you just because you are winning, forcing you to crush them. Basically this behaviour is killing any immersion, internal logic and is bad for those guys, too. Oh, and it transforms any game into a "kill them all" game.

I doubt you tried every 4x out there, some like Civilisation have rather good AI. It's a fact though that AI is one of the most overlooked feature in gaming, especially if the title can be played multiplayer (thus you can bypass the problem). Diplomacy in ES is quite borked and i doubt i have all the tools to fix it. I could improve the Alliance positive factor though; if set high enough it could offset the other factors and keep them happy with you. The devs have made some improvements there with the EXP, unfortunately it comes with a lot of issues.



Oh yeah, i did increase the odds the AI would declare war (by popular requests) because the AI was often too content to be at peace (even the military races). I could work on those again and see if i could strike a better balance...

You can see the three combat phases during automatic - look at the fleet health bar while combat resolves.

Alright, never noticed (i rarely use the feature).

I am not "abusing" something and you did not suddenly break something. It was the same way before I used your mod, likely even worse. I just do useful buildups and use just enough military development (research as well as fleet building) to get and hold a good territory for me, colonize just enough planets so my happiness does not completely tank and then build up those systems as fast as possible. As I said before, most of my systems are sitting on science conversion most of the time anyways. Its not hard to keep getting ahead that way.

There was a lot of those feature to abuse, like rushing new colonies used to be much more effective than being slow and steady. My tweak to expansion disapproval did make that strat obsolete and helped the AI greatly. There was also certain improvements that were just too good and crippling if not used early (like the second research improvement). It's quite possible there is still some of those quirks left around that make it too easy for experienced players but not used by the AI.