Also, P.S. quoting is still broken, I decided I would format the quotes in italics and add atrribution at the top for the hell of it.

Me too. If I don't find a better solution I'm going to cut away two thirds of the weapons-tree (so each race will only be able to either research beam, missiles or mass-driver). I used this concept in the past for the Minor-tree (which was radically cut-back in several places in order to compensate for their low RP and make them only research important stuff) and it might appear to be a radical change but if you think a little bit about it then it's not really a big deal.....

For example once a choice for one branch of the weapon tree has been made it is simply best to stay in that branch. I've never found it useful to research multiple weapons but I sometimes find the AI researching lasers although he already has ships with harpoons....

On a threefold weaponstree you can only circumvent that by steadily increasing the AI values for each following tech but this can bear the negative danger that the AI will transfix itself totally to weapons-research and never get back to social stuff (esp. considering that the low-tier weapons do need a reasonable high AI value, getting them early is of grave importance!)

Having only one branch available can fix that, the starting AI values could be very high and subsequently fall down from tech to tech, which would mean that, after some time, these values drop so low that the AI will re-focus on other stuff, ie. upgrading industry, terraforming....

Basically the AI values of the whole tree have to be this way (always slightly droping from tier to tier) so the AI will change oftenly thereby getting alot of easy low-tier starting technologies and developing his game equally on all fronts without getting stuck somewhere...

Additionally, if we would have only one weaponsbranch all those current worries about statoverkill (miniaturization + logistics etc) could easily be solved by integrating them into the weaponstree. Just reduce the standard miniaturizationsbranch to 3 techs but add the boni of the deleted techs to some high tier weapons, and this could be done to all sorts of "dangerous" techs (eg starbase defenses)- ie. making the weaponstree (with its huge techcostinflation) a pre-requisite.

Doing likewise with defenses (for fun) could add a totally new tactical element to the gameplay - ie. some races will be strong defensively versus other races because their defenses do always match the opponents weapons (although the opposite scenario could also happen). In these cases a human player could either focus on all-attack, or simply shy away from this particular threat and rather try to expand versus weaker enemies....

I'm afraid there's something missing because finally the hull will make a warship who will do combat in fleets. As such, there's the additional factor of firepower-elimination. To explain, if you send a fleet of huge warships vs medium warships (with everything maxed out) then the huge will have a (starting) advantage of factor 2.5. As the fight progresses the medium ships will have to face losses earlier as the huge ships, then it is very realistic to assume that in following combat rounds -n mediums have to go versus -0 huge ships. With every loss of medium hull the advantage factor should increase, although, it could also decrease if the huge hulls do also loose a ship etc...

Now I have absolutely no clue how to incorporate that    esp. because that using average rolls isn't quite what is happening within the game. We should have something like a combat simulator for this!

But nevertheless, these ratios can give a overview and hint towards possible unbalancedness - for example you'll see how strong medium is (the intro screenie is well justified. Capital Ships. ...) and I'm surprised by how weak small is vs tiny. I think if you throw in a few starbase assist modules tiny is going to blow small out of the water easily.

How much is that hardcoded limit on trade routes actually?

On another thing, if I remember correctly the Cathedral of Valour used to a small racial soldiering bonus, isn't it? Do you know why this didn't work, I mean racial stats comeing from spamable improvements... Because if it did that would open up very interesting possibilies

The module is a remnant of GalCiv1. Actually, a lot of data, pics, descriptions & some stats are, unused but not deleted. Such an over-repair also takes place if you take damaged ships back to planets, or with enough repair-ability. But it's not really a big deal. Esp. not with starbase modules who only work in a small areal. You could see it as a "hitpoint booster" or something alike.

Set the AI value to zero on high class farms, lower the food, make them a 1pp, its not as bad as gaunathor made it seem. Even just make the techs give the bonus instead!

@Maiden666

re: "Basically the AI "knows" the map inside-out right from the start, all colonizable planets, resources and he further doesn't have a fog-of-war"

While that is true, the AI can't colonise planets it hasn't "seen". I've tested this extensively, and it is the case. For example, in one game there was a habitable planet in a star system right next to the AIs. However, it didn't colonise it, despite having multiple colony ships. That only changed when one of the AIs scouts got close to the planet. In another game, there was a uncolonised cluster. The AI still had several colony ships left, and could easily reach the cluster without having to extend its range. However, it never colonised the cluster, because its scouts were busy in other directions. I finally took the cluster for myself, almost three years after the start of the game.

You can easily verify this for yourself. Just start a game, disable the FOW, let the AI play for itself, and watch what it does. It's always the same procedure:
-use the starting colony ship to colonise a planet within sight (usually the secondary world of the homesystem)
-build scouts, and send them out to find planets
-build colony ships, and send them to the scouted planets

I never saw any deviation from this.

re: "With spamable farms you either decrease their AI values then most worlds will be without farms and stay at 8b ppl(this is the case in my testgames the last 2 weeks in 80% of all worlds - playing without techtrade)"

The AI isn't yet researching the farming techs properly. Except for the Yor, but even they still need some work. 

re: "Or you increase their value then the AI will spam them which will do more harm then good."

The AI will never build more than four farms. With a low enough food bonus, this won't be an issue. I specifically set the farms up so that this will lead to a max population of 20b. Now the AI only needs to build enough morale improvements to keep the populace happy.

re: "However, on 1pp improvements you can increase the value to levels which ensures the AI will always build them."

You can never be sure about that, no matter how high you set the value. Even with a value of 500+, there will always be cases where the AI just doesn't build the improvement.

re: "Minor races are cut down by 50% in population!!!     If anything you should boost them"

This had to be done, because it  is impossible to balance for two different sets of base food production. No matter whether you go for 1pp farms or spammable ones, if the AI builds them on the homeworld, they'll have an over-population problem. If you balance the farms for the food value of the homeworld, then you'll screw over the colonies, because the farms will barely do anything. You'll also hurt the Yor, because their economy relies on high population.

re: "How much is that hardcoded limit on trade routes actually?"

12. The Korx already start at 10 (11 originally).

re: "The module is a remnant of GalCiv1. Actually, a lot of data, pics, descriptions & some stats are, unused but not deleted."

I know. It's just weird that the manual still mentions the module, even though it has never been in GalCiv 2 (except for maybe the beta for DL).

@Mabus

re: "A stat that works civ wide doesn't work on normal (1pp included) structures. It has to be an SP or GA."

Or a TG.

re: "The minors could could get a large soldering bonus to make them harder to conquer."

I was thinking of adding a planetary defense bonus to the Civ Capital. That would help everyone, not just the minors. It also makes sense, in my opinion. You'd want your homeworld to be as secure as possible.

Then this planet will be one which has a higher population, right now it is no different. The chances that the AI will build 4 farms onto a single planet are very low, I've probably seen this with Yor & Iconian but never with other races + ordinary farms. And to get to the standard 20b max they also need to complete the research into farms, and then still be in the process of upgrading planets. (some AI do sooner or later turn to 100% military global slider setting and totally forget about upgrades or empty tiles or research.....)

This all comes down that in a bigger game where an AI has +100 planets you'll find 1 or 2 such planets while all others stay at 8b ppl or maybe 10 etc.... and because of these 1 or 2 high populated planets the AI will generally adjust tax down for all planets - and isn't this precisely the problem that should have been fixed?

Granted, placing a +7b farm on a 400% bonustile will make the scenario far more worse because of the moral inflation beyond ++18b, but still the problem persist of having potential uneven max pop. Besides, a planet with 20b ppl also needs quite some boosts through moral improvements or resources otherwise it will be stuck into red moral long before 20b are reached....

I desperately need to learn more about this ominous bug. Maybe this was the thing that was responsible for my inconclusive results with tinkering with the Initial Colony building.  

Hmm, how about making farms indestructible then?

Another method could be to not use any upgrades at all.

Initial Colony +10b, Recruiting Center +3b 1pp, Counter Esp. +4b. Planets will mostly be +13b or +17b midgame if the AI researches into it. Give these improvements base +1 food and add in the rest with a foodprod modifiers. Bonustiles could be used but a overkill would only happen in very rare event. A single 200% bonustile will result in 1 +14b or +18b planet, only a used +400% tile used and with both improvements researched+placed+mid-gamesetting with pop already maxed out+no transports build would mean a negative and unusual high population. That's alot of factors.

Of course it would require a major overhaul of alot of techs/improvements because common farming would have to go, now I'm not proposing to do this here, all I'm saying is that it should be possible to let the AI have way higher populated planets (which does just add another difficulty level to the game).

I know, but question was *why* it doesn't work.... and if there's maybe anything that could be done to help this or make it running by some sort of out-of-the-box editing   . The possibilites for balancing, putting a racial stat into relation of the planet-count would be fantastic!

I just spend 2 hours with this and saw a deviation in 95% of all test. Always from turn 1-3. I assume my setup differs from yours:

-I took Thalans so the AI has to send his initial Colony Ship out of the system

-Took away all sensors from the starting ships

-Gave them negative sensors, so that in any case, they only had to +5 sensor range of their starting planet

-Gave them +100 range, on a tiny map they practically can go everywhere

-Started a game with them in order to check their sensor range

-Map tiny, suicidal diff, all rare, loose cluster

-Control N until no colonizable planets are near their system

-Used the cheats to clear the FOG. Sometimes used cheats to give me all their ships not including their colony ships.

If a habitable planet in the range of +10 to +15 parsecs was available, then the colony ship already does have a course set for it. Although the planet is quite some parsecs out of sensor range. Right in turn 1. This happened in about 30% of all cases.

In the remaining cases where habitable planets were even farer away then the colony ship will be taken into the planet in turn 2, but be released right in turn 3 with its destination set to a habitable planet totally out of sensorrange.

Mostly the AI will rushbuy another colony ship in turn 1 and this does do the cheating likewise. Occasionally he will rushbuy a scout, which will usually go somewhere else, even unhabitable planets.

Maybe it has something to do with AI personality, or probably the use of cheat codes does reveal the map equally. I usually don't use them. All I can say is that a few years ago, where I favoured a all-labs strategy with lots of techbrokering etc I made it one of my top priorities to get all enemy scouts ASAP to turn into Surveyors and also explore the map with them... if scouts would be a pre-requisite for colonisation then these AI wouldn't have been able to much of that because they seldomly had scoutships....

@Gaunathor.

No looks pretty good imo.

Is it worth considering turning the Precursor Library into an SP? I know it's not a structure you start with but the iconians practically trip over it. It's a good structure but perhaps it's not game braking if there's a few more of them in a game.

I ran a little experiment linking hull tech to industrial. The idea was you'd unlock factories then a bigger hull. Of course you could omit medium hull building and go straight for large by researching the next set of factories.

Suffice to say it wasn't a success.

*sigh* Hulls have been giving me a headache ever since apt1.0. I guess we're stuck with how it is.

I've given the AI a negative sensor-range so that in the case he gets something like the hidden-research multiplier etc he would still end up with a negative rating.

Then I thought maybe the increase is a flat +n put on top, and subsequently increased his penalty ffrom -99 even to -9999. No effect. Got espionage on him to check if the gamre took it in, confirmed.m Crazy that they would give him a +several thousand increase when there aren't even maps available in that range, and then there's a sensor-cap also...

But probably the AI's game isn't using the sensor-stat at all, so I changed the setting to immense galaxy and gave him additional range. Still the same.

Changed difficulty to cakewalk, then normal, then tough. Still the same.

This is a screenshot of the very first game on normal immense galaxy. Note the date is 15th Jan, it's the 3rd turn. The ship benath Thala is not a scout but his survey-vessel who lacks sensors+survey mod (he doesn't use this ship). On turn 1 the colony ship was parked back to Thala, but destined to a faraway planet in turn 2. On turn 3 you see it already moved.

The distance this ship has to travel is 55 parsecs!!! There are absolutely no scouts on the map nowehere. Or other ships in that region or somewhere else, except my own homeplanet which is even more far away. Because this system lies somewhat diagonally away from Thala, his sensor-range increase should have been somewhere +75 pcs (!!!) for him to be able to see this system. On normal difficulty!

But maybe there's still something wrong with my settings...

On the TechVic-boni:

They're fine and fitting except I find the Korath one rather weak. They don't need soldiering offensively at all. but defensively it will help them with their planets they got from colonization. The problem are there spored worlds. Population growth that starts at 0.01b ppl is so increadibly slow these worlds will still be very easy to inavde no matter how much soldiering is.... Then, if the Korath aren't at war with everybody all these planets will succumb to foreign influence. I've played games where this was the case every turn. Overpowered buildings like the Dark Influence are probably ment to mitigate this but without a reasonable amount of pop nothing can come out of a planet anyway, because everything is based on pop initally.

The same is with their economy, spored planets stay in red bcs forever, which means that the AI is going to sell them. Negative diplomacy rating further aggravates that matter. I consistently buy all their planets. Having Korath ingame and keeping a good diplomacy rating with them is an "easy win ticket".

Then, sometimes they don't research into Advanced Tocix Atmosphere Reductor, so only half production comes out of these planets, but the maintenance stays the same. That the use of spore doesn't wipe planets clean - instead all buildings stay intact - also aggravates that matter 

Now these are a lot of problems and I'm not suggesting in trying to balance them via the TechVictory, but nevertheless, that soldiering bonus won't do much if they have lots of spored worlds. Popgrowth & loyality would help them, but those don't fit thematically considering they live a rather destructive lifestyle and even go against their cousins... they're anything but loyal...

@Mabus

re: "Is it worth considering turning the Precursor Library into an SP?"

The Precursor Library bonus tile is once-per-map, so I'd like to keep the actual building the same. Still, if we remove the 10% civ-wide bonus, I might consider turning it into a SP. But then, this would also remove some of the uniqueness of the PL. Now it really would be just a Tech Capital with a higher base value.

@DARCA

re: "so maybe we can end up teaching the AI to build less farms"

We can already do that by reducing the AIValue of the farms. Getting the AI to build them in a reasonable amount is the problem, because three AIPs, with different attitudes towards farms (and morale improvements), use the same building. Even within the AIP itself is a difference, depending on which race they were made for. For example, the Altarian and the generic version of AIP 11 like to build farms, while the Arcean version doesn't (it's closer to AIP 8 in that regard).

@Maiden666

re: "But maybe there's still something wrong with my settings...  "

I don't know what else to say. If I set the difficulty to Suicidal, I see the same thing as you do, but not on the lower settings.

re: "On another note, now you have to have 4 tierIII farms to go to 20b, tradtionally it has been 2 tierII farms. And usually 1 or 2 moral improvements, which do make now 5-6 improvements for pop alone. Sure, the AI won't be able to play out this concept on all worlds, and when he does not, something will go astray."

The point isn't to make the AI go for 20b on all planets (except for the Yor maybe), but to build a reasonable amount of farms (one or two, three max) along with enough morale improvements to keep the population happy.

re: "That is why 1pp improvments are actually a very good choice in trying to help the AI"

With 1pps you want the AI to build them on all planets. However, that doesn't work. Even with very high values, the AI will still not build them everywhere. I've tried it before, with no success.

re: "wait wait 6% moral on such lowpop would automatically mean 0% moral on high pop worlds that's suicide. you've written you made moral improvements to 200 value, did he spam VLC's on big worlds and now adjust slieders accordingly?"

More or less. You can see the difference between Altaria and Wisp: three Temples (plus the Shrine of the Mithrilar) on Altaria, no Temples (or any other morale improvement) on Wisp. Again, that's with a value of 200. You'd expect at least one Temple on Wisp. However, that's another problem with the AI: it rarely builds morale improvements on (terraformed) low-PQ planets. It usually turns them into manufacturing worlds, as can be seen in the screenshot.

I'm still adjusting the value for farms and morale improvements, but, so far, it hasn't gotten any better.

@Mabus

re: "Is there an ETA on the next version?"

I'll put it up this evening (in 6-8 hours). I just want to make sure everything is working properly first. Once it is up, I have no problems with you doing some tweaking too. Maybe one of us can find some good values.

In the meantime, I've been looking for Frogboy's post regarding AI colonisation (or rather "cheating") again. I think, I've found it. In addition, I also found this and this.

@Mabus

Changing some of the categories is always necessary. I've no issue with that.

As for that particular line of research, I've seen AIP 11 do it, but not often. It seems to depend somewhat on what races there are, where they are, and pure chance.  But then, I haven't yet adjusted the AIValue for those techs. So that would be the first thing to look into, in my opinion.