RBTS9 - Final Frontier

[timmy827]

Checked the save, Hector has already become yellow-Kanji's vassal. And Orange is unwilling to capitulate (insert standard rant about vassal mysteries here ) - maybe this ends if we bribe the other Kanji to peace?

One last note - if we do vassal Lu after this world (looks like it as we have one vote for and none against), we might get more assigned citizens on some of his captured worlds by letting them starve. IE taking a world down from 10 to 8 pop will get rid of 2 "its too crowded" and probably 1 "motherland", so we can actually assign one more. Too bad we aren't Spiritual and whip some sports arenas out with them.
OTOH, WW is causing 2 in the larger cities, so it will get much better next turn assuming we finish the wars.

EDIT: Yep invasion ships are cheap. In the absence of suicide collateral, losses are expected. If the AI just spammed PDS's it would take a lot of sacrifice, but fortunately they load up on Destroyers and Battleships which just get trashed by invasion ships...

God-Emperor, thanks for the info on war declarations. But after what we just did to the largest AI it's clear we don't need to be worried whatsoever. I would be shocked if the AI actually attacked with carriers/fighters, which means that our bombers alone guarantee safety.

 

[God-Emperor]

The AI will never attack using carriers with squadrons. It never puts squadrons on carriers. It almost never builds carriers, maybe never - apparently as far as the AI is concerned it is just an incredibly expensive and weak ship that has no special features. If you give them some via worldbuilder, or force the AI to build one in CvAI.py, it won't load squadrons onto it. Likewise, it never puts missiles on cruisers. Or squadrons on starbases, and the only reason there are ever missiles on starbases is because they are automatically generated there (and it may actually move some or all of them off the starbases to its star systems - but not onto cruisers).

The AI simply doesn't understand that it can do these things. It is because the map is, technically, all land. There is no land AI type that the AI can use for a unit that carries other units, whether as a transport, carrier of air units, or missile carrier. The sea based AIs don't work so good on land (like UNITAI_CARRIER_SEA, used by the carriers even though they are DOMAIN_LAND units in FF).

In theory, switching the carrier's domain to DOMAIN_SEA and changing the bCanMoveAllTerrain setting from 0 to 1 might help a little. The Dune Wars AI has been tweaked (in the DLL) in some way that I don't know to give the AI the ability to use transports on land when set up like this, and it works (mostly), so a similar fix might work for the carriers and cruisers here. But it probably requires DLL changes to make it actually work. But without the DLL changes, it might work a little with these XML changes. I've been meaning to check what happens, but haven't yet.

 

[Olodune]

But it probably requires DLL changes to make it actually work. But without the DLL changes, it might work a little with these XML changes. I've been meaning to check what happens, but haven't yet.

I know T-Hawk has successfully compiled the DLLs for BtS, so I would bet he can do it here too I'm not sure how much effort he is interested in putting towards this though (just solving the opening Zombie mode is a major plus! ).

 

[T-hawk]

I know T-Hawk has successfully compiled the DLLs for BtS, so I would bet he can do it here too I'm not sure how much effort he is interested in putting towards this though (just solving the opening Zombie mode is a major plus! ).

Yeah, not so much interested. Solving zombie mode with 2 lines of code and putting the rest of the game back in balance (can keep wreckage in, don't need drastic Emperor level costs for a competitive game) was well worthwhile. But I'm not convinced it's a good idea to make AIs use carriers at all. As Timmy observed, the Squadron Defense building pretty well stops squadrons. Their role is to chip down big ships in open space, which the human won't leave exposed.

However, the AI can use squadrons themselves fine. Last night I had one AI declare war and KILL A STARBASE using bombers! Too bad its timing was too late - that happened 6 turns before an influence victory as the Forge.

Thanks for all the war info, God-Emperor.

EDIT: Yep invasion ships are cheap. In the absence of suicide collateral, losses are expected.

Suicide units are also expected losses.

If the AI just spammed PDS's it would take a lot of sacrifice, but fortunately they load up on Destroyers and Battleships which just get trashed by invasion ships...

Yeah, I was thinking about that. Invasion ships are cheap, but PDSes of the same generation are a third cheaper plus get the fortify bonus. So shouldn't attacking with invasion ships always be a losing proposition? The answer is costs, especially when in Pacifism. A PDS must be maintained forever, but an invasion ship costs money only during its short lifetime from production to smashery. If the AI spammed PDSes, it would resist invasion but kill itself on unit costs. I guess the right answer is to concentrate on PDSes when a credible threat is detected, but that's easier said than done.

 

[Olodune]

Yeah, I was thinking about that. Invasion ships are cheap, but PDSes of the same generation are a third cheaper plus get the fortify bonus. So shouldn't attacking with invasion ships always be a losing proposition? The answer is costs, especially when in Pacifism. A PDS must be maintained forever, but an invasion ship costs money only during its short lifetime from production to smashery. If the AI spammed PDSes, it would resist invasion but kill itself on unit costs. I guess the right answer is to concentrate on PDSes when a credible threat is detected, but that's easier said than done.

All true -- but another major factor is the usual one: An entire Empire's productive capacity can be directed at building an invasion fleet to attack a single garrison which is typically defended by only a fraction of the total "defensive" fleet. The Civ combat mechanics usually mean that if you have at least twice the number of attackers as defenders (in near parity battles), the attackers come out with a large hammer win/loss ratio. And a conquered system. Human MP players turn this on it's head by keeping most of their defensive forces at the point of attack.

 

[Olodune]

After sacking the Avower's system that our big stack was next to I was pretty much ready to get Lu to capitulate, and then I got a message that sealed it:

Hey! Something to keep our military busy even if I get peace with the Avowers Capturing the core Forge worlds should be logistically easy since they are so close to our jump lane network.

Then something very odd happened:

All our starbases lost their "sphere of influence" for a turn! If you thought motherland unhappiness was bad, try stacking -6 health/happy on top of it .... I never figured out exactly what went wrong, but next turn our influence popped back into place.

Another weird feature:

The "build here" planet is often assigned to a planet we don't have influence on (yet) in captured star systems. As far as I can tell these actually get built too, but if you switch the build target there is no way to switch back.

A tech trade with our cooperative vassal:

Lu is now researching Intuitive Computers fast enough to be useful to us.

Two of Forge's worlds have fallen and a division of our fleet is poised to take their homeworld next turn. If they will vassal afterwords I would take it since their remaining systems are far away.

Some notes for T-Hawk:

The time for Utopia may be near an end (we are only at ~24% science even with several starports finishing in my turnset). Many of our systems are running out of good planets and/or are at their happy cap.

There is a barbarian battleship next to a vulnerable invasion ship -- I'm pretty sure only one of them will survive the IT Sorry.

The Forge has been sending a couple BBs towards RC-1. I've been waiting to hit them with bombers until they are within kill range of our own BB.

I have scattered several Omega PDSs and Cruisers/DDs along our jump lanes to give more reaction options to encroaching barbs. It seems to work fairly well.

 

[Olodune]

Yellow Halis finished off Orange Halis too Which probably means they will attack someone new soon.

 

[Mojoqmeyvam]

My guess for the SB bug is that vassalization triggered it.

I'm not surprised by the "build here" bug. Fairly minor in the grand scheme of things, but useful to be aware of.

We've run Utopia for, what, 15 turns? It seems odd that it's worth a switch in and out in such a short period of time, but then again it essentially doubled our population growth rate .

Should we send Yellow Halis a thank-you note for their generous act of Fratricide?

 

[T-hawk]

Good work. I see the squadrons at RC-1 and the vulnerable Forge battleships. Remember that bombers can kill, I think two rounds from all three squadrons will kill it before it can get away, don't even need our own battleship. I agree on Forge vassalization, we could use the happy. If they won't capit after the homeworld, go after The Torch in the southeast. Might be worth cash rushing a couple invasion ships near there to speed that up.

Starbase influence disappearing is a known bug (see the FF thread in the official mods forum). It's a side-effect of the code that recalculates borders upon vassalization, in the regular game where the master always gets priority over contested tiles.

Lu says 13 turns to Intuitive Computers. No way is he actually researching that fast on two dinky systems, he must have started it long before the war.

I'm not going to play tonight - got way too absorbed in a solo game of FF. Red Syndicate diplo victory on turn 190-some, by beelining Green/Utopia, getting two votes from religious buddies, and conquered one small neighbor with Ω invasion ships to go over the top. I wanted to finish that for some experience with Utopia civic, which is relevant to us here.

Utopia does tend to crash into the limitations of available planet space, but we do want to keep it a while longer here. We have significant growth headroom still at Paradise (3), Horus Taurus (3), Xanadoo (6), Elysion (7), Valhalla (2), Nirvana (10 after 500 influence), Duat (10 after 500 influence). Happy headroom is coming at Ascendancy Theory. Also most systems can still build a hab facility; 80 hammers is a good deal to create 2 hammers 9 commerce on an improved gems planet in an otherwise full system.

So I don't think we'll switch out of Utopia just yet this turnset. Switching will be worthwhile eventually; once we are no longer growing, ditching the -15% penalties will make up for the anarchy within 7 turns. We'll definitely keep Green Economy, +1 planet size makes for by far the most powerful civic in the column. Back to Pacifism would be a net gain eventually (cost about 90 gold after inflation, gain 75 hammers now and more later) but probably not worth anarchy.

I might be tempted to 3-for-2 revolt to Pacifism/Technocracy/Mechanized and use the culture slider. 20% culture makes all the Sports Arenas give an extra happy, and Technocracy makes up for the loss to the slider. And the culture would help all the conquered worlds expand influence in a useful timeframe. This might be way overthinking it though; any opinions?

Anyway, I will play late Wednesday (US Eastern time) but not before then. Sooooo, you missed the battle action so far, do you want a piece? Feel free to swap with me and take it now, just post a got-it.

 

[Olodune]

I might be tempted to 3-for-2 revolt to Pacifism/Technocracy/Mechanized and use the culture slider. 20% culture makes all the Sports Arenas give an extra happy, and Technocracy makes up for the loss to the slider. And the culture would help all the conquered worlds expand influence in a useful timeframe. This might be way overthinking it though; any opinions?

I was thinking along these lines as well. Don't forget that we also have access to the religion temple (which is nearly identical to a sports arena). Also two arena (2 x 1 happy at 20%) apparently allow for one extra happy as 10%. So by building 2 arenas and 2 temples we could have 3 additional happy per 10% culture slider ...

... of course our entire productive economy is less than 30% of the slider so this plan may still need a little fine-tuning Not to mention that building those improvements themselves give +4 happy -- plenty in all but the largest conquered worlds.

 

[God-Emperor]

Starbase influence disappearing is a known bug (see the FF thread in the official mods forum). It's a side-effect of the code that recalculates borders upon vassalization, in the regular game where the master always gets priority over contested tiles.

This is a minor inconvenience for a human player - you just need to recheck all of your system's population assignments when the base culture comes back since you lost all the resources that were in the starbase cultural area and may have had systems deassign population due to the temporary loss of happy resources pushing down the happy cap. This can result in population being assigned to different planets than you'd prefer.

For the AI it is more problematic. As noted with the startup bug, the AI only does population assignments in some specific circumstances. If it gets a vassal, a system that has pop deassigned due to the temporary happy cap reduction will not reassign them until one of the rare occasions when it does a full population reassignment (when a system is nuked, captured, founded, or has a border pop all population is reassigned; also, when its population increases it assigns one population to a planet).

The game checks for excess population assigned to planets and removes it every turn, via doCityHappinessPopLimit run in doBeginTurnAI every turn for both AIs and humans.

Strangely, a human player's population is checked for excess (unassigned) population that can be assigned on every turn in the doBeginTurnAI function. This is run for every player, AI or not, but there is an "if" statment that does nothing if the player is an AI but runs updateHumanCityTurn for humans. In updateHumanCityTurn it checks for unassigned population and assigns it. This is run in the onCultureExpansion callback for all players, human or not, which is why that first border pop ended the initial no population assigned problem for the AIs. I see no reason why this function shouldn't be run every turn for the AI as well as the human. It doesn't fully solve the problem as updateHumanCityTurn does nothing if the system has the "avoid growth" flag set, and the AI does actually set this flag sometimes. This does lead to it not assigning all of the population sometimes, which is silly - all possible population should always be assigned or you loose production and commerce for no reason.

So a fix for a variety problems is to make the updateHumanCityTurn function run for everybody in doBeginTurnAI and remove the "if (not pCity.AI_avoidGrowth()):" limitation from that function.

 

[Olodune]

This is a minor inconvenience for a human player - you just need to recheck all of your system's population assignments when the base culture comes back since you lost all the resources that were in the starbase cultural area and may have had systems deassign population due to the temporary loss of happy resources pushing down the happy cap. This can result in population being assigned to different planets than you'd prefer.

This seems to be true any time the unhappy situation decreases -- newly available citizens are unassigned by default. In a SG its not much of a problem (I will check every planet at least twice in 10 turns and often much more), but is more annoying in a warring SP game with Halis (due to the many civics changes). If the AI doesn't know how to deal with this ... well I'm surprised honestly. As you mentioned it doesn't seem to be terribly difficult to fix.

Edit: Actually you are saying that for the human newly available citizens are assigned -- that was not my experience, so maybe there is a 1t lag? Does the global happiness situation get updated after this call to assign idle citizens?

 

[timmy827]

Nice job. Quick notes:

Keep an eye on Industrialism trade opportunities of course - not worth researching but if we can swing a deal we should.

I don't think Technocracy is possibly worth it if we need to raise the culture slider to 20%. Remember Tech's +15 adds with our existing multipliers (just buildings in this mod), probably something like 60% through the empire (most developed worlds have 2 Uni's and 1 lab). So Tech's actual ratio bonus is 175/160 = ~9.3 %
Even in the wildly optimistic case that we run 100% research all the time, that's a net loss. And if we can sustain 50% slider normally post-Utopia its a huge loss.

Also, turn on science and burn through our cash, stay near zero cash for the next 3-4 turns as we have a bunch of cash multipliers about to finish.

EDIT: Since we will be switching out of Utopia sometime most likely, take a close look at the captured worlds with severe cap issues; some Nutrition Facilities may need building to support them after Utopia.

 

[God-Emperor]

Edit: Actually you are saying that for the human newly available citizens are assigned -- that was not my experience, so maybe there is a 1t lag? Does the global happiness situation get updated after this call to assign idle citizens?

Well, it is supposed to. I don't know how it happens, but I have also seen unassigned population in my star systems from time to time. It should only be possible for it to last 1 turn except in systems that have the "avoid growth" function active or unhappy population (which is always unassigned). But it does seem like there may be some loophole that can let it happen that I am not spotting in the python.

 

[Olodune]

Also, turn on science and burn through our cash, stay near zero cash for the next 3-4 turns as we have a bunch of cash multipliers about to finish.

Yep, good catch. Since I was building mostly gold multipliers I should not have left any excess cash in the treasury. (Well besides maybe ~120 to upgrade a PDS to an Omega in a pinch).

 

[sooooo]

Anyway, I will play late Wednesday (US Eastern time) but not before then. Sooooo, you missed the battle action so far, do you want a piece? Feel free to swap with me and take it now, just post a got-it.

Can't play until then either so best that we keep the turn order.

But I would like a little battling - try to set me up to bash another civ

 

[Mojoqmeyvam]

With Thanksgiving, I won't be able to play Thursday-Saturday. Maybe Sunday evening, not sure. Skip, pause, whatever's fine. I've had my war fix, so ya'll play nice now. Just not too nice.

 

[timmy827]

With Thanksgiving, I won't be able to play Thursday-Saturday. Maybe Sunday evening, not sure. Skip, pause, whatever's fine. I've had my war fix, so ya'll play nice now. Just not too nice.

I'm in pretty much the same boat, unavailable from Wed. until late on Sunday, though it looks unlikely for us to continue the breakneck pace and roll around back to me. I'm fine with having a lull over the holidays, but on the other hand we probably don't need a lot of discussion either.

 

[T-hawk]

Do they have Thanksgiving in deep space? Maybe New Earth does, but what are the rest of us slackers thinking?

Pausing over Thanksgiving is fine. I'll still play tonight but no demands on the pace after that. Timmy, I do expect it'll get back around to you once more, we won't finish the tech tree in 40 turns from here.

sooooo, unfortunately there's not much likely left to attack after Forge. I could pass on the capitulation (if they'll offer) and let you attack the next world (The Torch in the southeast.) Or you could go after the Reds, though we might end up with a domination victory that way.

Re civics, did we decide what to do? I buy Timmy's analysis against Technocracy, and then Pacifism isn't worth a turn of anarchy by itself. So we want to switch out of Utopia in ~10 turns - to what, back to Mechanized?

 

[Olodune]

Re civics, did we decide what to do? I buy Timmy's analysis against Technocracy, and then Pacifism isn't worth a turn of anarchy by itself. So we want to switch out of Utopia in ~10 turns - to what, back to Mechanized?

I think a swap to Mech. is worth it in about 10-15 turns. I was also thinking that after Forge capitulates we may want to switch most of our systems to "Wealth" -- at least to get an idea of how many more turns of research are required to get the Gate techs (there is a free tech in the tree). Once we have an estimate of tech turns remaining the net present value of build options can be compared easily. Hopefully this will prevent us from "overbuilding" (something I tend to do in builder games).

I'll have plenty of Civ access this weekend, so if this game stalls (which is fine ) I/we might have to play a "Thanksgiving OCC" ...