RBTS9 - Final Frontier

[T-hawk]

Might want to attend to some MM. Paradise has one pop on blue 0-2-6, should be on 1-2-9 gems instead. Horus Taurus has one on 1-1-7 that should work 2-3-6.

Colony ship: Mu Artemis (southwest corner) and Poseidon Beta (far northwest) are the two reachable unclaimed systems. Unfortunately the colony ship in Eden is about diametrically opposite that (what kind of sense does a toroidal universe make anyway?) but the colship won't be any other use so may as well try.

Aggressive starbase: The prospects aren't so good as I'd been thinking. There's no spot to get one starbase into even Δ bomber range of even 3 systems. It could be between RC-5 and RC-2 quickly, or between RC-2 and RC-1 three turns later. Correct spot for the former is either 2W from the oxygen/silver (asteroid defense but a reduced vision field) or 2S 2W from it (in open space with a wide shooting area.) Not 1S 1W from the silver, even though that gets asteroid defense and a good shooting area, bombers can't reach RC-5 from there (Δ bombers can.)

Battleships: Three for one invasion stack is good, especially a capital that will need a couple turns of bombardment. More battleships are worthwhile if we want to start operating more than one invasion stack at a time against Avowers. Also, there's that outlying Avower system RC-7 to our northeast (past Voyage) that could probably fall to one battleship escorting a handful of invasion ships.

Squadrons: We have 6 bombers, that's enough at the moment. If we want more, they get built last with instant rebase to the combat zone. Carriers have to come sooner if we want some, though they are expensive and don't directly contribute combat value. No Avower system is within Δ bomber range of any other.

As our stack invades Aestoria, keep conships with the stack building lanes for quick getaway later. I think the timing will be about right to have jump lanes down to RC-2 about the time we finish with Aestoria, even if the southernmost conship goes to build a starbase. Might have to push a few extra conships out of Xanadoo, it's the best place to pause growth as it's over the health cap.

When attacking, look to promote one ship to a Medic (Upgraded Repair Staff / Improved Repair Staff). It does not require Combat I so any ship can get it. Carriers work great, although they won't make 5 experience for the second promotion; a destroyer works well for that.

We may not want Utopia right away, a 30% hammer drop from Mechanized to Utopia is a pretty severe weakness in wartime. We might even want Commerce Theory first (enables Spaceports, another bank analogue.) Up to you.

 

[Mojoqmeyvam]

Do you know which Halis is at war with New Earth? Hector just asked me if we should join in his war with them. Being asked to declare war the turn before I was going to? That would just be too good to be true.

Never mind, I turned down the request. If it's Orange Halis, we'll lose some con ships if I declared right now. If it's the other Halis, then it's a bad idea.

 

[Mojoqmeyvam]

WAR!

Halis has lost Aestoria. We lost 1 Delta Invasion in the glorious annexation.

We never got a chance to colonize any of the last systems, but Lessote (formerly Lu's RC-8) will fall soon, as well. Unfortunately, the Con ship that was near was eaten by a pirate, so warp lanes are a ways away. I have not moved the two SODs this turn, I was going to leave the honor of declaring war on Lu to the next player. I did not build a starbase in their territory, it didn't seem too useful, and I wanted to save the funds for research. If we want more airpower, it's better to build another carrier.



Utilitarianism came in this turn. I've put Commerce Theory in as a placeholder, Omega PDS and Spaceports seem immediately useful, but prioritizing the other techs and heading for Omega Invasion Ships (Long Range Weapons) might make sense.


I've seen Delta Destroyers from barbs. Nobody else can build Delta Destroyers. Not even us.
Aestoria comes out of resistance in 2 turns. I think we want it to build a School of Zealots, and then a Medic II Destroyer or 2.

 

[God-Emperor]

I've seen Delta Destroyers from barbs. Nobody else can build Delta Destroyers. Not even us.

The pirates ship mix is determined entirely by game turn, it has nothing to do with who has what techs.

Pirates first spawn on turn 30.
The target number of pirates starts low and grows until turn 250, at which point it is at the max and stays there until the end of the game. Each turn it checks the target number of pirate ships and if the actual number is less it will add a random number from 1 to half the deficit +1. They can be added to any tile that is not visible to anyone (and does not have a bad feature).
In the first 1/3 of the game (through turn 183 at standard speed for which the game goes to turn 550) the type list gives each spawn:

• 1/5 chance of destroyer
• 2/5 chance of invasion ship
• 2/5 chance of PDS

In the second 1/3 of the game (turns 184-336) the type list gives each spawn:

• 2/5 chance of delta destroyer
• 1/5 chance of delta invasion ship
• 1/5 chance of delta PDS
• 1/5 chance of battleship

In the final 1/3 of the game (turn 337 and later) you get:

• 3/8 chance of omega destroyer
• 1/8 chance of omega invasion ship
• 1/8 chance of omega PDS
• 2/8 chance of delta battleship
• 1/8 chance of omega battleship

The existing pirate ships are not upgraded, this is just for newly spawned ships.

Because of this it is fairly common for the pirates to get some ships types before anybody else, unless you focus strongly on the military tech line.

 

[Mojoqmeyvam]

Thanks. That's really useful to know. Omega Destroyers are going to be a ridiculous PITA to deal with, so T337 is a good upper limit on when we need to win by, and I think we'll be able to. Heck, regular battleships are going to be a near-term problem unless we get Omega PDS, which means we should probably prioritize Commerce Theory now, and get an Omega PDS in each system.

 

[God-Emperor]

Oops. The last turn for the 2nd list is turn 366, not 336. The final list starts on turn 367.

 

[T-hawk]

That is great info indeed. Thanks! So there's a REALLY sharp breakpoint on T184 when the pirates advance massively in weaponry. It's as if they get about a dozen midgame techs at the same instant. Now I wonder if regular Civ might even like such a mechanism; historical events like the fall of Rome to barbarians have never been properly modeled.

I've never found pirate battleships to be much of a problem. One or two bomber strikes and it's below the strength of a Δ PDS. The worst it'll do is pillage a couple lane links until you can get a heavy enough ship there to kill it. Ω destroyers (3 moves) will be a pain, but we should have no trouble winning within 150 turns from now. Need only about 10 techs x 10 turns each.

Anyway, sounds like good war progress, and timmy will get to have a smashing good turnset.

timmy827 <--- Up Now
Olodune <--- On Deck

 

[T-hawk]

Thanks to a suggestion from deanej in the FF thread in the mods forum, I put in a fix for the zombie AIs at game start. See here.

A few more notes from code reading, in CvSolarSystem.py:

Each homeworld system always has exactly 6 planets. It must have an Earthlike planet in the first ring. It must have at least 6 food total and at least 4 population support in the first ring.

Every system must have at least one 2-food planet in first ring. This avoids the situation where a Forge system might never grow, with the food penalty keeping it at 0 surplus permanently. Also, the first ring must always have at least 3 production and 2 population total. If a system does not meet all of these requirements, it is thrown out and regenerated.

The "best planet" in a system is selected as 2 * food + production. So if two planets are tied on food, the tiebreaker is production. If that's also tied, it's the innermost. (I would argue that size should be another tiebreaker.) There is just one case where a lower-food planet can win, an innermore 1-2-3 orange over a 2-0-3 white. The numbers are such that 0-3-3 gray could also win over 1-1-6 gems, but since there must always be a 2-food planet, that will win instead.

There is a hard prohibition on a size 3 Earthlike planet. If one would occur, it becomes a different size 3 planet instead.

 

[Olodune]

A few more notes from code reading, in CvSolarSystem.py:

Each homeworld system always has exactly 6 planets. It must have an Earthlike planet in the first ring. It must have at least 6 food total and at least 4 population support in the first ring.

Every system must have at least one 2-food planet in first ring. This avoids the situation where a Forge system might never grow, with the food penalty keeping it at 0 surplus permanently. Also, the first ring must always have at least 3 production and 2 population total. If a system does not meet all of these requirements, it is thrown out and regenerated.

The "best planet" in a system is selected as 2 * food + production. So if two planets are tied on food, the tiebreaker is production. If that's also tied, it's the innermost. (I would argue that size should be another tiebreaker.) There is just one case where a lower-food planet can win, an innermore 1-2-3 orange over a 2-0-3 white. The numbers are such that 0-3-3 gray could also win over 1-1-6 gems, but since there must always be a 2-food planet, that will win instead.

There is a hard prohibition on a size 3 Earthlike planet. If one would occur, it becomes a different size 3 planet instead.

Very interesting -- I actually think this looks like good starting location balance.

The "best planet" criteria is for citizen assignment AND planet specific builds? Or am I misreading your description. For citizen assignment planet size shouldn't matter, but for those +1/per-pop buildings the planet size is crucial.

 

[Olodune]

So there's a REALLY sharp breakpoint on T184 when the pirates advance massively in weaponry. It's as if they get about a dozen midgame techs at the same instant. Now I wonder if regular Civ might even like such a mechanism; historical events like the fall of Rome to barbarians have never been properly modeled.

It would be neat to have stronger barbs in standard civ -- probably as an extension of the "raging barbs" option. The barb spawning rules would have to be tweaked a bit since most games of civ will have little free land by the late medieval period. I suppose you could have strong barbarian filled galleon invasion fleets as long as an island remained open.

Anyways, I'm looking forward to the Space Pirate threat in this game (my test game never made it to t184).

 

[timmy827]

Was out yesterday with a long football nightmare, but see it now. Smashy-smashy coming tonight

It looks like Commerce Theory should be followed by Diplomacy (nothing useful) -> Ascendancy Theory (the +2 happy/health are the only things after Commerce's banks that really help us.

We are still planning Utopia for now right? Will do that.

 

[Mojoqmeyvam]

Utopia will drop our production significantly, but I think our fleet is large enough at the moment to do plenty of . Especially with Utopia, we're going to need the extra cap room, so a beeline for Ascendancy Theory sounds good. However, we'd have a 35% commerce penalty civ-wide. Break-even is currently at about 40% research, this will drop to something like 20-25%. Of course, we'll get a few hundred gold in pillages, too.

Interesting planet balance. A size 3 Earth-like at the start would lead to a huge boom. And because of how growth works, guaranteeing that the Forge CAN grow also guarantees that other civs grow in a reasonable length of time. Waiting 37 turns to get the 2nd pop...ouch. The Forge aren't weak because -1fpt slows down their late-game growth, it's that growth accelerates as you get bigger, and they lose 18 turns on a non-Earthlike system in getting that 2nd pop. I wonder: would replacing -1fpt with a -2 health penalty for the Forge make them overpowered?

 

[T-hawk]

The "best planet" criteria is for citizen assignment AND planet specific builds? Or am I misreading your description. For citizen assignment planet size shouldn't matter, but for those +1/per-pop buildings the planet size is crucial.

That "best planet" function is for the free nut/mining facilities when a new system is founded. Planet size is ignored. This is separate from the AI decision logic on selecting planets to build facilities, which does include size (indirectly, by looking at each planet's total production.)

Commerce Theory then Ascendancy Theory sounds right. I'm not completely sold that we want Utopia right now while still building military, but your call.

I was playing a game this morning with my AI fix, on Monarch, and was astounded when Halis beat me to what would have been my second system by turn 40! They sure do a lot better without 25 turns of handicap. Which also means that goody wreckage can be turned back on since the AIs will claim their fair share.

 

[timmy827]

Fun but long turns. I did the stack split set up by Cyneheard and it worked spendidly, taking four Avower worlds with fairly small losses (couple of squadrons, many cheap invasion ships). RC-3 will fall next turn.

The question is whether to kill Lu or not. In favor of not killing him, RC-6 is sort of far away (west of Forge), also I did a poor job of prepping and did not spare construction ships to rail in that direction (but Lu has built most of a jump land to RC-7, so we can spare a lot of power. ETA probably around end of next turnset). Also in favor of not killing him is that he has Squadron Focus which we need (won't give up now, but will capitulate and with the new tech trading rules we will be able to swap with him). Also, we can tell him to research Intuitive Computers and he might finish it in time...The argument against killing him is that Motherland unhappiness is severe, 5 or 6 per system, so his nice planets really aren't that helpful. Aestoria (the fake-Halis capital) is pretty sad too, was hoping yellow Halis would finish off their last faraway world but no luck yet. Left some construction ships unmoved, as their next action may depend.

Be careful with planets coming out of resistance - no population assigned!

Some notes:
I will retract my earlier requests for more squadrons and carriers, they were not as effective as I'd hoped. Lu had the Bunker-equivalent in most systems, the bonus from that plus Destroyers (including some Delta) nerfed the damage. Lu only had one fighter squadron, but destroyers managed to shoot down a couple too. They were good for smahing battleships he foolishly left exposed, and worked OK for bombing culture defenses, although the battleship animation is WAY cooler

Squadron recon acts as fog-busting right? We may want to do that (tedious though) with spare ones now that the war has cooled down. Interesting byproduct of the large distances and slower culture expansions, galaxy is all settled but still significant barb activity.

I did revolt to Utopia near end, so some more happiness buildings may have to get put into queues. WW is not too bad.

We should start a starbase near one of Lu's uranium sources; don't need it until the very end but would be nice to avoid some jerk AI stealing it from under our noses.

Had no clue if the colony ship can go anywhere, maybe we can delete.

 

[Olodune]

Sounds like some solid warmongering

I'd like to take a closer look at the save before I weigh in on the elimination of the Avowers.

Good idea to grab a uranium source while we can.

I'll play after work today.

 

[T-hawk]

I think I'd take the capitulation for the tech. Motherland unhappiness in former Avower systems won't add up to another tech before the end. It will be manageable; the vassal itself grants a happy face and we get 2 more happy at Ascendancy Theory. Also, is Hector still willing to be a voluntary vassal? If so we could take that too for another happy, since one vassal already causes diplomatic problems but we really don't have any need for diplomacy now (power graph lead means no attacks.) Note that we might have to do Hector first, since taking Lu might make Hector think we're too close to winning. (Speaking in generalities, I don't know the specifics of AI vassal decisions.)

Squadron recon acts as fog-busting right? We may want to do that (tedious though) with spare ones now that the war has cooled down.

I'm not sure. Depends on between turns whether the visibility is cleared before checking for barb spawns. A tip to reduce the tedium (read this, never tried it myself) - with an air unit (squadron) you can hold Shift and click the recon button several times to queue up several actions.

I will be out tomorrow night, but if Olodune plays early this evening as usual, I'll slot in later.

 

[Olodune]

A tip to reduce the tedium (read this, never tried it myself) - with an air unit (squadron) you can hold Shift and click the recon button several times to queue up several actions.

Interesting. That could be very useful in SP games with Airships patrolling the coast -- so thanks!

I will be out tomorrow night, but if Olodune plays early this evening as usual, I'll slot in later.

We'll see As long as I don't get held up at work there is a good chance.

 

[Mojoqmeyvam]

Two vassals? Sounds like a plan. Maybe we should see if Orange Halis will surrender, as well? I guess this means that the Syndicate is going to keep that starbase in our territory. Oh well.

We should delete the colony ship. I knew on my turn that it had nowhere to go, but I didn't for some y reason. Unless the barbs raze a system, and then we'd have to win the settling race to reclaim it, the colony ship's useless. And it seems that the barbs will capture systems (see Lessote, formerly Pirate, Avowers, and Orange Halis).

It's hard to avoid Invasion Ship losses unless you're using Omegas on Deltas, but, hey, they're cheap.

 

[God-Emperor]

we really don't have any need for diplomacy now (power graph lead means no attacks.)

In FF you can't count on that. In the python code it can force the AI to start a warplan, in the doAIWarChance function in CvFinalFrontierEvents.py. So in addition to the usual war related stuff, the python can force an AI into war in a purely random way. Even if they are friendly.

Every turn it checks every team vs. every other team that team has met and uses a purely random chance to force a warplan. The odds are 6 in 1000 if furious, 4 in 1000 if angry, 2 in 1000 if cautious, and 1 in 1000 if either pleased or friendly. Although these sound like small chances (and they are), it is every turn. If you spend 100 turns at cautious, that 0.2%/turn chance works out to about an 18% chance that they will get a warplan forced on them. This doesn't insure that they will actually attack, especially if it is one of the "preparing" options - it does the standard warplan stuff. Note taht this is purely random - there is no factor that causes it to ignore the check, so it doesn't matter if they are already at war with someone else or what your strength is relative to theirs.

The actual warplan that is forced is purely random as well. There is an equal chance of it being any of WARPLAN__PREPARING_LIMITED, WARPLAN_PREPARING_TOTAL, WARPLAN_LIMITED, WARPLAN_TOTAL, or WARPLAN_DOGPILE.

Edit: oops - aside from having to have met the other team, there is one thing that causes it to skip the check: they won't get a warplan forced for another team to whom they are a vassal. Or something like that.

 

[Olodune]

In FF you can't count on that. [...]
The actual warplan that is forced is purely random as well. There is an equal chance of it being any of WARPLAN__PREPARING_LIMITED, WARPLAN_PREPARING_TOTAL, WARPLAN_LIMITED, WARPLAN_TOTAL, or WARPLAN_DOGPILE.

Presumably this was added to make wars more common? Since the last third of the tech tree seems to be focused on military techs, I think it makes some sense. With the random rolls it conveniently gets around the DoW-Immunity that is too common with the standard diplomatic model ...

Still, I don't think I would like this "logic" to be included in the standard game