France/Japan discussion (patch 3.6)

Buildings don't give levels, they give XP. The "choose a promotion" code has no idea what granted the XP. The "gain XP" code doesn't know what gave the XP either.
Would it be possible to disable the yields on levels on the same turn as the unit is built though?

Unless you have the units stored promotion option enabled (not a default option), units soft-lock your turn until you pick their promotions. But if you fight with a unit on the same turn it is built that XP will only contribute to a level the following turn.
So maybe it's possible to disable levels from free XP on training indirectly?
 
Science part aside, I think the :c5culture::c5faith: on military/defensive buildings would be the best place for a change or nerf. That ability was designed to help with the early game, but scales quite well and is a push towards Wide, in contrast to the rest of Japan's kit. It is also the overall least thematic part.
In contrast to the rest of their kit how? Japan doesn't have much of a push towards Tall, if any.
If the ability was designed to help with the early game, and the issue is it scales too well, maybe it could be capped at a certain era? :c5culture::c5faith: from pre-renaissance/industrial/etc military/defensive buildings.
 
In contrast to the rest of their kit how? Japan doesn't have much of a push towards Tall, if any.
If the ability was designed to help with the early game, and the issue is it scales too well, maybe it could be capped at a certain era? :c5culture::c5faith: from pre-renaissance/industrial/etc military/defensive buildings.
If you need to cap a UA by era to make it balanced, it deserves to be axed. I think limiting it to defensive buildings only is enough of a nerf, if it's needed at all.
 
If you need to cap a UA by era to make it balanced, it deserves to be axed. I think limiting it to defensive buildings only is enough of a nerf, if it's needed at all.
Why?
I don't think it needs to be nerfed either, mind you. Dojo science changed to faith would probably be enough.
 
In contrast to the rest of their kit how? Japan doesn't have much of a push towards Tall, if any.
Japan's yields from its core mechanics (UA's GWAM and Dojo leveling) don't scale with empire size, and end contributing proportionally less with the increased policy culture costs from having more cities. And Japan's extra source of tourism is purely through great works, which also don't scale with empire size. Last, owning more cities means you get a higher tourism penalty against other civs (-5% :tourism: tourism per city your empire is larger than your opponent's). My experience with Japan is that they win culturally more easily if they keep a small number of core cities. Preferably resorting to puppets if they conquer cities, as puppets don't count for empire size for neither culture/science costs, nor for tourism penalties.

If you need to cap a UA by era to make it balanced, it deserves to be axed. I think limiting it to defensive buildings only is enough of a nerf, if it's needed at all.
As it is implemented, the trait has to insert yields directly in each individual building, the buildings themselves don't have a coded category. The whole military/defense thing is just a UA naming thing. If we want, we can easily just add "pre-Renaissance" to the UA's description and keep the yields to four buildings (Barracks, Walls, Dojo, Castle).
 
Japan's yields from its core mechanics (UA's GWAM and Dojo leveling) don't scale with empire size, and end contributing proportionally less with the increased policy culture costs from having more cities. And Japan's extra source of tourism is purely through great works, which also don't scale with empire size. Last, owning more cities means you get a higher tourism penalty against other civs (-5% :tourism: tourism per city your empire is larger than your opponent's). My experience with Japan is that they win culturally more easily if they keep a small number of core cities. Preferably resorting to puppets if they conquer cities, as puppets don't count for empire size for neither culture/science costs, nor for tourism penalties.


As it is implemented, the trait has to insert yields directly in each individual building, the buildings themselves don't have a coded category. The whole military/defense thing is just a UA naming thing. If we want, we can easily just add "pre-Renaissance" to the UA's description and keep the yields to four buildings (Barracks, Walls, Dojo, Castle).
When I play Japan, I do warring a lot, the more units you have an keep using and upgrading, then you use UA for GWAM more. I take religion which gives GG/GA points when you take cities, so it's also count. Japan is very good being a wide conqueror.
 
If we want, we can easily just add "pre-Renaissance" to the UA's description and keep the yields to four buildings (Barracks, Walls, Dojo, Castle
Removing the post renaissance boosts is so minor it wont impact balance. Another nerf that doesn’t nerf anything. Leaving only the walls line is a real cut with real impact.
 
If you need to cap a UA by era to make it balanced, it deserves to be axed. I think limiting it to defensive buildings only is enough of a nerf, if it's needed at all.
I think people are being too conservative, if a civ is in the top civ contenders I think we can go a little harder.

I say toss the mechanic. It won’t even really change Japan’s gameplay that much. I’ll still go barracks with authority regardless, I’ll still want my dojos, I’ll still build walls when I need them.
 
Re: giving Japan more :c5faith:,that Would make Japan a:c5faith: faith civ, or more of one than he is already. Is that what people want? I feel like we have enough of those.

It’s not just about what would make Japan balanced; what roll do we want him to fulfill in the lineup of civs?
 
Last edited:
I think people are being too conservative, if a civ is in the top civ contenders I think we can go a little harder.

I say toss the mechanic. It won’t even really change Japan’s gameplay that much. I’ll still go barracks with authority regardless, I’ll still want my dojos, I’ll still build walls when I need them.
So, we can do it infinitely: if a civ in the top, we nerf that civ, then other civs are going to be top civs and we nerf them. Or we do it only to Japan? In my games Japan never outperfoms as AI civ, it closer to bottom.
 
I think people are being too conservative, if a civ is in the top civ contenders I think we can go a little harder.

I say toss the mechanic. It won’t even really change Japan’s gameplay that much. I’ll still go barracks with authority regardless, I’ll still want my dojos, I’ll still build walls when I need them.
Re: giving Japan more :c5faith:,that Would make Japan a:c5faith: faith civ, or more of one than he is already. Is that what people want? I feel like we have enough of those.

It’s not just about what would make Japan balanced; what roll do we want him to fulfill in the lineup of civs?
Japan had :c5faith: faith for quite some time in the Dojo leveling mechanic. Before the experimentation with leveling in the UA, the Dojo would give :c5culture::c5science::c5faith: when a unit leveled up. Overall, the role that Japan kept distancing from was the economic/materialistic one, namely :c5food::c5production::c5gold:. Even when they had the fishing boat + atoll bonus from BNW, it was a :c5culture: culture version, instead of Polynesia's current :c5food: food one. The only time I remember them having an economic bonus was in very old patches, when the Dojo would give +2 :c5culture::c5production: to iron and horses tiles iirc, when they were still had the fishing boat bonus, and I think the Dojo didn't have yet the leveling mechanic. Most militaristic civs tend to focus on :c5food::c5production::c5gold: instead, and use that to quickly raise a large army. Other ones with a similar yield leaning to Japan's are Sweden, who gain :c5culture::c5science: from their UB, and Assyria, who is heavily focused on :c5science: science and doing so already from Ancient and early Classical eras. Japan trading :c5science: science for :c5faith: faith means some differentiation from those two.

One thing about :c5faith: faith is that this yield gives Japan one peacetime method of generating Great Generals and Admirals for their UA. Especially in human hands, who can plan either the policies, or the religion (To the Glory of God), to secure those purchases. While Japan doesn't have to focus on :c5faith: faith, their UA can benefit a lot from it during peacetime.

One role that has been explicit was that Japan would not focus on conquest, so to differentiate them from France. Japan was often envisioned as a more defensive militaristic civ, and one that wouldn't have to rely on conquest to thrive. That, however, means Japan needs some form of generating generous amounts of yields outside of conquest, as avoiding conquest means giving up plenty of yields that normally allow warmongers to snowball. The :c5culture::c5faith: on military/defense buildings is a nod to how to play this intended design, favoring defense and setting up leveling for the Dojo's yields.

The other role Japan differentiates from others is the whole GWAM part, as militaristic civs normally don't focus on great people outside generals. Also one reason why France's current UA sticks so much when discussing Japan, even if we ignore that it's the exact same mechanic under a different trigger. The GWAM mechanic by itself gives Japan a clear focus on :c5culture: culture, the question then is what other yield gets to mix with it. Assuming we don't want to make Japan so unidimensional towards :c5culture: culture; very few civs get only one type of yield with their bonuses, and Japan's design has been shying away from :c5food::c5production::c5gold:, so it's expected that their secondary yield focus would be either :c5science: science or :c5faith: faith, or both in lower amounts.

I'm not against Japan having :c5science: science as a secondary focus, instead of :c5faith: faith; in fact, if the UA's :c5faith: faith was switched for :c5science: science, Japan would have an easier time pulling the Longswordman Rush that the civ leans to. But, for now, I want to make AI Japan stop winning scientific victories better than dedicated scientific civs, hence why I'm focusing on :c5faith: faith.
 
So, we can do it infinitely: if a civ in the top, we nerf that civ, then other civs are going to be top civs and we nerf them. Or we do it only to Japan? In my games Japan never outperfoms as AI civ, it closer to bottom.
Congratulations, you discovered balancing.
 
I think people are being too conservative, if a civ is in the top civ contenders I think we can go a little harder.

I say toss the mechanic. It won’t even really change Japan’s gameplay that much. I’ll still go barracks with authority regardless, I’ll still want my dojos, I’ll still build walls when I need them.
Nah, that changes a lot. The early faith almost guarantees them a religion, and early culture also goes a long way.

But do you prioritize walls even if the city doesn't need them?

Restricting that to defensive buildings only:
1. They'll probably still found, but not almost always.
2. When you think about Dojo yields you no longer have to also take the culture and faith into account. That building (and Barracks too, to be fair) is stacked enough already.
 
I think there was an neat idea once from someone that was to give France a similar war score ability to the Aztecs something like this:

"When you complete a favorable peace treaty (Warscore of 25+), you gain a tourism modifier of 50% and one of 25% with all other civilizations (for x amount of turns) (stackable?)"

Otherwise I like the stacking mechanic of France as something unique. They have some synergy there in more mobile melee and the logistics promotion in 4UC, so putting more emphasis there seems fun too.
 
I really like these 2 new ideas for France and Huns now:
1687794303457.png

1687794337822.png

The French UA name could use a bit of work though ;)
Spoiler me gushing about these new French and Hunnic kits :

This fixes a rather glaring issue with both civs: They don’t have UUs that worked with their UAs.
  • Multi-Attack bonus
    • The multi-attack bonus works best with more low damage attacks, but France had a melee unit with a very high base CS. The ability is almost tailor-made for skirmishers, and the Huns have a unique skirmisher
    • The multi attack bonus had no incentive aside from more damage, which meant it didn't matter at all after the target died, or would have died anyways. More up-front bonuses were always more useful. Now with added yields on stacking attacks there is a durable economic motive behind stacking damage on targets. Much more interesting.
    • Dealing damage to units also affects war weariness, so this complements the Huns' existing bonus very well.
  • Unit capture
    • The hunnic horse archer cannot convert units. Only melee units can
    • This makes it really hard to use the Huns, because their UU line competes for the same strategic resource as the UA requires for a different unit. The Horse Archer doesn't require horses, but it upgrades into a regular heavy skirmisher that does, so you have a competition for resources created by your own UA. Not fun
    • Also, the unit capture probability scales with the difference between the base CS of your unit and the target. France has a unique Tercio -- a melee unit -- with a very high CS bonus, so his UU augments the conversion chance
    • Huns had no use for the trash units it created except to create traffic jams. France's new UA has a use for supply-free units that it gets on conversion. Trash units merely existing gives yields per turn on empire.
    • Capturing a unit means your attacking melee unit doesn't advance. You know what also stops melee units from advancing on kill? tile defenses. France now has a % chance of no advance for all his infantry units in his UA, and we can give his Chateau a 100% chance for no follow up too. The two abilities compound each other in a subtle way, so France uses melee units a bit differently.

Spoiler Update on my Japan rework proposal :

I have taken other users' feedback into consideration, and I submit this altered version of my proposal:

Draft 2 of Sakoku Japan
1687797363949.png

Blocks civilizations sending TRs to you unless you have a TR Or Embassy with them
  • You can allow anyone to trade with you and ignore that part of the ability, and your trade partners aren't limited by your trade cap
  • What is important is that you retain control. You have the final say on who gets to trade with you or not. Who gets to benefit from trading with you, and who gets the :tourism: tourism bonus on you.
All cities gain +1:c5happy:Happiness and +15%:c5greatperson: Great Person rate for each unused trade route
  • Functions as a 3rd kind of trade route vs the :c5gold::c5science::c5culture:ETR and :c5food::c5production:ITR options. You can send trade routes Nowhere and get :c5happy::c5greatperson:
  • It was 10% in my first draft, but after playtesting I didn't feel the impact strong enough. I think it should be good at 10% though.
I also dropped the idea of super-closed borders. As other users pointed out, it’s situational and hard to gauge the value of. It would probably also be a nightmare to code.

Playtesting Notes:
I played a game until Renaissance, turn 204 on standard with this civ, but then my save games got corrupted and CTD'd when I tried to load back in, so that's as far as I got.
King difficulty, Continents Map, Standard speed
Spoiler a view of my continent. My capital is under the tech banner :

1687814907071.png


Gameplay overview
  • I was sharing a continent with France and Zulus, with a large space occupied by city-states that kept us all fairly separate at first.
  • I had a Truffle / Incense start. Went GoHunt. Founded turn 95 for Ceremonial Burial and Teocallis. Reformed on turn 152 for Crusader Spirit. Enhanced turn 153 for Martial Arts and Symbolism
  • I went settled a 5 city core by turn 80 and then went to war against France until he was eliminated on turn 149.
  • I settled my 6th city, Nara, far away from my core around turn 100. Zulu declared war on me on turn 120, and laid siege to Nara while I was still trying to finish off France.
  • I had a lot of trouble with happiness because of this 2 front war, and the +2:c5happy: in all cities from unassigned TRs was the only thing that kept me in the fight and still able to found my 6th city in this time. I had 1 revolt but after that it was manageable.
  • I unlocked Steel (Samurai and Dojos) on turn 142 and that allowed me to crush Orleans, and relieve Nara with my best units. Dojo Samurai are more than a match for Medieval Zulus, so Shaka was asking for capitulation by turn 172 and I had his capital by 191.
  • I had met all the other 5 players on the other continent by that point. I founded the WC, only to find out that the other continent had let Austria go completely unchecked, build 11 wonders, make 8 CS marriages, and control the largest religion. This gave her 16 votes in the final session.
  • I was gearing up for a joint war against 4th place Indonesia, whom the entire other continent already hated, when I left for the night. When I tried to get back in today, the game CTD'd
Notes:
  • Since I was at war with my entire continent, I only deployed TRs to city states or they sat in my capital. I could keep 1-2 caravans unassigned and still get good value out of them.
  • The added flexibility of the unassigned TRs was surprisingly useful for early diplomacy. I always had extra unassigned TRs on hand, so I could send them out just to complete CS quests. I didn't feel pressured to have all TRs running as soon as they became available, so I could coordinate TR quest completions with emissaries' diplomatic missions to get the maximum influence benefit. This was really handy for keeping CS allies as proxies in my war, which was really important given the strategic position of the city-states on my continent.
  • I had the GP rate set at 10% per unused TR. This felt a bit low in practice. I didn't notice much increase in the amount of GPs I was getting. Would increase to 15%.
  • Japan is probably a little more dangerous on offence with this new UA, because I can cut TRs to prevent War weariness and unhappiness from lowering my troops’ damage
  • Martial Arts (unimplemented belief, but passed congress) has a huge synergy with Dojos. +75% XP on melee units. My samurai levelled fast.
  • Teocallis are really good. They should be nerfed
  • Tribute is currently awful. The yields are maybe 30% as high as they ought to be.

So yeah, I've now played part of a game with my version of 'Sakoku Japan'.

I definitely think it has every bit as much of a strong theme going for it as 'Shogunate Japan'.
I like that it keeps Nobunaga's Great Person focus, and therefore his VC focus, more open-ended. He has a tool that will help slow down another civ's CV, and a tool that can be used to create more GWAMs, or focus on another GP type like scientists.
You can play this Japan wide and aggressive like I did to leverage the global :c5happy: per city, and use your trade routes to gain strategic CS allies around the map, or you can focus a tall Artistry game and use the %:c5greatperson:GP rate, combined with a few :c5food::c5production:ITRs to pump out lots of great people, with a dollop of extra happiness for more :c5goldenage:golden ages.

Overall it felt like a success. Overall I think which one is better hinges on whether people want Japan to be :c5culture:/:c5faith: (Shogunate) or :c5culture:/:c5science: (Sakoku). I would prefer to drop the faith; the fewer guaranteed founder civs we have the better, imo.
 
Last edited:
Proposal for France
Napoleonic wars: Gain a free great work upon city conquest. Great works grant 2 culture, 2 Science and 2% strength to land units. Total great work yields halve every era.
 
This fixes a rather glaring issue with both civs: They don’t have UUs that worked with their UAs.
  • Multi-Attack bonus
    • The multi-attack bonus works best with more low damage attacks, but France had a melee unit with a very high base CS. The ability is almost tailor-made for skirmishers, and the Huns have a unique skirmisher
    • The multi attack bonus had no incentive aside from more damage, which meant it didn't matter at all after the target died, or would have died anyways. More up-front bonuses were always more useful. Now with added yields on stacking attacks there is a durable economic motive behind stacking damage on targets. Much more interesting.
    • Dealing damage to units also affects war weariness, so this complements the Huns' existing bonus very well.
  • Unit capture
    • The hunnic horse archer cannot convert units. Only melee units can
    • This makes it really hard to use the Huns, because their UU line competes for the same strategic resource as the UA requires for a different unit. The Horse Archer doesn't require horses, but it upgrades into a regular heavy skirmisher that does, so you have a competition for resources created by your own UA. Not fun
    • Also, the unit capture probability scales with the difference between the base CS of your unit and the target. France has a unique Tercio -- a melee unit -- with a very high CS bonus, so his UU augments the conversion chance
    • Huns had no use for the trash units it created except to create traffic jams. France's new UA has a use for supply-free units that it gets on conversion. Trash units merely existing gives yields per turn on empire.
    • Capturing a unit means your attacking melee unit doesn't advance. You know what also stops melee units from advancing on kill? tile defenses. France now has a % chance of no advance for all his infantry units in his UA, and we can give his Chateau a 100% chance for no follow up too. The two abilities compound each other in a subtle way, so France uses melee units a bit differently.
All are fine, but I'll miss the Eki/tile gain synergy. Moving the tile gain to Shoshone won't maintain the synergy since Encampments have the adjacency block so it's never worth building them outside of working range.

Also, there's a weird incentive to keep a unit alive to do the most number of hits possible in order to milk the :c5culture:, which quickly goes out of hand in the early game (and doesn't matter in the late game). The AI, on the other hand, will finish the unit off in at most 3 hits, gaining much less than a human.
 
L. Vern posted another AI test, with both Emperor and Warlord difficulty results, and the test reinforces my suspicion that AI Japan relies heavily on the interaction between handicaps and leveling to perform. In those tests, Japan was the 3rd best civ in Emperor, but scored 36th in Warlord, the difficulty with no AI handicaps. Moreover, the confidence interval puts Warlord AI Japan firmly in the "95% certain need a buff" category. This suggests that we need to find a way to tone down Japan at higher difficulties and, at the same time, buff the civ for the lower difficulties.

I still think that the adjustment for higher difficulties has to target the leveling mechanic. It is the only plausible explanation for how AI Japan can compete in equal terms with scientific civs (not the case when played by a human), it has a previous case of getting out of control on higher difficulties and it has an exponential formula for its yield output that high difficulty AIs can exploit. Replacing the science on leveling for faith still looks promising, as well as the idea of replacing the AI's experience handicaps for a comparable combat bonus instead.

How to buff Japan for lower difficulties is another matter, since we've been fixed on the Emperor performance. But I don't see grounds for nerfs by now; dropping Warlord AI Japan from 36th to even lower makes no sense. I want to check what thoughts everyone has on this point.
 
L. Vern posted another AI test, with both Emperor and Warlord difficulty results, and the test reinforces my suspicion that AI Japan relies heavily on the interaction between handicaps and leveling to perform. In those tests, Japan was the 3rd best civ in Emperor, but scored 36th in Warlord, the difficulty with no AI handicaps. Moreover, the confidence interval puts Warlord AI Japan firmly in the "95% certain need a buff" category. This suggests that we need to find a way to tone down Japan at higher difficulties and, at the same time, buff the civ for the lower difficulties.
Unfortunately I just can't buy in to the warlord results. For example, Japan has a solid amount of SV wins in the Emperor run, but no one is getting SVs in warlord. So any civ whose win % includes a good amount of SV of course is going to go down quite a bit.

I think only a warlord run with no time victory would give us a realistic test of how the civs are doing.
 
Back
Top Bottom