I enjoy reading your posts
Axioms, regardless of how far-fetched they sound on paper.
If I may have a request, as a fellow hobbyist game dev I'm also interested in the more technical details of your work. Having so many interconnected systems, you probably have some clever architecture managing it. How do your systems communicate? How do you handle modifiers/effects/triggers while also avoiding tight coupling? Please dedicate some of the posts to the coding nitty-gritty.
I always get this claim of being unbelievable or some related term. This never makes sense to me because for all practical purposes these mechanics are individually simple. Outside of the AI there is nothing difficult going on. It is almost all about flexibility.
I guess I'll give an example of the vassal mechanics. So a typical game has one type of vassal. This vassal type has to have most of the mechanics of a more flexible system like tribute/taxes, some military value, and so forth. So aside from the interface as code to choose vassal interactions you actually aren't increasing complexity very much, again outside of the AI. CK3 added individual vassal contracts, almost 7 years after I discussed them on my old blog and wiki for Axioms before my long break, if even Paradox can do it you know it was a failure of imagination and not technical capacity. Axioms still has a deeper vassal system than Paradox, but this mostly involves more granular interactions and a more serious survey of history. Programming wise my interactions are broken up so they are actually simpler rather than more complex.
In fact in Axioms the only thing that distinguishes "vassals" from any other diplomatic relation is a single variable and the user interface display of that variable. This allows for the implementation of a system substantially similar to the Roman Socii. The Socii system was relatively unique in history.
The way this worked was that on conquest the Romans would annex a small amount of land for a "colony", where Roman and other "Latin" colonists would live, these gave up citizenship in the city of Rome if they had it or their city of origin, in order to be citizens of this new colonial polity. These confederates would sign a treaty, of the type styled foedus, where they would be totally sovereign domestically but Roman foreign policy was their foreign policy and they were required to provide a levy for Roman wars, and the Romans were at war continuously for 98% of the ~5 centuries of the Republic, of a certain specified number of troops. There was no tribute. No fancy feudal marriages or obligations.
What Creative Assembly and companies like Paradox used to do 10 years ago was hardcode vassal types in a very rigid way. To borrow from your post the mechanics were tighly coupled. So functionally a large portion of what I am doing is sort of Object Oriented Design. Of course I only call it that because of the context of your post there is actual non-code design terminology for that. For the user interface and for the data structures the interface presents to the player what this means is that you really aren't doing that much work for the gains you are getting.
So to get back the the Socii example there was a rigid design distinction between "vassals" and "allies" with distinct code, interfaces, and other stuff like that. This meant more work to design a simple mechanic, more "tight coupling" in the data, code that was confusing for new developers, and adding a new feature required more work and at best you'd be copy pasting rathering than truly "re-using" existing code although often you wouldn't even be doing that.
In fact classical writers discuss the idea that the Roman Socii were actually vassals but the Romans and the Socii themselves wrapped the system in a thin narrative of "alliance" and "equality" because that made the Socii feel good about themselves and for the Romans those feelings reduced friction and the chance of revolts. Eventually of course the "Socii" started the "Social War". -AL is a Latin suffix that means "pertaining to" and thus the Social War is the "war pertaining to the Socii" or in pure English translation "the war pertaining to the allies".
Because my design is just implementing "loose coupling" in the meta-relationships of the data and mechanics you get most of the benefits while actually improving the code. Similarly by having flexible loosely coupled mechanics you have a framework to which you can attach a new mechanic very easily.
If you do indeed read my posts, and by posts you mean the +99 crit damage walls of text on my Substack you would have read the inagural post in the Esoteric Arcana series about Divination. This is a fantastic example of the benefits of my high level design choices. I can implement "prophecies" and even common "scrying" with practically no effort, mostly just some new user interface panels, because what is happening at the programming level is that "scrying" just applies an arbitraty level node of the "Intelligence Network" system to a province. It would be difficult to implement meaningful "scrying" in the common fantasy novel sense if one was already giving the player perfect information on every province, character, and population in the whole world. That is why the mechanic, if a game can figure out an implementation at all, is usually something like the spells in Dominions(no relation) 3-5 that disocver appropriately aspected "magic sites". Similarly in the same game series "divination" themed magic simply modifies the frequency and positive or negative value of scripted random events. My Divination magic, which covers all the classical forms, doesn't require anything meaningful code wise at all.
Additionally as far as the general claim of my design being implausible I suggest people consider that back in 2013-2014 I posted here and on my wiki and blog about mechanics which people had the same comments about and which are now, shittily to be fair, major mechanics in Paradox games. Secrets in CK3 is a subsection of something I posted about in 2014. Stellaris implemented something substantially similar to my "Intelligence Network" idea in 2021. CK3 also implemented a version of individual vassal contracts. There are some other things I believe. Royal Court implemented a frankensteinian Paradoxian amalgam of several of my systems in their new Culture system. Not that they copied me but just that the mechanics are substantially similar. Field Of Glory Empires implemented the Empire Age decadence system and a few other things similar to ideas I discussed here and elsewhere. I think this is pretty good evidence that the other things I am planning for Axioms are perfectly doable. It is just a matter of imagination, proper understanding of political concepts, and willingness to do something without an ironclad assurance that their is an audience. Game design conservatism is partially a limitation of the imagination and understanding of developers but I'd argue that it is something like 45-55 that and publisher conservatism respectively.