[R&F] When will we get a AI overhaul mod?

[leif erikson]

Moderator Action: The trolling and flaming needs to cease in this thread or it will be closed. Get on topic and cease discussing each other.
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889

 

[Aristos]

As for the "they won't this time around"... Beyond Earth never got its DLL released. But of course the community for that game was mostly dead shortly after release, so that doesn't really tell us that much.

Didn't know that, this reinforces my "intuition".

 

[bbbt]

I believe they released dll source code for both 4 and 5, so hopefully they will for 6 as well. I'm honestly quite surprised that they do so, t's pretty rare for game developers in general and almost unheard of for AAA ones.

I'd assume it would come after they finished releasing DLC/Expansions for the game. And god knows, they might be skittish after the cross-platform play issues, as any DLL mods won't work on Mac/Linux (tho they can't go on Steam either so it's likely a small audience).

Edit: Well I'm wrong, I googled and looks like the Civ 5 dll source code came out ~2 years after Civs release: before BNW!? https://forums.civfanatics.com/threads/civilization-v-dll-source-code-coming-with-fall-patch.476593/ That would put it inline with this upcoming fall if they did the same pattern, though as noted, I wonder if they'd be concerned about re-introducing cross-platform issues.

 

[tzu]

any DLL mods won't work on Mac/Linux

(Modified) Windows DLLs won't work on Mac/Linux. Ie. players have to be careful, that they download the correct DLL for the platform they use.

With the DLL source code available, DLL mods developed in any platform can be ported to the other platforms by creating the specific DLL.

 

[Gedemon]

I don't know for Linux, but I don't think mac can use something similar as a DLL.

That didn't prevented civ5 DLL mods to be posted on steam.

My fear for civ6 is that AFAIK this time they never promised civ6 to be "most moddable ever", and that they've made a modding framework which is fantastic for new civs/leaders with UU, UB (is the bug for Unique District still here ?), ie what was more popular in civ5 player/modder base, but not so great for complex rules change (I mean even when comparing civ5/6 without the DLL) , which means no total conversion or overhaul like for civ4 (which where possible for civ5, but only a few where made).

So maybe they consider the DLL release to not be worth the dev time for the targeted player/modder base...

Which would be a shame as I still think that civ6 does have the potential to be the "most moddable ever"

 

[ashendashin]

My fear for civ6 is that AFAIK this time they never promised civ6 to be "most moddable ever", and that they've made a modding framework which is fantastic for new civs/leaders with UU, UB (is the bug for Unique District still here ?), ie what was more popular in civ5 player/modder base, but not so great for complex rules change (I mean even when comparing civ5/6 without the DLL) , which means no total conversion or overhaul like for civ4 (which where possible for civ5, but only a few where made).

That's pretty much my own thoughts on it. They've been purely focusing on fluff mods when speaking on moddability from what I've seen.

The whole deal on it being the most moddable I don't actually know. That was a long while ago. Could well be extrapolated journalism or even my own poor memory.
I found an old Civ6 article that specifically took that from Civ5's advertising in comparison to 6.

Edit:
Sorry about that @tzu . I probably let that get to me back when I still had plenty of hope for the DLL.

 

[Trav'ling Canuck]

I don't know for Linux, but I don't think mac can use something similar as a DLL.

That didn't prevented civ5 DLL mods to be posted on steam.

My fear for civ6 is that AFAIK this time they never promised civ6 to be "most moddable ever", and that they've made a modding framework which is fantastic for new civs/leaders with UU, UB (is the bug for Unique District still here ?), ie what was more popular in civ5 player/modder base, but not so great for complex rules change (I mean even when comparing civ5/6 without the DLL) , which means no total conversion or overhaul like for civ4 (which where possible for civ5, but only a few where made).

So maybe they consider the DLL release to not be worth the dev time for the targeted player/modder base...

Which would be a shame as I still think that civ6 does have the potential to be the "most moddable ever"

It would also be a shame because supporting the modding community would let the development team play to their strengths - cool new features, artwork - and let the community flesh that out in areas that are less economical for the development team to spend time on or simply of low priority to them - providing accurate information through the UI, game balance, AI behaviour, etc.

I wonder if the longer term plans for the franchise also play into the release/not release decision? Not just whether more expansions are planned and if so how many, but also whether Civ 7 will be built on the Civ 6 engine or something entirely new? I'm not sure if planned expansions/intentions for Civ 7 would preclude a release of the current DLL, but general conservativism might lead them to keep it in house until they're finished with it and move on to something else.

 

[Aristos]

So maybe they consider the DLL release to not be worth the dev time for the targeted player/modder base...

I know you are trying to be polite, but I would have phrased the above as "they are scared to be put to shame again by the modders"...

 

[Gedemon]

I know you are trying to be polite, but I would have phrased the above as "they are scared to be put to shame again by the modders"...

That shouldn't be a consideration on their side, and if it was, then pointing at the weaknesses of their code would be a mistake on our side...

But IIRC we've already discussed this, we have the advantage of being free of development constraints like release date and priorities, and we work on an almost finished product, ie we don't write from scratch, we improve existing code in a larger framework.

 

[bbbt]

That shouldn't be a consideration on their side, and if it was, then pointing at the weaknesses of their code would be a mistake on our side...

But IIRC we've already discussed this, we have the advantage of being free of development constraints like release date and priorities, and we work on an almost finished product, ie we don't write from scratch, we improve existing code in a larger framework.

Absolutely, if anything it's a resourcing issue on Firaxis/2k's side not a coding issue. But without armchair quaterbacking we probably wouldn't have half the thread content

And speaking of armchair quarterbacking:

Their consideration would presumably be primarily if it will increase monetary return. I'd say it would as total conversion mods increase audience/sales on the tail end (leading to retention/initial sales on the next version) and increase social media PR/'awareness'. This would imho outweigh drawbacks like increase support costs/churn by less mod-savvy users (i.e. who contact support/complain about the game not realizing it's a mod issue).

Things I could see giving them hesitation:

As you said, there are not nearly as many total conversion mods as in 4, especially with 'clear hooks' so to speak (I'm thinking on a marketing perspective).

Large corps are generally crazily IP-protective (as said, I think Firaxis is pretty rare amount AAA to even release some of the source code). I could especially see them getting pushback if they were planning on re-using the engine for other titles (BEII; Civ: Age of Mythology, who knows). Granted, tho that didn't seem to stop it with 5, but maybe that decision was made before the decision to make BE.

 

[Aristos]

That shouldn't be a consideration on their side, and if it was, then pointing at the weaknesses of their code would be a mistake on our side...

But IIRC we've already discussed this, we have the advantage of being free of development constraints like release date and priorities, and we work on an almost finished product, ie we don't write from scratch, we improve existing code in a larger framework.

True in general, but the fixes that put them to shame were fixes so simple to make, yet very influencing in gameplay terms, that the above argument does not hold.

Absolutely, if anything it's a resourcing issue on Firaxis/2k's side not a coding issue. But without armchair quaterbacking we probably wouldn't have half the thread content

If you never tried to mod something for civ, then the above definition/qualification certainly holds true (for you). But for many that try to make civ better in UI, AI and gameplay terms, your sentence sounds a little offensive, don't you think?

 

[cvb]

I cannot help, anytime I see the word 'shame' in context of software development, my internal exception handler throws one and reports a 'type mismatch terror' - but many people simply feel better after telling other people, how gray and bad the world etc is ...

That shouldn't be a consideration on their side, and if it was, then pointing at the weaknesses of their code would be a mistake on our side...

But IIRC we've already discussed this, we have the advantage of being free of development constraints like release date and priorities, and we work on an almost finished product, ie we don't write from scratch, we improve existing code in a larger framework.

Right! And for example here:

[...]
I see nobody on this forum, who wholeheartedly says "All is fine". On the other hand: "strong" words won't help us either. Confucius said "Who speaks load in strong words will be heard, but forgotten." I believe a cooperative style will gain more than a confrontative one.
[...]
I think, the best option is to give the community access to the sources and let them mod.
Btw, you cannot compare the results 1:1, because when I modify, it is a piece of love - I don't count the hours ... and hell, no: there is no specification book, product requirements document ... nobody is going to tell me, what I do or not!

Talking about pieces of love - I think, this one has the potential to become the absolute Taj Mahal of civ6 mods. The only concern I have, is that in modern times with a lot of affected units it may become real WORK to play it

Der Logistik bleibt es vorbehalten, den Traum der Taktik zu gestalten. It keeps reserved to logistics, to shape the dream of tactics.

 

[Gedemon]

Talking about pieces of love - I think, this one has the potential to become the absolute Taj Mahal of civ6 mods. The only concern I have, is that in modern times with a lot of affected units it may become real WORK to play it

Der Logistik bleibt es vorbehalten, den Traum der Taktik zu gestalten. It keeps reserved to logistics, to shape the dream of tactics.

Thanks, I'm trying to implement the mechanisms in a way that won't cripple the AI (or will require minimal coding on the AI side), which should also allow casual playing without need of micromanaging everything (but should still provide a reward when doing so)

IMO the biggest difficulty in making a good AI for a Civilization game is to balance the rules between the requirements for a competitive AI and the human's demand for an interesting game, and that starts at the design level.

 

[Trav'ling Canuck]

IMO the biggest difficulty in making a good AI for a Civilization game is to balance the rules between the requirements for a competitive AI and the human's demand for an interesting game, and that starts at the design level.

It sure does. Every time you add in a rule that only the human can use effectively, you need to add a compensating bonus to the AI somewhere else.

It's no small feat to come up with a new game element that both adds to the fun decisions faced by the human player and can be efficiently navigated by the AI. I should give the Civ 6 development team credit more often for coming up with a lot of great ideas that fit this criteria.

There's a lot of potential in Civ 6, buried under a cavalier attitude to the accuracy of the information provided to the player, willingness to let errors go uncorrected, and apparent disinterest in providing a challenging game for experienced players (hope I get proven wrong on this last point if there's a future expansion).

 

[cvb]

IMO the biggest difficulty in making a good AI for a Civilization game is to balance the rules between the requirements for a competitive AI and the human's demand for an interesting game, and that starts at the design level.

I like the way your approach in the end limits the number of (fully operational!!) units of the HUMAN player at all, of only highly decorated (ie. experienced / "5stars&overstrength") units, of only full upgraded units (because of also temporary, "auxiliary" units) etc.
In this context I like also this one:

I don’t think Knights are overpowered and so require a nerf. They’re powerful, yes, but not overpowered.
Knights need to be powerful. In principle, the game actually makes you work hard to get them - you need iron, you need to research a tech path that is focused on military not infrastructure, you need two builders to make six farms for the eureka (plus researching the rest of Feudalism), and then you need to build them. [...] Where it goes wrong is this: [...]You ideally shouldn't actually have that many Knights. Just a few to help bolster your existing melee (or whatever) army. Knights should be like Varu - you have a few, they’re powerful, it there not your whole army. [...] They just need to be harder to get, harder to build in large numbers, and have better counters.

apparent disinterest in providing a challenging game for experienced players

They cannot make a game for people like me, with an brain routine about '2 2 0' or '3 1 gold' tiles directly connected to the eye & burned in more than 30 years ago ...

 

[ThunderLizard2]

It sure does. Every time you add in a rule that only the human can use effectively, you need to add a compensating bonus to the AI somewhere else.

It's no small feat to come up with a new game element that both adds to the fun decisions faced by the human player and can be efficiently navigated by the AI. I should give the Civ 6 development team credit more often for coming up with a lot of great ideas that fit this criteria.

There's a lot of potential in Civ 6, buried under a cavalier attitude to the accuracy of the information provided to the player, willingness to let errors go uncorrected, and apparent disinterest in providing a challenging game for experienced players (hope I get proven wrong on this last point if there's a future expansion).

The problem I see is the wholesale change of game mechanics between each release. Instead of building on something that worked, the devs start over with de novo mechanics and the AI always seem to get short shrift. Why not build on what worked in Civ IV and Civ V? This seems to be the way many games are going these days.

 

[Leathaface]

Civ VI's Australian Summer 2017 Update announcement on Steam has the following wording:

"ModBuddy will also receive additional updates in the future, as part of a modding SDK update later. These tools do not include DLL source for Civilization VI at this time."

This sounds like the DLL source could be possibly released in the future. Though the update was released Febuary 21st, 2017.

Here is the link to the announcement page: (search DLL source) https://steamcommunity.com/games/289070/announcements/detail/576861888092920827

 

[Ryika]

This sounds like the DLL source could be possibly released in the future.

You're falling for the oldest trick in the book, companies make those sort of statements all the time. All it means is that they're aware that you want this thing, and that they're not giving it to you at this time, so keep hoping, because they might give it to you at a later time. All that statement is, is a phrase to calm down the people who would otherwise be: "Where's the DLL??? Game's dead!!" - it says nothing at all about their intent going into the future, they may just be staggering the negative reaction to the realization that there won't be DLL access until a later point when less people care and the people who cared have continued playing in hopes that there will be DLL access in the future.

 

[Spudsie74]

I have put my two cents in before, but can resist....Can someone name a strategy game that DOES have an AI that can compete with an experienced player? I maintain that the complexity of Civ make it cost-prohibitive to develop a "good" AI. Mods, and tweaking the source code can make it better, but until they develop real AI (that learns, like a human), and have computers that can process massive amounts of information simultaneoulsy (a Quantum Computer, or the Trinity Point), there is no way you can cook up a program as adaptable enough to beat a good human player. The AI is basically working off a script...until that changes, you will have a "dumb" AI. If they DO come up with a scripted AI that is competitive, they would have to re-do it every time they tweak the game, which would be prohibitively expensive.

That said, Civ 6's AI is particularly lame..but I think that mods can take some of the rougher edges off it, and, if released, (and someone takes the time to do it), a recoded AI would be less objectionable. BUT STILL INFERIOR TO A GOOD HUMAN PLAYER.

 

[Trav'ling Canuck]

Can someone name a strategy game that DOES have an AI that can compete with an experienced player?

Through the Ages, and the AI doesn't cheat.

Also, Civ 4 & 5 at Deity, although with those the AI gets big bonuses to compete. Still, even unmodded those two have much more efficient AI at winning than Civ 6 does (so far).