Slashman wrote:

1) increased Ai awareness of when a player is nearing victory and the desire to counter this.

I don't know how to do this precisely, but in AIpersonnalitys, I saw some parameters about a "attitude" level when another empire lead score or number of regions. As I already use the file, i could write a more aggressive value here. But I have no idea of what it produce exactly !

Each major faction get also a personnality file, but i never tweaked that, i don't know what is into.

2) Alternate means of hindering a player who is far ahead that don't necessarily mean declaring war. (Note: the Morgawr's Black Spot ability does not achieve this in any meaningful way)

Iam thinking about to make a sort of fair balancing during the game, in the same way of progressive difficulty, so nothing magic, but I hope it could limit the snowball. 

3) Environmental dangers in a grand scale. Weather effects on land, Earthquakes, giant monster attacks etc.

Weather changed, and my old mods about that look totaly outdated (i even don't know if they still work !). I played Harsh winter difficulty, it's quite good. But I would like a hell winter mod. As I have no time for doing that actualy, and I need to understand new vanilla winters before.

There is a way to do some limited effects (world effects) or to create monsters, as you can see it in the quests. But all of that need to modify or add new contents in the quests definitions files, where it's possible to use a sort of script able to trigger effect or verify conditions, it's a totaly dark code looking like black magic.

Unfortunately, quests are the most difficult part of the game to code, and it's worst when you are just a modder totaly almost blind about mechanics used.

Above that, I constated i couldn't modify a quest file because reference file seem to be outdated, and unless amplitude provide the new one, there is no solution.

Dragar wrote:

To some extent, fighting against the snowball is to fight against the whole design of a 4X. The accumulation of small advantages into bigger ones over time is how good plays early on translate into a winning position later.  

So true!

The AI on this game is NOT designed to win. They just grow their empire (war, war, war...), to finally win by accident. Improved AI makes the AI reach their victories easier but doesn't change the core issue.

There's a diplomacy penalty when you're going to win, but that's about it, you can simply throw a few techs to the AI and keep things going.

Limiting snowball is quite tricky. Taking a look back con Civ 5: So much resistances about growing/expand made a game that making a new city past turn 80 would push your empire back instead of fordward.

I find usually the key point for snowball once you get your Era 3 key buildings up and running, and you start with the winter 3 pearl buldings, make your first level 2 distrcits, etc. At that point you will speed your game and surpass any AI most of the time (if you're not too damn far behind).

To me the most annoying thing is to play a catch-up game against the AI, so the progressive AI is a good idea to at least minimize that.

Dragar wrote:

To some extent, fighting against the snowball is to fight against the whole design of a 4X. The accumulation of small advantages into bigger ones over time is how good plays early on translate into a winning position later.  

Maybe so, but I still don't want my wins to be essentially the same each time I play. Either that or the AI needs to get better at opposing players with snowball effects. Now don't get me wrong, I'm not saying I want good play to be punished, but it really annoys me to see the AI seem to simply ignore the fact that I'm obviously going to win.

Even if the AI are ignoring me...then the gameworld itself should have some kind of reaction and give me something else to deal with instead of just hitting End Turn til completion.

To some extent, fighting against the snowball is to fight against the whole design of a 4X. The accumulation of small advantages into bigger ones over time is how good plays early on translate into a winning position later.  

Hi, 

I made a "snowball limiter" in my Progressive Difficulty mod.

Still not in release as I am testing it, so it take many time, as I do it while playing.

My first test was on Impossible Difficulty. As the AI was nerfed at start (and become stronger after), I reached the snowball because I was not attacked during a long time. 

Here, I should use an higher difficulty, as my Progressive Difficulty about Impossible seem to make it a few weaker than vanilla (but as I made also many new difficultys, it's not a problem).

With the limiter, snowball curve was not so heavy, but I set low values on the mod because I don't know what it will make, so I don't take too much risks.

As it is, it could be release, but It need more tests on some property for verify all trigger as it should, and it take time too.

After that, maybe I will set higher values for making the snowball limiting more effective.

Main director is technology count, affecting almost everything.

Example, as it is set for now, if you get 50 techs, it make (average) +75% cost on production and research. (in real, tech count effect on production is not so high because production is affected also by anothers parameters).

Then I can set higher multiplier, making it reach 150% for example. 

But if you get 25 tech, it make the effect in proportion. so if it is set for a max of 150% for 50 tech, effect will be 75% when you get 25 techs.

As the snowball is exponential, I need also to ameliorate this, making there is more modificators coming from others attributes (example : number of districts, population, etc). I already made some, and they look great, but it's not enough I think. There is some I don't know to tweak like city growth. As I know the code for doing that, I do not, because there is side effects. Example if city growth is nerfed, how it happens with broken lords ? If they are not nerfed too, then the game will become unbalanced.

My conclusion is there is in the game a lack of slow down, and actuals one are not enough strong. Slow down effects are : the more you research technology, the more it cost. Or the more you have population the more it take to gain much. Or the more you have citys, the more it cost in Empire plan and Approval.

So it's why I speak of "game unbalance" because in late game, these slowdown effects are not enough strong. So my actual project is just to make new one, and progressive one, they grow with rise of power. 

Slashman wrote:

Playing the more difficult races also helps. Forgotten or Ardent Mages for instance.

In any case, you have to give props to Amplitude for creating a game with races which are so very asymmetrical. That makes their job several times harder compared to other 4x games which all sport similar races with much smaller differences between them.

+1

Playing the more difficult races also helps. Forgotten or Ardent Mages for instance.

In any case, you have to give props to Amplitude for creating a game with races which are so very asymmetrical. That makes their job several times harder compared to other 4x games which all sport similar races with much smaller differences between them.

Thanks for the exhaustive reply Enchanteur, but I didn't mean to push you to try to make changes you aren't sure about. 

I think that what I referenced in my post would have to come from Amplitude themselves. And I doubt seriously that those things would ever be included in the current EL. We'll need to hold out for EL 2 before we see that kind of gameplay (if it is ever considered by the devs). 

Hey Enchanteur,

I've been taking a look at your mod as well as the Improved AI mod and while I suspect they will both help with certain things, the core of the problem is simply that there are no surprises in the late game which can help to mix up the power balance.

One of the reasons I mentioned Sword of the Stars earlier is because of things like their Grand Menaces which can literally turn a game on its head. Amplitude seem to not want to enable that type of gameplay, but I honestly think it would work rather well for the planet itself to start throwing some monkey wrenches into the works when the end game approaches. The other issue is that the AI do not specifically react to a winning player. Now I get that having all AIs declare war on a player at once is not pleasant for everyone, but there are no alternate routes that other players can truly take to hinder another empire sufficiently in the late game to cause that winning player to break out of the "hit end turn until victory" rut. Even espionage doesn't truly have that much of an impact because of strict limitations in its use.

In my mind it would take a combination of things to alleviate this issue, but those things would also need some control settings in the options for people who want the standard steamroller gameplay to remain the same.

So my short list would be:

1) increased Ai awareness of when a player is nearing victory and the desire to counter this.

2) Alternate means of hindering a player who is far ahead that don't necessarily mean declaring war. (Note: the Morgawr's Black Spot ability does not achieve this in any meaningful way)

3) Environmental dangers in a grand scale. Weather effects on land, Earthquakes, giant monster attacks etc.

Hi,

This discussion is very interesting.

I feel the same about the snowball effect and the no brain late game "search all, build all, fast".

After many thinking about what you say in this topic, I think my mod Progressive Difficulty should help a lot, but maybe not for all.

I made it because I also really dislike these late game problems.

It will help AI to not be overpowered, but it don't solve the snowball problem itself.

I think this snowball problem come from something unbalanced in the game. It's just when you get many city improvements all togethers, systematics luxurys boosters, maxed empire plan, etc... then the total resources you can produce is far more than costs. And in the city building queue, it can take only 1 to 3 turn to build something with a huge production cost, at this step of the game. 

I feel I have to do something about that.

And I have some ideas, I don't know what is doable or not, so it will need some testing for checking faisability.

It will not be perfect or fantastic, but If I can apply these ideas, the snowball effect could be far limited or even gone.

Ninakoru wrote:

Yeah, snowball potential on this game sometimes makes you reach a moment you're unstoppable.

There are a few interesting mods out there to rise the AI capabilities (Improved AI) and to make less useful that breaking point you're unstopable (Progressive difficulty).

I've been hearing about these mods but I haven't really seen anything conclusive as to how effective they are. I'm all for giving them a try if they really do make a difference.

Yeah, snowball potential on this game sometimes makes you reach a moment you're unstoppable.

There are a few interesting mods out there to rise the AI capabilities (Improved AI) and to make less useful that breaking point you're unstopable (Progressive difficulty).

Yeah... min/max is the least interesting way to play and require to have an AI that get incredible bonuses to keep up. I usually don't play games like that for the reason I gave... but most other games usually make it less repetitive since you need to use most features to succeed. I suppose every faction will need a slightly different min/max playstyle on the hardest difficulty, you are also dependent in a good start or it will be very difficult. The early game is what decide a game on Legend difficulty. As for victory type that would be whatever seem the most easy one to achieve... but I rarely stomach to finish a game over starting a new campaign.

I would also agree that Hard difficulty is enough and make the game really fun if you don't min/max your game-play and go for a more varied strategy, follow quests and do a bit of role-playing.

Sword of the Starts was a really fun game, many fun memories of that game... :)

Jorgen_CAB wrote:

So... I enjoy the game... just don't get how min/max play is so repetitive and boring and principally is all about pressing end turn more or less the first 60-80 turns.

The thing is that for me min/max play is always repetitive and boring. It is maxing out everything and taking every advantage. I assume you use a specific formula and always build/research the same things in the exact same order no matter what faction you play? I assume you ignore quests and don't go for any specific victory types?

I'm interested because I've had some surprising things happen to me over the course of several games playing on hard where the AI declares war at the worst possible time when I'm focusing on something else etc. 

Still I don't disagree that the game could use a few more surprises. Specifically, I think we need the equivalent of the giant sea monster that the last expansion brought, to appear on land and threaten to upset the balance of things. I don't think that black spot thing from the Morgawr is the end game upset that the game needed.

I think the game that does this kind of thing best that I can remember is Sword of the Stars where you get normal random encounters, but then grand menaces start to show up when the game progresses past a certain point. I remember losing a game that I had the upper hand in to the Puppet Master once. Fun times. 

One thing that I found interesting since I recently started playing Total War: Warhammer: That game actually has a greater emphasis on diplomacy than EL by far. You literally can't just ignore diplomacy and excel unless you're playing certain factions (and even then...it is still sometimes necessary). In TW:W I find I need to use diplomacy to make sure I don't fall behind economically as well as to ensure I don't get swamped by potentially hostile empires before I'm ready. It's kind of ironic since most people would logically assume that EL would have the more advanced diplomacy system

I'm curious as to how you 'do nothing' on the hardest difficulty and still win with no effort and what faction you are playing when you do this.