Here is an example of missions that the AI just straight cannot do despite
being super easy and giving you several free Provinces.
Rather than teaching the AI to do this these should never have been designed like that.
- 2
- 1
-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Here is an example of missions that the AI just straight cannot do despite
being super easy and giving you several free Provinces.
Rather than teaching the AI to do this these should never have been designed like that.
I was speaking strictly of AI RNG, which is what we monitor in those runs. Of course that we have people playing constantly! That's a core part of our development and QA process.
- 8
- 4
I've seen the AI dismiss it's own cores. For years. As for the AI, it is bad and adding missions to it has made it worse. I think a new system or a tweak is needed to it because it's an added bonus to the player. But there are SO many bonuses in this game already, and add to the fact that the human player is better at getting them is sad and makes me personally quit my campaigns in the early 1600s.
- 2
Meta always ends up in equilibrium and gets stale. That is why PvE players change nation and try for different achievement each time. For PvP, its just different.
Most online pvp games remedy this by regular updates keeping meta in constant flux. This gets +1 and that gets -1. Next month its reversed etc. and the "meta" "changes", in a rather artifial manner. Its a good thing though, to direct meta into a more fun gameplay. When the gameplay is really fun, the staleness of meta becomes more like aging in cheese, really
At this age of horrible mercenaries, quantity was in the meta of QOL. Same reason everyone kept pushing slacken button when it still did something. Now you are stuck with mercenaries that are just superduper good, and management nightmare
Also, while streams was mentioned, I want to bring up the assumption that vast majority of EU4 gaming is not PvP, and that PvP feedback is not only useless, but often detrimental to the PvE balance. PvP should instead be approached as a entirely different playmode, with possible custom rulestes for the host, like nerfed manpower or whatever.
Mission trees are going to be impossible to balance, if the AI is not able to utilise them properly, which seems to be the case. There are four approaches to fix :
1 improve Ai.
2 rework and streamline missions to make them easier for ai.
3 nerf mission rewards to the point of no one caring. This is good thing because while i like flavor, i hate railroading the game and some trees are doing it too much.
4. give AI different rules from players, concerning mission requirements, where applicaple. A shortcut really. could also be a guaranteed mission completion at certain time or interval. It's not very clean, but it's still better than the power creep player gets from being able to tap on mission rewards.
1 and 2 require too much resources to fix everything, but could still alleviate some of the worst excesses.
3 and 4 are very easy to do but are not very clean solution.
Same thing with estates, but to a lesser degree. AI might require some help with managing them. Personally I think there is a readability issue with the sheer number of descriptive text, with all the complex new privileges) but the estates are harder to manage now with nerfed monopolies. estate disasters are still not an issue, as they should be, but at least now you can't add privilegies at your whim and then revoke them at your leisure.
- 5
EU 4 is the dota of grand strategy games
If everything is op nothing is op
If everything is op nothing is op
- 1
.png){kind=avatar}
Some thoughts on the last replies:
1. We're trying to nerf perks across the board when possible. We nerfed Manpower and Force Limit sources across different features, as we acknowledged that there was an 'inflation' of those modifiers. We also reworked the sources of Development Cost reduction, as we thought that it was too trivial to stack them. We also nerfed the Loyalty given by Estate Privileges, as with the expansion to 6 possible, plus the number of available ones, it was too easy to stack it as well. Why don't try to dodge or avoid these rebalances, but we have to be careful, because the fix might be worse, as it may affect different game systems.
2. AI RNG is still a thing. We have a number of automatized nightly runs, and we observe different outcomes on each game. We obviously try to redirect things that might be troublesome for the player (for instance, AI Russia was broken for a few weeks during the development of Domination, making the Ottomans even more powerful as they had free reign to expand to the north until we found what was causing the issue), but we don't want to force specific outcomes, as we know that experiencing different outcomes is part of the game's core.
3. I beg to disagree with the idea that no one at Tinto is taking a look and assessing the 'meta'. One example, regarding Idea Groups: last year, we experienced in the Grandest Lan, but also from videos and streams, that Quantity+Econ ideas were widely used by almost everyone. So we took a look at how to change this, and we introduced a number of changes that peaked at 1.35, with the new Idea Groups (something that had been static at 21 since 1.0). Now a new meta will appear, obviously, but this is something we're not ignoring at all. The same could be said with the changes to the Combat System and Unit Pips in 1.34 and 1.35; these changes already make the game different from 1.30 and previous versions. I could look for even more changes that we've been introducing since 1.32 (the rebalance of religions such as Catholicism, Protestantism, or Reformed, the changes to older systems such as Army Professionalism and Slacken Recruitment Standards, etc.), but my point here is that we indeed take a look at rebalancing older systems and mechanics, and how they interact with the new content we add.
4. Mission trees are probably the hardest to balance regarding 'power creep', I concede that, as there's a fine line between making rewards appealing, overpowered, or meaningless. We try to adjust and readjust them as much as possible, and we'll keep polishing them among versions, that's something we're committed to, although not being an easy task.
5. There is no such as 'undetected to Tinto'. We proactively encourage reporting existing issues and making suggestions to improve the game, because we read this feedback, and try to answer it (and this is why I'm here, discussing the design philosophy and state of the game with you all
The real problem with mission trees is not that their rewards are overpowered (although many of them certainly are). It is the fact that the AI is wholly incapable of factoring in the existence of missions into its calculations. Were this not the case, I would actually enjoy playing against AI coming at me with overpowered missions boosting them.
A player who knows that some future mission will give perma claims or a union CB or permanent +0.5 army tradition down the road can make plans around using this mission later. So far as I am aware, an AI will only fulfill missions by accidentally bumping into their success conditions, without any accounting for missions not yet fulfilled or further down the tree.
So in effect missions become a massive buff to just the player and no one else. When missions were simple this buff was smaller. But as mission trees grow more intricate and longer, this buff accumulates.
I am not sure if there is an easy way to fix this. Presumably making a smarter AI is out of the question, otherwise you would have already made it by now. Maybe the best solution is to devise a set of alternate mission success conditions which apply only to the AI and are much easier to fulfill. This won't change the problem with the AI not seeing future missions, but at least it will help move them down the tree.
- 7
The real problem would then become that AI plays nations every time exactly the same
If you don't account for the power creep, in current situation, AI nations still behave at least with some unpredictability. This can't be said about players though, who are following the directions of the missions diligently :-(
- 1
And then you have things like AI picking idea groups, event choices, govt reforms based on the need to steer the game in a certain direction, rather than given thing being actually valuable from the gameplay perspective
This is probably a pipe dream even for EU5 with how relatively underresourced the development seems to be, but at this point I really feel like the only real solution is for the game to have a "Role-play" and "Ironman" mode with different rules for missions, and how AI picks idea groups and such. (For instance, in "Ironman" make it so that out of available options AI always picks the best one - the one with the highest weight- as opposed to a RNG choice between these weights)
- 1
The real problem would then become that AI plays nations every time exactly the same
If you don't account for the power creep, in current situation, AI nations still behave at least with some unpredictability. This can't be said about players though, who are following the directions of the missions diligently :-(
I think these concerns are exaggerated. The AI has all kinds of hard-coded factors weighing into its calculations already. But that doesn't mean the AI takes the same idea groups, gov reforms, or rivals every time. I don't believe letting it go through missions more easily will suddenly change this.
- 4
Well yeah no, and of course better AI would even out the situation, giving welcome increase in difficulty, as has been pointed out. But on the avenue of determinism, I think it would still be another step to a wrong direction.
Also I would like to bring up again my personal number one grievance with railroading AI nations: they suddenly become crazy hostile wanting a lots of my land, against their apparent self interest. An fruitful, 200 year long alliance with 100 trust will break because they want to complete a mission suddenly?
I am not against having nation specific goals, but i'd prefer them to be embedded into the environment, like byzantine cores, that are a natural prize to aspire for. An HRE nation does not need mission to become emperor, because they want to be emperor already anyway. If some nation should have extra incentive, they could have a mission spawned for them like estate agenda. If they should get extra reward, just make event that fires when the condition is set. And it there is not at least speculative historical lore they are based to, neither should exist.
Some mission contain nice lore, but i would just put that lore into events that pop up at relevant time, because reading all the missions in one go, when its not relevant is not doing a service to lore. Also some mission descriptions are so lazy it looks like a bored intern has written them. Consider, f.ex. some of the purple phoenix missions.
"The recovery of the Aegean Islands will secure the coasts of Anatolia" is not exactly packed full of the rich roman history of that region. It's just dry mission, and could as well not exist at all.
- 3
- 1
- 1
I only read the first few posts, so sorry if this was already mentioned, but it's not just specific nations that get this power-creep, though the missions trees are definitely the worst offenders both because how much free stuff they give but also because they railroad the entire playthrough. Also individual mechanics have been rendered pointless since you now get so many sources of things like merchants, accepted cultures, cost reductions everywhere, etc.
- 5
- 2
What you wrote was literally a Paradox! THE signature EU4 event is the greatest LAN party (massively attended, fun, and very entertaining PvP match in an amazing venue). And to make matters worse for PvP, there're frequent complaints by the PvE players when PvP balance is just-simply discussed or even mentioned (e.g. We don't want PDX wasting resources on PvP, etc). So almost any PvP discussion gets massively downvoted. And yet again, THE signature EU4 event is a huge PvP match.
We PvPer's already have pages of rules listed on discord, and often use tailored PvP mods. We've done this for many, many years.
Last edited:
- 6
- 1
The need to have mods and houserules underlines the problem, that the game is not really built to be played that way. That it is incomplete or defective in some way. Then again, if it's still being played regardless of the hassle, it has potential that players are willing to overcome to get the experience they can't get more easily elsewhere. (if i would want to play strategy multiplayer PvP, i would probably also try my luck with EUIV too )
So yes i see it from that side too. I have played some multiplayer with friend but more as a co-op and believe i would found real PvP too confusing and frustrating. Most of the game design simly does not support real time competitive play, and i dont believe in casual PvP.
PvE design goal is also detrimental to PvP. I remember first time I faced this problem when I played World of Warcraft, and as I was, what they called a /carebear, PvE player first, I found it often frustrating that fun class mechanics were nerfed away because of PvP balance, and for same reason mostly there was a OP rotation, where you'd buff the least popular idea group, i mean class, and nerf the quantity ideas, i mean the previous OP class, so that while there was never balance between classes, there was balance in how long each was considered OP
In a perfect world, all audiences would be catered to. Maybe that was never option in an mmorpg, but that would be option for EU4, because pvp and pve games are always played separately. To have built-in differences in the game for multiplayer (lobby with checkmarks for specific PvP rules, such as slow in game speed if there is a war going on, events spawning to players only simultaneusly and short pause associated to them, some chess clock time stored for every player to be used in national emergency and possibly even changes to gameplay, such as reduced manpower, or idea values that get's randomised at the gamestart etc.
Someone should give data on what is the % of serious EUIV players who actually do PvP though. I understand that Grandest LAN is an event to have fun and put up a show, and i don't think it has a lot to do with actual player behaviour overall. To make design choices based on PR happening is REALLY bad design choice.
All this does not mean I am against EU4 multiplayer the least, or that I don't like the event and would not want to attend. I really would love to!
- 3
- 1
To be honest, i think having a timer where the AI just gets the mission reward is a great solution. A player with the same amount of time would be able to complete it, so why not just let the AI do it so that the in game difficulty keeps up pace after the dreaded 1600s. Improving the AI always gets brought up in these discussions, but i frankly think its really hard in such a complex gameMeta always ends up in equilibrium and gets stale. That is why PvE players change nation and try for different achievement each time. For PvP, its just different.
Most online pvp games remedy this by regular updates keeping meta in constant flux. This gets +1 and that gets -1. Next month its reversed etc. and the "meta" "changes", in a rather artifial manner. Its a good thing though, to direct meta into a more fun gameplay. When the gameplay is really fun, the staleness of meta becomes more like aging in cheese, really
At this age of horrible mercenaries, quantity was in the meta of QOL. Same reason everyone kept pushing slacken button when it still did something. Now you are stuck with mercenaries that are just superduper good, and management nightmare
Also, while streams was mentioned, I want to bring up the assumption that vast majority of EU4 gaming is not PvP, and that PvP feedback is not only useless, but often detrimental to the PvE balance. PvP should instead be approached as a entirely different playmode, with possible custom rulestes for the host, like nerfed manpower or whatever.
Mission trees are going to be impossible to balance, if the AI is not able to utilise them properly, which seems to be the case. There are four approaches to fix :
1 improve Ai.
2 rework and streamline missions to make them easier for ai.
3 nerf mission rewards to the point of no one caring. This is good thing because while i like flavor, i hate railroading the game and some trees are doing it too much.
4. give AI different rules from players, concerning mission requirements, where applicaple. A shortcut really. could also be a guaranteed mission completion at certain time or interval. It's not very clean, but it's still better than the power creep player gets from being able to tap on mission rewards.
1 and 2 require too much resources to fix everything, but could still alleviate some of the worst excesses.
3 and 4 are very easy to do but are not very clean solution.
Same thing with estates, but to a lesser degree. AI might require some help with managing them. Personally I think there is a readability issue with the sheer number of descriptive text, with all the complex new privileges) but the estates are harder to manage now with nerfed monopolies. estate disasters are still not an issue, as they should be, but at least now you can't add privilegies at your whim and then revoke them at your leisure.
- 2
- 1
I thank PDX for the continued expansion of EU4. It's greatly appreciated.
I'd like to write that the game is far, far easier than EU3 or the early release of EU4. For example, a popular YouTuber just uploaded a video today playing as Bohemia. Before 1515, he already has Poland, Lithuania, and Hungary under a lesser personal union. He's already vassalized Brandenburg, Saxony, Byzantium, and Bulgaria with a Claim Throne casus belli on England. He's already inherited Burgundy. He's already the Emperor of the HRE. He's defeated the Ottomans in two wars.
This is an example of the game's current design philosophy. There's no challenge left for the YouTuber and stops his Bohemia Guide. There's nothing left in Europe for Bohemia to worry about. Bohemia has only disaster to deal with (30-years' war), which can be averted if a player takes religion and quickly removes the reformation centers. And the 30 years' war lacks any deeper content, because it devastated much of Europe for nearly 100 years. Doesn't matter, because a player can avoid the league war through some diligence.
- 3
- 2
Quantity - Eco for the -30% dev cost it used to bring, of which stacking is absurdly broken in an MP context (limited expansion, min-maxing tall bonuses to get dev clicks at 5 or less mana each). Quality just happens to go along really well with Eco as well.
I started up a game as Scotland today and noticed that, unlike previous patches, this time I was able to stack up enough diplo rep modifiers from reform type, estates, and advisors to get a royal marriage with Burgundy on the first day.
So now Scotland has an even easier way to beat England: just sit around and wait until you get the Burgundian inheritance, after which you can easily swamp England's force limit. And you can even increase your chances of getting the inheritance by becoming HRE emperor, which is easy thanks to the fact that Austria doesn't even try to contest the election.
This is a relatively minor issue, but it highlights how all of the gradual stacking of DLC features over time leads to players now being able to do things they couldn't do before simply by pushing all the right buttons.
So now Scotland has an even easier way to beat England: just sit around and wait until you get the Burgundian inheritance, after which you can easily swamp England's force limit. And you can even increase your chances of getting the inheritance by becoming HRE emperor, which is easy thanks to the fact that Austria doesn't even try to contest the election.
This is a relatively minor issue, but it highlights how all of the gradual stacking of DLC features over time leads to players now being able to do things they couldn't do before simply by pushing all the right buttons.
- 3
- 2
- 1
Didn't the event change so it's much less likely that you get the inheritance even with that 1st RM?
The last time I checked the event, your best chance of getting it was to be both HRE emperor and the strongest marriage partner. If Burgundy hates the HRE emperor then the marriage partner is more likely. Scotland has a good chance to be strongest partner since Burgundy tends to marry OPMs, so even without being HRE emperor it has a good chance to get the inheritance.