Dune Wars 1.9.4 Patch Feedback

ChrisAdams3997 · Jun 7, 2011

Suppose you have a system where it picks a random tech.
Suppose they have two techs you don't have, with costs of 500 and 800.
You capture a city that gives 50% of a tech; the expected beaker yield is 325.

Now suppose that they have three techs you don't have, with costs of 100, 500, and 800.
You capture a city that gives 50% of a tech; the expected beaker yield is 231. So, by adding another tech, you are now worse off.

I was struggling with the same realization, but I have a couple of ideas that I need to flesh the math out on so I'll come back to it. Ideally the player shouldn't feel cheated just because the enemy had one low level tech. This would push players to game it by researching those low level techs just to eliminate them from the pool of possible techs, something like what people do with RA's in Civ5, which is just... gamey.

This sounds very promising. Do they actively build more workers, or just send the workers towards spice tiles?

No, I've found they produce plenty of workers, but the worker logic issues defining what plots were improvable and the lack of appropriate valuing of additional spice tiles were a big hurdle before and explains much of the lack of spice production by AI's. It didn't help that they didn't think Spice Extraction was even a very useful technology to get :lol:

, but like I said, they'll now place value on it based on how much spice is in their borders.

Interesting. I have generally found massive settler spam to be a reasonable strat in Dune Wars, at least relative to vanilla. Getting the good city spots is important.

It is, but the timing has to be right. In the AI's case they were throwing new settlers out when they were already down to 40 and 30% research, plunging them down to nearly zero research sometimes in the early game (and then sometimes still spending on espionage at the same time). You really need to get the necessary techs that allow you to make new cities profitable before you spam them (or, at least the AI does), and if you tank the economy too early, you never get to those economy building techs until you're already too far behind.

Something else a player can do to expand early is found extra cities than what the economy can really support, but set a number of less important cities to idle proccesses to make up for it while they grow. I'm looking at teaching the AI to do this if in bad economic trouble, though only in the lower production cities and after the basics needs of the city are taken care of.

I'd also like to get them to emphasize culture if there is more than one spice in the next ring of city expansion and if that city expansion can be achieved in a reasonable amount of time. This of course assuming the city has already taken care of basic defenses and such.

Ahriman · Jun 7, 2011

This would push players to game it by researching those low level techs just to eliminate them from the pool of possible techs, something like what people do with RA's in Civ5, which is just... gamey.

I see a slight risk of this, but mostly they will trade for these techs anyway through diplomacy. There is much less ability to game things here than with Civ5 research agreements.
[In terms of Civ5, I only play Thal's balance mod, because I find the vanilla game horribly unbalanced, so sometimes I will mess up because I forget which mechanics work differently in vanilla Civ5.]

No, I've found they produce plenty of workers

Ok. In my experience they often lagged in producing improvements, but that was in part because they didn't adopt the +50% worker speed civic. [They used to all adopt Slavery, before we made that Harkonnen only.]

In the AI's case they were throwing new settlers out when they were already down to 40 and 30% research, plunging them down to nearly zero research sometimes in the early game (and then sometimes still spending on espionage at the same time)

Interesting. I think its fantastic that you're looking hard at what the AI is doing here. I only ever did fairly cursory analysis.

I'd also like to get them to emphasize culture if there is more than one spice in the next ring of city expansion and if that city expansion can be achieved in a reasonable amount of time.

I'd probably take it easy on this on this. The economic costs of emphasizing culture are seldom worth just grabbing an extra spice tile or two, and the spice tiles might be gone by the time the border expands anyway.

ChrisAdams3997 · Jun 7, 2011

I see a slight risk of this, but mostly they will trade for these techs anyway through diplomacy. There is much less ability to game things here than with Civ5 research agreements.

I've played with tech trading off for so long I almost forget it's there sometimes

.

What I want to do is add up the total of all of the techs they have that you don't, maybe only adding in half the tech cost for those you don't have the prereq's for, and use that as a basis of how many tech points you steal. From there if it's more than the cost of the first random tech picked to give points in, and there is more than one you have the prereqs for, I might spread it over two of them. If you only have the one you have prereqs for, then you'll get that tech and the balance of the stolen points into one of the newly unlocked techs it allows you to research. This I think should fairly handle a combination of cheap/expensive and prereqed/non-prereqed techs in the mix.

Ok. In my experience they often lagged in producing improvements, but that was in part because they didn't adopt the +50% worker speed civic. [They used to all adopt Slavery, before we made that Harkonnen only.]

If that's based on experiences before Deliverator built 1.9.1 off of Better BUG AI, that merge added in a lot better handling of AI worker training.

I'd probably take it easy on this on this. The economic costs of emphasizing culture are seldom worth just grabbing an extra spice tile or two, and the spice tiles might be gone by the time the border expands anyway.

If I did this, it would be for the case of getting to the third ring only where it would add at least two or three spice quicker. Right now the AI looks for a way to get the first border pop early as possible, with only a few other priorities ahead of it, but after that pays little attention to adding extra culture per-say (unless going for culture victory). This would maybe say, if that third ring has > x spice, and borderpop > y turns, and we have a cheap building adding culture not built yet, and there is no danger, not losing a war, etc., etc., go ahead and build the culture building. As a player I do this all the time because getting more spice early in the game can make a big difference in your research and expansion capabilities, it makes sense for the AI to consider it as well.

Unfortunately the methods of checking those plots for spice are relatively 'expensive' computationally for the small AI benefit yielded, which is the real reason I probably won't do this unless I come up with something more efficient than what's currently in my head. I try to keep functions called by the AI this often as streamlined and efficient as possible to avoid lag for people with poor old computers like mine, so sometimes there's a choice between functionality and efficiency.

Ahriman · Jun 7, 2011

What I want to do is add up the total of all of the techs they have that you don't, maybe only adding in half the tech cost for those you don't have the prereq's for, and use that as a basis of how many tech points you steal.

That sounds good.

If that's based on experiences before Deliverator built 1.9.1 off of Better BUG AI

Most of it is, yes. Though I thought we had earlier versions of BBAI built into earlier versions of the mod? What I thought he did was rebuild just on BBAI, rather than on some of the overmodcomps we had running that included BBAI? But I'm not sure.

On the spice; your reasoning sounds very sensible, and I'm glad you're including runtime in your decisionmaking. Go with whatever you think works best.

Jester Fool · Jun 7, 2011

Currently (1.9.4), on Immortal the AI will have the tech lead (sometimes overwhelming). On Emperor, the human player usually can get an early tech lead (but that disappears later in the game due to AI bonuses). If you accelerate the AI tech rate (not a bad thing mind you) AND make unit spam cheaper (allowing the same size armies), Emperor+ will not be playable IMO. Unit spam on Emperor is a problem as is and it gets boring and tedious killing unit after unit later in the game because you can't win without taking over a large portion of the AI's territory. All the victory conditions require a large amount of land (# votes, spice plots, terraformed plots etc.) I stopped playing a recent game because although I would win eventually, I couldn't stomach killing 30+ units for every city I took. I had over 20 cities and even though I whittled Ix down to 6 cities, he was doing fine on his own (ignoring my war success maybe? - Ix: -8 cities, Me: 0 losses (excluding a few shot down air units)). Long story short, go ahead and give the boost to research but don't lower unit costs. From a fun standpoint, I think that is the better way to go.

Ahriman · Jun 7, 2011

I dunno, I don't mind making the game harder. The game *should* be very hard on the highest difficulty levels. If that can't be handled, then turn the game difficulty down slightly.

I normally play on Immortal, but maybe I'll need to tone it back, but that seems fine. It doesn't seem like he's planning to boost research, just to get the AI to make better decisions about research priorities.

Rebalancing the AI away from unit production without decreasing unit cost seems like it might be a bit dangerous. But let's playtest and see.

Does the AI already get an advantage on civic maintenance costs? If so, then we're probably already favoring the AI by boosting civic costs.

I agree that the the victory conditions still require a lot of contrast, I seldom actually finish games.

Another thing to consider; some game options (no tech trading, no tech brokering, raging barbarians) can make the game a lot harder for the AI. So I think its important to have some room at the top of the difficulty scale so that the game is still tough on Immortal/Deity even with some or all of those options on. I normally play no tech brokering, but I leave tech-trading on; without tech-trading, then the advantages of a positive diplomacy modifier are too small.

ChrisAdams3997 · Jun 7, 2011

You should actually have less unit spam on immortal

. If I remember right, the only value changed in the handicaps xml is iAIUnitCostPercent as follows:

Code:

        [U]Original[/U]          [U]New[/U]
	<>100</>	<>100</>
	<>100</>	<>100</>
	<>100</>	<>100</>
	<>100</>	<>100</>
	[COLOR="Red"]<>95</>		<>90</>
	<>90</>		<>80</>
	<>80</>		<>75</>[/COLOR]
	<>70</>		<>70</>
	<>60</>		<>60</>

While at the same time I reduced the amount of the total budget civs will spend on their military. So for any where it hasn't gone down, they will theoretically build less units (unless there is somewhere else in the code that bypasses what I've done, still needs testing). At the higher difficulties I've assumed it'll be making enough units regardless, I just didn't want to cut them short in the 'mid range' difficulties (between noble and immortal), as if it cut back how many units they would build too much, it'd be too easy for the human to out build and conquer them. The human will always be better at war, can't get around that.

Currently (1.9.4), on Immortal the AI will have the tech lead (sometimes overwhelming). On Emperor, the human player usually can get an early tech lead (but that disappears later in the game due to AI bonuses). If you accelerate the AI tech rate (not a bad thing mind you) AND make unit spam cheaper (allowing the same size armies), Emperor+ will not be playable IMO.

That tech lead of course comes from massive bonuses and starting with three extra techs. I don't like playing on the really high difficulties myself because it's a much more artificial challenge than a competent opponent. It starts you at a big disadvantage, instead of giving you an opponent that is capable of putting you at a disadvantage.

I don't think we'll ever have that happen for a good player at noble difficulty, but if we can achieve it with less hokey AI bonuses than those given at immortal, that makes for a more fun game. If you can still beat it at immortal with a smarter AI, then I have more work to do :lol:

.

If you accelerate the AI tech rate

Just to touch on this in particular, the AI tech rate is only faster because they aren't doing stupid things they did before. One of the few things Civ4 has absolutely no AI handicap code for is cheaper technologies, they always pay full price. Where they get advantages in tech pace in higher difficulties is that everything else is cheaper so they have more disposable income to expand and research at the same time.

Does the AI already get an advantage on civic maintenance costs? If so, then we're probably already favoring the AI by boosting civic costs.

Yes, they do. In fact they kind of double dip in this one because they have a bonus for civic maintanence and inflation, so not only do they pay less for the civics, they pay less % inflation on it in the late game. But I really don't want to get into a big effort to rebalance the handicaps except in special cases like the unit costs (and that's only because it affects total units that the AI will build). I'd rather that good players get a better challenge with fewer AI bonuses. After all, why would we make Immortal easier just to make it playable when you can play a lower difficulty and get the same effect :mischief:

Jester Fool · Jun 8, 2011

After all, why would we make Immortal easier

I don't even play Immortal anymore - I can't stand the massive unit spam after the BetterAI merge. I play Emperor now and the unit spam is barely tolerable (I finished one game and stopped playing one). So, play a complete game (to the bitter end) on Emperor (or Immortal but I wouldn't advise it) and tell me how fun you think it is. Compared to vanilla, Dunewars already plays a level harder (no doubt the BetterAI). Before the merge, I could win on Immortal (got spanked on Deity though). Now, Emperor is challenging but it is hard finishing games (due to tedium). A better question would be why have Emperor, Immortal, or Deity if nobody can finish a game due to the late game becoming boring and tedious.

Ahriman · Jun 8, 2011

Compared to vanilla, Dunewars already plays a level harder

Really? I always found Dunewars at least a level easier than vanilla. The all-terrain transports weaken the AI, as does poorer choice in tech priorities, as does poorer synergy in civics, as does poorer ability to specialize cities, as does poorer ability to utilize aircraft.

Jester Fool · Jun 8, 2011

Really?

AFTER the BetterAI merge, yes. I used to be better at Dunewars than vanilla. Now I find Immortal vanilla easier than Emperor Dunewars. I think it is the sheer number of units the Dune AIs build (since spice adds more available commerce compared to vanilla).

ChrisAdams3997 · Jun 12, 2011

Dealing with Unit Spam
And other thoughts on AI, handicaps, and life

I know it's been quiet around here a few days but don't let it fool you, I've been working throughout the week on several things

. To start with, I agree about the excessive unit spam, especially on higher levels and have been trying ideas and testing as extensively as I can to get a handle on the mechanics of it. As a result I'm trying several things to reign it in with a good balance across difficulty levels and have made some changes to how the AI evaluates how much military to build.

For the most part, the AI builds up its military until it reaches certain economic thresholds, where that threshold is influenced by the leaders personality, current wars or warplans, etc. I've kept this basic idea, but have changed some of indicators and the balance of how much of their budget they'll spend, how sensitive they are to economic trouble, etc.

One of the basic problems with late game unit spam is also the non-linearity of economic growth towards the middle and late game, yet unit costs remain linear. To address this I'm experimenting with a system where the cost per unit increases as you build more units over your 'free unit' cap. So the first unit over the cap costs 1 gold, but by the time you have 30 paid units, they might each cost 1.6 gold each, or 48. I've just started testing this, so the exact math is still not finalized and I may change some of the parameters I'm using.

For now, the amount extra is based on a ratio of the (paid units : free units) so it does scale with empire population. However, to avoid conquerers getting big discounts to new units from conquered cities, the population used is based on population * %culture in each city, so conquered cities count for less 'military support' here, and it's unaffected by civics unlike the base free unit calculation. There's already enough mechanics that encourage snowballing, I didn't want to add another :crazyeye:

.

I've also decided to generally rethink the way handicaps are setup. I'll be honest, I've never liked the way Firaxis set up the higher difficulty levels, just don't think they put a lot of thought into it. I've moved away from giving the AIs a bunch of free stuff at the beginning of the game (workers/techs/etc.) that gives a big immediate lead that the player has to catch up to, but tends the fizzle toward the late game, make early wonders/religions impossible, etc. The setup also heavily favored unit spam with the way unit cost handicaps were set.

One of the worst things here is that the 'perEraModifier', which affects a number of things for the AI, giving them increasing bonuses across the game based on era, was coded to decrease AI unit costs throughout the game, meaning in the late game when unit spam is already encouraged by non-linear economic growth, the units were also getting cheaper for the AI. I've removed this and also made unit costs hold much more steady between difficulty levels rather than having them fall off sharply in the upper levels.

What I've aimed for instead is fairly equal starts at all the high difficulty levels for the player, but setting up the AI bonuses intelligently to encourage stronger AI economies for the human to compete against without encouraging massive unit spam. I've shifted a lot of things around, included a new tag, and altered some existing behaivor. I'm still testing and this kind of thing has a lot of variables all affecting it at the same time so is difficult to tune and perfect--everything is interconnected--but I'm making some nice progress it seems in my test games. I'm hoping to put out something for others to test in the next week or so, so for those watching, thanks and be patient

.

Oh, and I might have lied about the thoughts on life part, sorry :lol:

Jester Fool · Jun 12, 2011

Can't wait to try it out! I have never liked fake difficulty and from the above post, neither do you. Better AI + more balanced unit spam = A MORE ENJOYABLE CHALLENGE. :goodjob:

ChrisAdams3997 · Jun 13, 2011

I appreciate the support

Ahriman · Jun 13, 2011

To address this I'm experimenting with a system where the cost per unit increases as you build more units over your 'free unit' cap. So the first unit over the cap costs 1 gold, but by the time you have 30 paid units, they might each cost 1.6 gold each, or 48.

I would tend to oppose this. I think that:
a) It is confusing and non-transparent to the human player, relative to constant cost per unit. Maintenance is already confusing enough, thanks to the idiotic inflation mechanism.
b) It favors the human player over the AI, since the human player can use units much more efficiently and so gains relatively much more from a system that favors small armies over large ones.

I've moved away from giving the AIs a bunch of free stuff at the beginning of the game (workers/techs/etc.) that gives a big immediate lead that the player has to catch up to, but tends the fizzle toward the late game, make early wonders/religions impossible, etc.

One of the worst things here is that the 'perEraModifier', which affects a number of things for the AI, giving them increasing bonuses across the game based on era, was coded to decrease AI unit costs throughout the game, meaning in the late game when unit spam is already encouraged by non-linear economic growth, the units were also getting cheaper for the AI. I've removed this and also made unit costs hold much more steady between difficulty levels rather than having them fall off sharply in the upper levels.

These two things seem in conflict to me.
I agree that the goal is an AI that is of roughly equal challenge throughout the game (call this Design A) - though initial boosts are important I think because the AI is incredibly vulnerable in the early game, to barbarians and to humans.

But your current view of the AI is that the AI is too strong to start with and then weakens over the course of the game (call this Design B).
If we remove a mechanism that gives increasing benefits to the AI at higher eras, doesn't that move us away from Design A, not towards Design A? It seems to me that removing this will only exacerbate AI weakness.

I am very skeptical that there is an obviously superior way that will improve AI performance.

What I've aimed for instead is fairly equal starts at all the high difficulty levels for the player, but setting up the AI bonuses intelligently

Can you please describe a few examples of what you mean by "intelligently"? It's a bit hard to analyze without this.
But I am very skeptical about removing the early game boosts, I really worry that they will get slaughtered by barbarians.

I'm worried that what you are doing is generally weakening the AI. What actions are you taking that are actually giving the AI an advantage?

I think it's great that you're attempting this, but.... people have been tinkering with the AI for ~8 years now. I'm not sure that this mod is the best place to totally redesign how all the AI bonuses work.

ChrisAdams3997 · Jun 13, 2011

*sigh* People really should have more confidence in me

.

I'm not going to go line by line through what I've done, as most of it only makes sense if you know the underlying code and the philosophy and logic behind it. I will say I've done a lot of study of that logic and AI behavior and performance - especially in the early game

- identified weaknesses, built in new 'debug' information to track AI economies, research, and troop buildups, and have considered more than you might guess as I've approached this. It is technical, and while the ultimate 'goal' is arguable, the means of getting there are less so without having someone else with the technical knowledge to argue it.

And, please, I don't say something like that arrogantly. It's just that it's easy to argue all sorts of things about a complex system and how to change it, but there are a very few people who have studied and understand the structure of the AI's decision making enough to have informed opinions on it.

But your current view of the AI is that the AI is too strong to start with and then weakens over the course of the game (call this Design B).
If we remove a mechanism that gives increasing benefits to the AI at higher eras, doesn't that move us away from Design A, not towards Design A? It seems to me that removing this will only exacerbate AI weakness.

It's very important that I point out there are different ways that an AI can give a challenge throughout the game. One is indeed to give the AI production and economic bonuses aimed in a way so that they can pump out massive amounts of units. But as Jester Fool pointed out and I agree with, this isn't really a fun way.

There are also, as I mildly suggested, many areas of the code that that 'perEraModidier' plays into that are appropriate and I haven't touched. This particular area however was not an appropriate or constructive use of it. The continued use of this tag along with inflation cost reduction for the AI still gives progressively higher bonuses in a fairly strategic way that helps keep them in the late game.

I've shifted the balance more towards the AI's at higher difficulties being more competitive in research and expansion, but cut back (hopefully enough) the parts encouraging unit spam at the same time. This isn't to say higher difficulties won't have more AI units compared to lower levels, they will because they'll have more robust economies to support more units. What they won't have is a more robust economy and cheaper units that encourages massive unit spam.

So to really sum this up, where, when, and what type of bonuses you give the AI have a big impact of the gameplay environment in as far as how they shape AI decisions. That's why setting up those bonuses intelligently is important.

I've also seen no evidence of AI's routinely failing in the early game due to removing the 'freebies' and playing at Monarch and Emperor in test games have routinely been able to only get to second or third in points by the middle game and am usually not in the tech lead. Though you'll probably have to play it for yourself to see.

It is confusing and non-transparent to the human player, relative to constant cost per unit.

I'm estimating the average players' ability to cope with such a mechanism to be quite adequate :lol:

The concept that more troops = more money is simple enough, the only players that would have any problems with this is ones that want to lay everything out on a spreadsheet before hand... Most players will look at their troop costs and decide if they can afford to build more or not. That said, I'll only be keeping this if it achieves what I want it to in terms of balance, I can disable it by commenting out about two lines of code

.

thanks to the idiotic inflation mechanism.

lol, you may not like it, but it's probably one of the better ideas Firaxis had. This goes back to the typical non-linearity of economic growth. Without it the late game would have ridiculous balance issues.

Jester Fool · Jun 13, 2011

though initial boosts are important

I don't have a problem with initial boosts. I want the AI to be challenging. However, sometime after the BetterAI merge (1.9.x) the unit spam became unbalanced. I know because I always finish my games (except for the one I will get to). Maybe I never noticed before since I typically won games (pre-merge) ~ turn 315 to turn 375.

This is the only game I haven't finished (out of > 50 complete games - that's where "Dune Messiah" comes from). Units killed is more than 525. Make note of my casualties. Statistics screen :

Spoiler :

The demo screen :

Spoiler :

Victory conditions :

Spoiler :

First thing to notice is that it is turn 375. I am not even close to victory but the game is obviously won. I have 77 spice but will never reach a spice victory. Diplo is out (unless I can take even more cities from BG - Corrino will be destroyed but that won't get me enough votes unfortunately). Ix won't surrender (but is helpless to stop me from taking one of their cities whenever I want - they had 14 when they attacked me). So even with the obscene amount of units the AI can field (I am third in troop count), the AI can't mount a REAL challenge, but it is terribly boring to kill unit after unit when the game is obviously won. If I proceed to kill my way to victory (which is really the only option), I will probably need to kill another 400+ units (most units killed so far belonged to Ix). FAKE difficulty fail. This was not a problem before the merge. I would rather have a fun experience (defined as finishing the game win or lose) than a supposed "challenging" AI (which is really just massive unit spam). Before the merge, I never could beat Deity and had to play well to beat Immortal (so I consider the balance better than now). There is no reason to make Emperor (which the screenies are from) such a boring experience to actually finish. That is what Deity is for (or Immortal if the goal is to be harder than vanilla).

ChrisAdams3997 · Jun 13, 2011

I'm also curious what the tech differences were in most of those wars?

or Immortal if the goal is to be harder than vanilla

I'm not aiming to be 'harder than' so much. What I would like is a setup where each difficulty level is a reasonable step up in difficulty where it's noticable, but you don't feel like you've gone from your buddies basement to the world poker tournament (or what ever it's called

) in one step. Noble of course remains the 'even' set point where there are almost no AI handicaps, Monarch is the average difficulty I expect most casual players who've played a few games to step up to, with Emperor being just a little tougher. Then it gets a little hairier.

In all cases, I've continued the idea in the AggressiveAI option that it'll favor more unit production and warfare than with it off, so if you want the more casual tech pace of Monarch with more units for the AI similar to a higher difficulty level, you can achieve that with this option. It's in a way a half step between difficulty levels, though it does carry some other changes besides the number of units they'll build.

Ahriman · Jun 14, 2011

*sigh* People really should have more confidence in me

Trust, but verify.

I think all gamers have heard many claims about AI performance even from professional game developers to be pretty skeptical about such things.

And not knowing anything about you or prior work you've done, I don't think it's unreasonable to raise a few issues to check that you've considered them or thought of the consequences.

I'm certainly no AI expert, but I'm pretty good at designing models and in thinking about unintended consequences. If you've already thought of the issues I raise, then great.

I guess what worries me is the AI's (lack of) strategic combat ability. One way that the AI can be hard to invade is if it has fairly large armies in all their cities. They have to do this, because they are very bad at shuffling the armies they have around (using transports) in order to repel an invasion. If you shrink down the number of units they have, I am worried that it might become too easy to destroy the enemy army piecemeal, because they can't bring the units they have to bear.
They need such a large army because their efficiency is so bad, because they can't predict and follow what the human player is doing. Ideally what a human would do when facing an invading force is to have much of their own army also in transports, within their territory, so that they could rapidly respond to whichever city the invader actually besieged.
If you can improve their defensive strategic AI to improve their ability to concentrate force - using transports or whatever - then I would be much more sanguine about cutting back on it's unit numbers.

I think it's great that you've picked up the torch on this mod, and I really appreciate what you're doing. But I think it would be disappointing if you were not interested in discussing the changes you were making. If nothing else, it is important for people to understand the changes in order to have a good idea of what might be causing some particular behavior that is observed. If it's all a black box, then it becomes difficult to do effective play-test.

My problem with inflation and non-linearities is one of transparency. The human needs to be able to figure out what the actual cost of a particular unit is, and why, and more importantly they need to be able to figure out the marginal cost of any particular decision. Inflation wrecks all this, because it means that if a city has maintenance costs of 6 but inflation is at 50%, then an effect that reduces maintenance costs by 1/3 actually saves you 3 gold per turn, not 2. So it becomes much harder to evaluate whether I should build a market, or a courthouse, for example.

@Jester
I wonder if the problem here is not that really about unit numbers, but it simply that all the victory thresholds are set too high.
Ideally, a victory condition should set in when it's clear that you will win and that no-one else (or other alliance) could stop you. Nonlinearities in economy size mean that this happens long before you control half the planet (unless there is only one other player).
I think this is a problem in vanilla BTS as well; domination and conquest can be very dull, and our victory conditions are a bit too close to theirs.

Jester Fool · Jun 14, 2011

I'm also curious what the tech differences were in most of those wars?

I had a tech advantage vs. BT. I then had to recover my economy from that war. Ix had 2x my power at that time. By the time Ix attacked me, Ix was the tech leader by ~6 techs. After the war, Ix is done with the tech tree (future tech) and I have 3 left before future tech. Corrino is maybe a tech behind me and BG are done IIRC.

Jester Fool · Jun 14, 2011

I wonder if the problem here is not that really about unit numbers, but it simply that all the victory thresholds are set too high.

Back in version 1.8, the AI didn't build quite so many units and the victory thresholds were fine. I enjoyed winning domination (but still had to kill huge armies). The difference was that you could kill off (or cap) 1-3 AIs before your tech lead disappeared. The 1.9.4 patch plays much different. Now, after killing off one AI, your tech lead is gone, the economy stalled, and all the remaining AIs field huge armies. So I think that the changes have definitely shifted the victory conditions because they all are based on land (which is now much harder to attain). Compared to Immortal vanilla, Emperor DuneWars is harder to win domination and the reason is unit spam. I have never killed 525+ units in an Immortal vanilla domination game before.

The solution would be to tone down the unit spam (more than in 1.8 but less than 1.9.4) (my preference) or to readjust the victory thresholds, which would be fine too. Right now, I don't think spice victory is even possible, diplo a long shot, terraforming probably impossible as well and domination is just boring. Conquest is obviously out as well. Not sure about the other conditions although you can win by time I guess (but do you want to win by time after playing 30+ hours?)

Dune Wars 1.9.4 Patch Feedback

Prince

Tyrant

Prince

Tyrant

Emperor

Tyrant

Prince

Emperor

Tyrant

Emperor

Prince

Emperor

Prince

Tyrant

Prince

Emperor

Prince

Tyrant

Emperor

Emperor

Similar threads