1.9.5 Feedback

with a few additions that I happen to like.

Click to expand...

Ok, but let's discuss these too.
The main thing I was unsure of with my proposal; we let Feudalism have OR tech requirements, but Great Houses have and requirements. But then, Education is a fairly expensive tech, while Light Manufacturing is a cheap one, and we don't want it to be easier to get to Great Houses than Feudalism.
I'm also a bit worried that having weakened the requirements for Fanaticism will make it too easy to get to Feudalism.
Part of the reason why we had the OR option in the first place was because Feudalism was such a bottleneck, but with that bottleneck weakened, maybe we should remove the OR requirement and just make Feudalism require Education. Fanaticism is already a very strong tech, I think it is fine that pushing for offensive military shouldn't give too easy a path to Feudalism and Law of Arrakis.

In which case, maybe Dune topography is too high of a requirement, and we'd be better off with Light Manufacturing and Desert Trade? Or maybe Great Houses requires: Desert trade AND {light manufacturing OR combat ornithopters}? I'm not sure.

Religion Influence is 1 for a holy city, and all the shines add 1 more. Influence is multiplied by the spread factor.

Click to expand...

Ok, thanks very much for the info. Wasn't intending to cast aspersions, I had just mis-remembered this.

This is something that should probably be scaled by game speed--e.i. increase the spread % at quick, decrease it on epic/marathon--so that you expect roughly similar observations regardless of game speed

Click to expand...

This is a very good point. I only ever played Vanilla on normal - on slower speeds, did the earlier religions tend to spread even further? I'm surprised they would have coded it that way.

Also:

Only part I'm not sure about in principle is moving the Wet Planet Conservatory, I see it a lot like the wierding room at the Arrakeen Palace (I always assumed that's kind of what it was supposed to be, just on a bigger scale) and is kind of a symbol of Imperial power in a way.

Click to expand...

I agree with this thinking in general and think it is reasonable to be at a political tech, but caste systems isn't a great fit, and more importantly I worried that moving the Imperial Estate building to Feudalism makes Feudalism a bit too strong, which makes it worth moving one of the civics (Hereditary Rule or Serfdom) to Caste Systems, which then makes it worth considering reducing the value of Caste systems by moving the Conservatory somewhere else.
But this may not be a big deal.

Well, I wasn't committing immediately to any of the exact balance decisions because you first have to see if the general idea is even feasible in how it shifts the tech tree around. There are as many visual and spacial requirements (or more) than there are logical ones involved in changing the tech tree because ultimately it all has to go onto the tech screen with each tech easily showing what you need to get it. You can't cross OR lines, you have limited space to fit techs in, etc.

I spent several hours last light playing with this and I'll just say, it creates some difficulties in how it shifts the rest of the tech tree after it. That's the biggest problem in my view. Looking at the tech screen I posted above for guidance, you can't place OR arrows from either DuneTopo or Light Manufacturing to a Great Houses above or below Feudalism because it crosses with the OR arrows going to Feudalism. I got by that by having Education be the OR for both, with Light Manu and Dune Topo as ANDs for Great Houses.

But after that the way it shifts the techs in the next column or two forces you to redo (visually) the whole rest of the middle of the tech tree. I found a way to make it all fit, but it's not as visually easy to read as before in my opinion. One thing I really like about the current tech tree is how through the whole middle of it, nearly all the military units are along the top, economy/science through the middle, with ecology on the bottom. The revised tree I was able to make certainly had to make some compromises on that aesthetic to fit everything in.

If you've got the time, you might want to try your hand at it and see if you can make something work.

because you first have to see if the general idea is even feasible in how it shifts the tech tree around

Click to expand...

Fair point. This does take some trial and error.

On a personal level, I find the aesthetics less important than the balance. The Civ4 system is nice in that it lets you add pre-reqs without needing lines for more than one of them.

I'll see if I can come up with anything workable.

The main problem seems to me to be that if you have Great Houses and Feudalism adjacent then it is tough to have GH allow Spice Indiustry and Protected trade while still Feudalism as the pre-req for Caste systems.

Yep, and if you want to see what I did, I can post the xml, but I'd rather get a completely fresh take on it and compare afterwards.

This is where I got to:
Feudalism requires Education.
Great Houses requires Light Manufacturing (and Desert Trade).
Harsh Conditioning requires Feudalism (and Fanaticism).
Law of Arrakis requires Divine Mandate (OR Feudalism) (note: though, are we forced to have connection lines for all OR requirements? If so, that is a probelm.
Spice Industry requires Great Houses (and Desert engineering)
Protected Trade requires Great Houses.
But then we want: Caste systems requires Feudalism. Is it possible to force the requirement without drawing an arrow connection, or to allow any line crossing at all?
Alternatively, we could have Caste Systems require Great Houses (or Feudalism) and just show a line for the former, if we don't have to connect all OR requirements.
But then I worry we're making Great Houses too much of a bottleneck.

We could have Caste system require Great Houses and Protected Trade require Feudalism, but that is less of a good thematic match (ideally political religion/political tech should go together, and economic religion/economic tech).

though, are we forced to have connection lines for all OR requirements? If so, that is a probelm.

Click to expand...

Right now, yes, all OR prereqs draw lines. Could I create a tag that functions just like OR requirements but doesn't draw arrows, yes. But I'm afraid it'd be confusing for players if part of the OR prereqs have arrows and part don't.

Line crossing makes it very disorganized looking and will often overlap arrows going in different directions making it difficult to see what leads to what. I wouldn't ever do it.

However, if there is only one requirement for a tech it doesn't matter if it's an AND or OR(well, almost doesn't matter, I'll come back to that), you'll need it, so for instance we could get away with making Caste System have only a Feudalism AND prereq, so no lines going to it.

The only actual difference in such a case is the OR prereq will give a +20% research modifier where the AND won't (kind of an odd quark of the system), so if we want to keep the real cost of the tech constant, the defined 'xml' cost of the tech has to come down by approximately 17% (16.666% to be exact) When removing an only OR prereq. Just something to keep in mind.

but that is less of a good thematic match (ideally political religion/political tech should go together, and economic religion/economic tech).

Click to expand...

This. We definitely want to keep that logical flow in tact.

But I'm afraid it'd be confusing for players if part of the OR prereqs have arrows and part don't.

Click to expand...

It is far from optimal, but might not be too bad, if (and only if) you could still include the requirement icon on the tech label (as gets done with non-linked AND requirements).

Line crossing makes it very disorganized looking and will often overlap arrows going in different directions making it difficult to see what leads to what. I wouldn't ever do it.

Click to expand...

The only line crossing I had in mind would still be pretty clear:

so for instance we could get away with making Caste System have only a Feudalism AND prereq, so no lines going to it.

Click to expand...

That seems potentially workable.

The only actual difference in such a case is the OR prereq will give a +20% research modifier where the AND won't (kind of an odd quark of the system)

Click to expand...

To be clear; you get a 20% research cost reduction for a tech that is an OR requirement for a tech you already possess? Am I interpreting that right? So if A or B are needed for C and I want them all, then it is cheaper for me to go A-C-B than A-B-C?
Embarrassingly, I didn't actually know that.

Ahriman said:

It is far from optimal, but might not be too bad, if (and only if) you could still include the requirement icon on the tech label (as gets done with non-linked AND requirements).

Click to expand...

In fact, I think this is the only possible way that we could get GH and Feud to be adjacent and give the tech allowances we want - unless we considered totally re-arranging the rest of the tech tree too (ie more than just these 4 columns).

We'd just need to decide if the slight aesthetic blemishes were worth the increased gameplay value (*edit* increased gameplay value... of changing from the status quo).

Well, the problem is it won't draw it like you did, it'll draw it like this:



I'm looking down the line, and it might be feasible to swap Caste System and Protected Trade (and the techs after them in the tech tree of course) so that we can group the two Great Houses 'derivative' techs together and beat this issue.

It is far from optimal, but might not be too bad, if (and only if) you could still include the requirement icon on the tech label (as gets done with non-linked AND requirements).

Click to expand...

While I'd be very cautious using this much, I'm looking at doing this for Law of Arrakis(let's call it LoA for the moment), as I'd like it to be reachable from any of Divine Mandate, Feudalism, or Great Houses, but you can't draw lines from all three. It's too important of a tech for expansion to be hitched to only Feudalism or Great Houses, it should be reachable from either to keep them roughly equally viable options (where the decision is based on your strategic plan). If for instance it had OR techs for Divine Mandate OR Great Houses only, It would make Great Houses somewhat better than Feudalism just because it leads to LoA.

To be clear; you get a 20% research cost reduction for a tech that is an OR requirement for a tech you already possess? Am I interpreting that right? So if A or B are needed for C and I want them all, then it is cheaper for me to go A-C-B than A-B-C?
Embarrassingly, I didn't actually know that.

Click to expand...

I think you got your letters backwards, but yes, for every OR tech you have, you get an additional 20% research mod (I'll come to that) towards that tech.

Research mods are something not shown in the interface, so most players are blissfully ignorant of how they work or when they are working for you, but it's also one of those things that shouldn't be explicitly shown as it breaks immersion a little bit and would encourage a more min/max play style. For the average player it's enough to know that having additional ORs does make it quicker to research (I've known this principle for what seems like forever as I know RevDCM at least has help text when hovering over a tech saying it speeds up another, is that in DuneWars?).

For the sake of balancing though, it's good to know the mechanics. A research mod does NOT affect the COST of a tech, instead it's a multiplier to a player's current research/turn. So if you are producing 100 research, a +20% research mod means you actually apply 120 research to the current tech that turn. So a tech with a cost of 5000 takes 50 turns @ 100 research per turn, but 41.6667 turns with a +20% research mod. That's where the cost reduction to keep them equal required above comes from:

1.0 - (41.6667 / 50 = 0.8333) = .166667

OR

16.6667%

Tech Diffusion also acts on a player's research modifier, adding to it based on how many players they know that have the tech already (with a bonus if that player has open borders) and how far behind they are in techs compared to the tech leader.

Now, the BetterAI merge gives us a GlobalDefines setting that differentiates between the first OR tech and each additional OR tech. This means we can set the first OR tech to not give a research mod bonus, but at present it's set to 20% for the first and for each additional, and I've avoided changing that as it affects the research balance considerably from what we are used to seeing since almost all techs have at least one OR prereq. It would in effect slow down tech progression.

But it is rather illogical that setting one OR tech makes a tech cheaper than giving it an AND tech. The mechanic should work in principle to encourage you to seek additional ORs before researching an expensive tech if you can put it off, but giving a bonus for the first OR doesn't make a lot of sense, nor do players realize at all that that's happening. The good news is that despite the fact that it's illogical, it's a hidden mechanic that doesn't really break anything, so it's not really imperative that we change it either.

the problem is it won't draw it like you did, it'll draw it like this:

Click to expand...

Ah, understood.

I'm looking down the line, and it might be feasible to swap Caste System and Protected Trade (and the techs after them in the tech tree of course) so that we can group the two Great Houses 'derivative' techs together and beat this issue.

Click to expand...

That would be an excellent solution. I don't have the mod on this PC, so I didn't really look outside the screenshot.

While I'd be very cautious using this much, I'm looking at doing this for Law of Arrakis(let's call it LoA for the moment), as I'd like it to be reachable from any of Divine Mandate, Feudalism, or Great Houses

Click to expand...

I think I would limit LoA to just Divine Mandate (religion line) OR Feudalism (governance line).
Otherwise, Feudalism ends up being too unimportant and you can just ignore it and go with Great Houses too easily.
Great Houses in the design is needed for Spice Industry and thus Offworld Trade, which is really key, and it already leads to good economy boosts. Feudalism also needs to be a requirement for something important.

I think you got your letters backwards, but yes, for every OR tech you have, you get an additional 20% research mod (I'll come to that) towards that tech.

Click to expand...

Ah, I didn't get the letters backward, instead I had misinterpreted the mechanic.
Ok, so you get a bonus on the advanced tech for having multiple pre-reqs, rather than having a bonus on the pre-req if you already have the advanced tech.

In terms of balancing, there is a lot of scope for changing tech costs. The one thing I would be very careful about though is that the current layout is for graphical convenience. Techs in the same column are *NOT* necessarily intended to be the same tier. There have been a lot of situtations where a high-tier expensive tech has been shuffled left for graphical experience, but it remains expensive, and in general should stay that way.

since almost all techs have at least one OR prereq

Click to expand...

Are you sure this is true? From memory in the tree, there are relatively few OR requirements, nearly all the requirements are AND requirements.
I think it is fine to leave the mechanics for now, and then rebalance any particular techs that feel like they have issues. I think it is an ok hidden mechanic that mildly punishes beelining, and there are so many profitable beelining opportunities in this mod that a little punishment is probably good for balance.

I think I would limit LoA to just Divine Mandate (religion line) OR Feudalism (governance line).
Otherwise, Feudalism ends up being too unimportant and you can just ignore it and go with Great Houses too easily.
Great Houses in the design is needed for Spice Industry and thus Offworld Trade, which is really key, and it already leads to good economy boosts. Feudalism also needs to be a requirement for something important.

Click to expand...

That makes sense, we'll try it. I'm going to try making the Caste System/Protected Trade tech swap later today and see how it comes out.

Techs in the same column are *NOT* necessarily intended to be the same tier.

Click to expand...

I wasn't planning on changing the base cost of any techs (unless there are specific cases where our changes dictate it's necessary for balance in it's new position), so no worries there

Are you sure this is true? From memory in the tree, there are relatively few OR requirements, nearly all the requirements are AND requirements.

Click to expand...

They may function as an 'AND' because there is only one OR prereq for a tech, but any prereq with a line coming from it is coded as an OR, so gives the research mod bonus, even if - being the only OR - you absolutely need it. Count how many actually have no lines to them . That's the point I wanted to get across, nearly every tech gives a defacto +20% research mod when researching it because it has at least one OR prereq.

Therefore, if we restricted it to only give a research bonus for ADDITIONAL OR prereqs (as would be logical), we'd be essentially increases the length of time required for almost all techs by removing that defacto research mod bonus for having one OR tech (which would have to be compensated for somewhere else).

I want to stress that this is by no means a necessary change I'm advocating, things are balanced with it, so no need to buck the system. I just raised it so you know what the underlying mechanical difference is between using an OR or AND prereq.

but any prereq with a line coming from it is coded as an OR, so gives the research mod bonus, even if - being the only OR - you absolutely need it.

Click to expand...

Huh. Did not know that. How bizarre.

Thanks for letting me know how this works.

Proposed Tech Tree


Despite the visual changes from swapping Protected Trade and Caste Systems, nothing has really changed beyond that point, so here's the highlights:

• Feudalism is disconnected from Fanaticism, otherwise remains the same
• Great Houses requires:
  • Light Manufacturing
  • Desert Trade
  • Arrakis Habitation
• Harsh Conditioning may look different because I couldn't connect any lines to it, but it still has the same prereqs, the cost has been adjusted down to account for no OR prereq bonsuse (so essentially stays the same)
• Suspensor Devices also may look like it's in a different spot, but also retains the same prereqs (just switched the AND and OR) and cost

I did have to cross some lines in two spots, but I think in both cases the path is clear and obvious, which is the goal. It ought to be balanced where Feudalism is the quicker to found typically, but it's perfectly possible to get to CHOAM before or not long after too some games.

I didn't feel the need to move around the civics/buildings/wonders for any of the techs in question. Since it's now possible to go straight to Caste Systems from Feudalism, it'll be easier to get missionaries and 'temples' up for it quicker after founding if a player prioritizes it. All in all this seems like a pretty elegant solution, though it still needs to go through some testing. I'll probably put out beta in the next day or two for those interested. I've also got some new AI work (related to tech choices/valuing), game speed scaling for religious spread, great nobleman/merchant for founding Imperial/Choam with the political trait, and a few odds and ends to release as well.

Proposed Tech Tree

Click to expand...

Looks good. Nicely done!

I'll try a new version if you put it up soon.

A potential warning: I am getting married in a week, so I'm nearing a point where I will be off for a month or so.

Ahriman said:

I am getting married in a week

Click to expand...

Congrats!

@ Chris : The way the game works is 100% improved. However, after a few games, I think the AI bonuses to unit maintenance need to be increased SLIGHTLY (on Emperor and probably the higher levels as well). If the AI had a few more "free" units (total gold spent on units = an extra few units) I think that would be better for game play. In other words, unit spam can be increased a little as long as the tech rate doesn't suffer. Keep up the awesome work. Fine tuning is always a biatch.

I think the AI bonuses to unit maintenance need to be increased SLIGHTLY (on Emperor and probably the higher levels as well).

Click to expand...

Sounds plausible, but: is your goal to get the AI to build more units, or to advance its science faster?
If the former, then unit maintenance changes might not do that, I think we would need Chris's specific understanding of what the right handle is. If the latter, is unit maintenance the right lever, or should we boost science or economy directly?

Bringing the unit maintanence down is the most direct route if you want them to be able to field a few more units, and a 5% shift down from Emperor onward wouldn't be unbalancing.

I'll bring up something else I've been thinking on for a while. What I'd like to do is create a new game option (as in, has a check box in create custom game) called something like 'Adaptive Difficulty'.

Now I'm not talking about the crazy 'if you get near the lead, we'll bump up difficulty level till you reach Diety' everytime FfH mechanism (man that annoyed me!). Instead, what I'm thinking wouldn't change the 'difficulty level', but instead would give additional passive bonuses to the AI based on how well the player is doing. The idea revolves around the fact that there are good starts and there are bad starts, and the game is generally balanced around those middle of the road starts. But when you get a really good one...

Basically this would give bonuses to things like city maintanence, tech costs, unit costs, etc. based on the relative score of the human vs. some algorithm of the average AI scores (probably ignoring any 'outliers') to start to give a soft ramp up as the player begins taking a dominating position, then if he's still able to continue to put up more and more score compared to the AIs, the bonuses become more and more severe. This would go a long way towards balancing out those really good starts compared to more normal ones (without resorting to setting the difficulty level much higher, which ends up precluding lesser starts).

It's important I guess to note that the goal of such an option would never to be to make it impossible or frustrating for a player to ever get into a good leading position. Instead it's goal is to make sure the AIs don't become too much of a push over when you are doing well in order to keep enough challenge to keep the game fun. I think we've all been there where you've done too well and the rest of the game just feels like mop up time.

Also, since it would never change your actual original 'difficulty level', that difficulty level's handicap bonuses act as a starting point for any additional adaptive difficulty bonuses. Effectively therefore that original difficulty choice still acts to set what 'maximum' bonuses the AI could end up with (original_handicap + max_adaptive_handicap = max_total_handicap for any particular area).

edit: Oh, and Ahriman, congrats

The idea of "adaptive difficulty" is discussed for many different games. If you have an option to *decrease* the difficulty level if the player is doing badly, I think it dilutes the achievement of actually winning. The game is giving you a break when you mess up.

I think *increasing* the difficulty has a similar problem. If you find the game too easy at a certain difficulty level, you should win this one quickly and play again at a higher difficulty. "I just barely beat the game on Chieftain level" would be a possible outcome for both a good, experienced player and also a poor, beginning player. In a more quest-related game, where you only play through once, I can see adjusting the difficulty; but for a game which you replay many, many times I think holding the difficulty level constant is a better approach.

In FFH, and other games, they have discussed mechanics for faster wins. Once you get into a dominating position, mopping up other AI's who have no chance just burns time. So either, adjusting the victory conditions to make mopping up less necessary, or adding this type of "faster win" mechanic can help.