Diplomatic GUI

[LDiCesare]

One thing I've been thinking about: It should be possible to trade units (give units as in civ2, and, apparently civ4, or galciv), but I think it should be possible to trade units for a limited period, thus have mercenaries. This would be quite natural and easy to do in the negotiation screen.

[Lord God Jinnai]

I think that might be better handled by the ruler paying a lot of money of front and making the units free maintance, but have them live off the land and what they make.

Else we could just have a hidden mercenary civ from the start that advances. its who point to keep track of technology level and abundance of potential mercenaries in an area and if there are too many they can also become bandits or pirates. You might also be able to have this keep track of merchants as well.

I dunno just a thought.

[LDiCesare]

My point is not to simulate real-world mercenaries but more to allow the player to lend his troops to another civ. You can give them outright (for a payment) or lend them. This makes sense in the dipomatic gui.

[LDiCesare]

Here's a screenshot of the current state of the ui.
It lacks a statusbar and buttons in the lower part.
I'll probably try to prevent the selection of a line making it blue. The middle pane also needs some tuning, and scrollbars appear out of nowhere for the moment... Anyway here it is.
I may show the available amount of money in that screen too. I don't know yet how I'll display the gold/gold-per-turn dialogues. I've even less of an idea of how to allow tech-trading, or lands. Units would be easier (listing task forces), though I don't want to show the list of units of another civ unless I hae some agreement with them.

[LDiCesare]

I've added right of passage agreements. Stilll have to display them.
The way I handle this is:
If you have a right of passage (rop) with someone, this civ allows your troops to cross their civ. Note rop is not 2-ways. You can force another country to accept your troops and not allow them to send their troops to your country.
For the moment, if you set foot on someone else's territory and don't have rop, they'll declare war immediately. Actually, they should first check whether the war canbe won, and if so declare war. Otherwise, they'll contact you to propose a right of passage, and if you refuse declare war.
The ai will have to check rop's when planning its moves.
Next step will be ai/player and ai/ai diplomacy.

[Mark_Everson]

Sounds Great Laurent! Looking forward to checking out the new version.

[alms66]

I just wanted to let you know, Laurent, that I have now played GalCiv and I can understand why you like it. The knowledge of acceptance is immediate - much better than civ3 in that regard. But at the same time, it is much worse than civ3 in the tweaking department. I found myself tweaking 1 or 2 influence more dozens of times per deal. I hate that. That's why I want the system I described earlier over one like GalCiv or Civ3 - one where the breakpoint is concealed from the player and doing such tweaking can only lead to madness, that way I and many others won't even attempt it.

[LDiCesare]

I will try to get a tradebar implemented. The problem I face with this is that some deals do not have a straightforward value, like declaring war or making peace. I don't know whether it will be confusing to show the value the ai gives to peace for instance.
Typically, let's consider a war. I as Hannibal propose to end the war against Scipio. Scipio values peace at, say, 1000 gold. If I want to show this in the tradebar, do I show 1000gold from my side to Scipio? I could also show the value I attach to the peace (which would probably be something like -11000 if the ai controlled Hannibal) but this doesn't make much sense I believe (it means things like "you have more to win than they have to lose" which are only useful for decision-making by the ai).
The ai currently goes for a deal when it believes the net gain is >= 0 (or > 0, can't remember, it doesn't change much). Deals are possible, in theory, between ai's if they value the same thing as different costs. I'll have to add several more heuristics, like relationships (friendly, allied, etc should let you accept bad deals or give gifts), and other considerations (like if I trade for X at a loss with Y I can trade X for something else with Z and have a net benefit).

[Mark_Everson]

Good points Alms, we Really need to avoid micromanagement of the sort that you bring up.

Originally posted by LDiCesare
Scipio values peace at, say, 1000 gold. If I want to show this in the tradebar, do I show 1000gold from my side to Scipio? I could also show the value I attach to the peace (which would probably be something like -11000 if the ai controlled Hannibal) but this doesn't make much sense I believe (it means things like "you have more to win than they have to lose" which are only useful for decision-making by the ai).

I think that most of the playtesters at this stage would find both values interesting. There is also a micromanagement-reduction aspect. If I'm the Romans and I propose peace to Hannibal, I will see that I need to give him a HUGE offer to end the war. Seeing that, I will not waste time going to the negotiation screen every time something changes hoping that the balance has shifted.

The ai currently goes for a deal when it believes the net gain is >= 0 (or > 0, can't remember, it doesn't change much). Deals are possible, in theory, between ai's if they value the same thing as different costs. I'll have to add several more heuristics, like relationships (friendly, allied, etc should let you accept bad deals or give gifts), and other considerations (like if I trade for X at a loss with Y I can trade X for something else with Z and have a net benefit).

The AI should also take risk into account when startin wars. If there is a likely chance of very small gain, but a possibility of ruin for starting a war, then it is irrational to start the war under most likely levels of risk aversion. For that reason, until we have a good way to evaluate risk, I'd put a threshold of something like a few percent of my civ's value that needs to be exceeded. That would obviously need to be tweaked to allow picking off very small opponents that have no support from other major powers.

[LDiCesare]

The going to war code takes into account the enemy strength (thus risk). The problem is it doesn't remember armies that went away from a visible square into the fog of war. I must code that knowledge and some decay for it so the ai will not attack when it just saw a big force but may be tempted to do so after a while.
I also made provision for preventing a civ from entering too many wars at once. A civ shouldn't enter a war if it's already waging war unless it's guaranteed to be an easy win that doesn't divert too many resources.
I know that in other games, Galciv and CivIV for instance, there is a global military rating available that simplifies the risk evaluation a lot, but I think it's a bit ridiculous if you can't explain how that information came to be known.

[alms66]

Originally posted by LDiCesare
I will try to get a tradebar implemented.

Perhaps it's better to wait for the next demo? I've always been one to push for more and more in the demos, but this is one of those things that I think is better done in one full shot, sort of like you did with reworking the riots model. I would finalize what you have working now (peace and money transfers), then finalize the demo, and only after the next demo is out, turn your full attention to diplomacy, trying to tackle as much of it in one go as possible. Regarding the rest of this paragraph, I agree with Mark's comments above.

Originally posted by LDiCesare
The ai currently goes for a deal when it believes the net gain is >= 0 (or > 0, can't remember, it doesn't change much). Deals are possible, in theory, between ai's if they value the same thing as different costs. I'll have to add several more heuristics, like relationships (friendly, allied, etc should let you accept bad deals or give gifts), and other considerations (like if I trade for X at a loss with Y I can trade X for something else with Z and have a net benefit).

Again, I'd wait on this until the next demo. I also have some comments I'd like to make on coding the diplomacy AI, which I'll post later today. I suppose I'll put those in the "coding the diplo model" thread - which is around here somewhere.

[LDiCesare]

I'm wondering how to manage technology exchanges.
Knowledge of applications require several tech levels. The tech levels are quite abstract, so exchanging techs in the Civ sense doesn't work. Exchanging +1 in farming for +1 in mathematics sounds like accounting and not very thrilling.
Providing knowledge in an area (teaching plan) should be possible, and providing knowledge of the techs to get an application should also be possible, only if the tech levels are not too far removed. The tech exchange should probably result in additional research points rather than moving the tech levels up, too. But how to present that?

[alms66]

I think for Clash, tech "trading" is definitely out of the question. What can be done, is to have research treaties. There need to be two basic types of treaties - "gift treaties" and "partnership treaties". There could be treaties of the big, broad categories (communication, transportation, etc), down to treaties for a single application-enabling tech (gun making, shipbuilding, etc.).

For gift treaties, in order to set "ending" conditions so that a tech doesn't continuously grow forever with a single treaty, I suggest having to set a slider (kind of like setting a gpt value for cash deals), that represents what level you are willing to give the tech up to. For example, by setting "communication - 50%" you are saying that the treaty is active until the other party has reached 50% of your communication tech level - at which point, it expires. If the condition is never reached, the treaty never expires automatically.

I also suggest not giving bonus levels (+X) via these gift treaties, but rather extra RP's in whatever field the treaty relates to. The amount of extra RP's given should be determined, not only by the difference in each party's tech level, but also by the percentage set above. The larger the gap of tech levels and the larger the percentage set, the faster the tech level grows.

For partnership treaties, simply apply a bonus of 25% (or, alternatively, implement several different levels of cooperation treaties each giving different bonuses) of the RP's produced by the other parties each turn. So, for example, if CivA and CivB have a "Communication Research Partnership Treaty", CivA receives 25% of the RP's produced by CivB in the Communication field, while CivB receives 25% of the RP's produced by CivA in the same field. To put hard numbers on it, let's say CivA produces 3000 RP's and CivB produces 4000 RP's each turn. As a bonus to each turn while the treaty lasts, CivA receives an additional 1000 RP's, while CivB receives an additional 750 RP's.

And, finally, to prevent runaway tech growth via such treaties, each additional treaty signed, after the first, is halved in effect.

[Mark_Everson]

Good points all. I agree that the teaching or trading approach seems best given the way that we do tech. The best way I can think of to present limits or goals to the player in a meaningful way is to phrase it in terms of key applications. "We will help you in chemistry and gunpowder until you reach Muskets". or "We will help you in seafaring until you can build Caravels." The player would be presented with" It will take X turns (or years) to get to Muskets with current spending, and Y turns with the teaching / trading treaty.

[alms66]

Originally posted by Mark_Everson
"We will help you in seafaring until you can build Caravels."

This is the same as "seafaring - 100%", just to clarify my post a little more.

But, anyhow, I came back because I wanted to add a little more (I thought about it and realized I never explicitly answered your question of how to present it to the player).

So, basically we have the tree structure (as pictured in post #94 above), with Science (or technology seems like the better term, IMO) as one of the branches. Beneath that branch we have the major Tier 1 Tech categories (communication, transportation, etc.), beneath those are tier 2 techs, beneath those are tier 3 techs, and so on. This will allow you to select any given tech at any level in the tree and put forth a treaty with it - giving great flexibility. To allow an entire branch (say all communication techs) to be put into a treaty though, there needs to be an "All" option, probably right at the top of the list, so that you might have the following:

-Technology
      -Communication
           *All Communication
           -Tier 2 Communication Techs
                *Writing
                *Other Tier 2 Comm. Techs...
           -Tier 3 Communication Techs
                *Alphabet
                *Other Tier 3 Comm. Techs...
      +Transportation
      +Military
      +Science
      +Social
      +Infrastructure

The example above, shows each expandable node as either "+" or "-", with the player selectable options for placement on a treaty preceded with a "*". I'm sure you can take that and extrapolate from there if you go this route.