Diplomacy

I wish I could change AI willingness to trade huge sums of gold for cities. I think the loss in happiness from getting a weakened city is already enough of a penalty, and losing gold too is just even more of a hit. I avoid it since I feel it exploits misjudgements on the part of the programmer who set those numbers.

Thalassicus said:

I wish I could change AI willingness to trade huge sums of gold for cities. I think the loss in happiness from getting a weakened city is already enough of a penalty, and losing gold too is just even more of a hit. I avoid it since I feel it exploits misjudgements on the part of the programmer who set those numbers.

Click to expand...

I do it all the time, and here is how I rationalize it!

First, the AI usually trades surprisingly little gold for most cities. The offers shoot up when either the city is a loaded metropolis, or (interestingly) when it's the second city you sell them, and it's adjacent to the first. Given the value of a serious position in either case, and how cheap gold is for the AI on the levels most of us play at, the cities are not rip-offs.

The only exception to fair-value is when you offer to sell a civ one of their own cities. In these cases they often (but not always) agree to pay an inflated premium. To me this makes sense from a patriotic and cultural perspective, but agree that it may be an edge we don't need.

In my case, selling cities is also part of puppet-mastering, something I really enjoy in Civ.

That makes sense... so it's only when we re-sell cities that they give larger sums of gold.

Thalassicus said:

That makes sense... so it's only when we re-sell cities that they give larger sums of gold.

Click to expand...

I think they pay similarly if you sell one of your own cities. But yes, they only pay absurd amounts when it's one of their own cities... and then only sometimes.

There's also the expected discrepancy between expansionists and isolationists.

Thalassicus said:

New Method
Each turn +1 point is added to a militaristic pool for players with militaristic friends or allies. We get a unit whenever our points pass the threshold, and the pool resets to 0. If we have 9 points and our threshold drops to 7, we get a unit on the next turn with 2 points' overflow.

This approach has several advantages. Say we have 9 points stored and the threshold is 10, but we lose our militaristic friends. If we later get militaristic friends again, the pool is saved so we'll get a reward on the next turn. Points never go to waste.

Click to expand...

This whole design with a threshold number of turns still seems a bit confusing.

Here is how I would do it:
Suppose we wanted an alliance to give a unit every 15 turns, and a friendship to give a unit every 30 turns.
Then, I would create a pool. It starts at zero. For each ally I have, I get 2 points per turn. For each friendship I have, I get 1 point per turn. On any turn on which the counter is at least 30, I get a new unit and subtract 30 from the counter.

These methods are *almost* equivalent, but I think mine has fewer potential rounding issues and is arguably clearer. It works better to have a threshold as being the actual number of "points", rather than a number of turns.

Then, display to the player their current counter (like the GPP counter), their incremental points per turn, and the number of turns needed at the current rate.

So for example, if I had 1 alliance and 1 friend, and I had 5 points already, then the message would read:
"5/30, +3 per turn, 9 turns until next unit."

In the "new method" above then I would be in an identical situation whether I had 3 points, 4 points or 5 points, because all of these would take 9 turns and then reset the threshold to zero, but in my method, you are better off with 5 points already accrued than you are with 3 points, because with 5 points after 9 turns you'll be on 32 points, which will give you the new unit and then drop you down to 2 points, whereas the "new method" would give you the new unit and drop you down to 0 points.

If 15/30 turns aren't the target, then just change the thresholds and the relative ratios.
For example, if you wanted 12 turns allies and 18 turns friends, you could have allies give 3 points, friends give 2 points, and have 36 as the counter threshold, and subtract 36 when a new unit was formed.

This way, the threshold number of turns is clearly a function of the actual primitives, rather than a variable in its own right.

Thank you! I was trying to figure out a way to invert it to a static threshold and variable rate. The reason the current version has static rate and variable threshold is I adapted it directly from what Firaxis does... their method changes the threshold. I've made this change for v106.2 beta.

I dunno about displaying the details of militaristic civ unit giving to the player. I thought the idea that you don't know exactly when you get units is integral part of the whole process.

If you'd argue that due to deterministic effect, the player could know anyway by crunching the math since last unit, well, that's a bit torturing yourself info addict gaming to me. Personally I'd furthermore prefer a random component to how many points are added to the unit donation pool in addition to not knowing the turns left. (Total variance of unit donating time will still be small, unless you put a huge random component to points added each turn. You're summing up random components there, so variances diminish.)

I thought the idea that you don't know exactly when you get units is integral part of the whole process.

Click to expand...

Lack of transparency is never a good thing. The player should always know how the mechanics work, and should be able to do as much long-term planning as possible.

Personally I'd furthermore prefer a random component to how many points are added to the unit donation pool in addition to not knowing the turns left

Click to expand...

Why? What does randomness add? We already know exactly how many culture or food we're going to get, and when every other building or unit we get is going to be completed.

Ahriman said:

Why? What does randomness add?

Click to expand...

Randomness adds a crucial component to strategy: margin of error. Just like it does in combat, for example. With some randomness you need to play somewhat safe, though just how much risk you tolerate is also a strategic choice.

The need for safety margins also gives the AI a break: Otherwise the human can always game to win it just with the skin of his teeth. In this case the lack of safety margin would be that that you know exactly how much gold you need to give to get from the CSs #N units. I think that's rather absurd. You need to strategize in more wide, abstract ways: "I be friends with these people, and I get units about so and so". Exact optimization, as done by Excel sheets, isn't strategy - at least not the fun kind in my book.

tlaurila said:

Randomness adds a crucial component to strategy: margin of error. Just like it does in combat, for example. With some randomness you need to play somewhat safe, though just how much risk you tolerate is also a strategic choice.

The need for safety margins also gives the AI a break: Otherwise the human can always game to win it just with the skin of his teeth. In this case the lack of safety margin would be that that you know exactly how much gold you need to give to get from the CSs #N units. I think that's rather absurd. You need to strategize in more wide, abstract ways: "I be friends with these people, and I get units about so and so". Exact optimization, as done by Excel sheets, isn't strategy - at least not the fun kind in my book.

Click to expand...

I don't think many here would argue with your overall point. It's all about the specifics. For example, do you think it's fair to know exactly how much gold you need to buy an alliance? I think it would be unrealistic not to know.

I don't see why margin of error is an important thing to have in terms of when you get military units. I can't see any obvious way to game the system, and whether you get a unit this turn or next turn doesn't make a crucial strategic difference, it is just useful to know and to understand how the mechanic works.
There is already some uncertainty; you don't know what unit you will get.

As Txurce says, it would be very frustrating to not know how much gold you needed for a CS alliance, or how many turns it would take for you to research a particular tech, or how many turns it would take to construct a particular building or unit.

And as for combat: I agree that having multiple outcomes from a fight is useful, but as it happens I find the lack of transparency annoying. Ideally I would know the probability distribution over the various possible outcomes. I think people are working on UI mods that would display this - as they did for Civ4, or as is possible in say Battle for Wesnoth.

Some (small) randomness would be a secondary thing, anyway. My main point was that in terms of gameplay, there is such a thing as "too much under the hood info to the player", and exactly telling the player how many turns for CS unit donation is such. CS units are, by design as it seems to me, a more unreliable way of getting units compared to producing them yourself. And I don't think the unreliability involved is a design flaw or a bad thing, it actually adds to the game. Of course some knowledge of the process is needed by the player, so that he can make an informed decision whether it makes sense to donate to CS, or to directly purchase units. In my opinion the CS should yield about the same units/turn (plus other CS friendly/allied benefits), but direct purchase gets the units that you want when you want them.

Oh, and probably you guys remember Master of Orion 2 research, where after a set number of "beakers" was achieved, you started to have a probability of inventing the tech you were shooting for. As more beakers beyond the threshold was accumulated, that probability went up, and of course the agglomerate probability to achieve the "breakthrough" went up geometrically. It didn't change the game much, because the odds cancel out to average in the long run, but I think it was a nice touch. So in research you don't need to know exact number of turns, in fact it's sort of silly that you do ("In 10 years and 2 months our wise men will discover the wheel, my liege!"), but you do need to know magnitudes of expected times, and relative times between different techs, to make strategy decisions.

In alliance, yes you should know how much gold you need for alliance, since that should simulate more like a negotiation. The way it superficially works, that you give them gold and hope for the best, is sorta silly. Of course you'd do a negotiation and agree on prices. Which you can think happens when you know the mechanics, you can decide whether you want to invest enough to make allies.

The unit reward itself is still random. I think this adds enough unpredictability to it, without needing a random factor on when the unit is rewarded.

and exactly telling the player how many turns for CS unit donation is such

Click to expand...

Why? I didn't see an argument as to why number of turns is something that needs to be random. Other effects that are based on number of turns aren't random (eg: great person generation, unit/building construction).

And I would distinguish between two things; there are cases where uncertainty is valuable; eg a probabilistic chance of different outcomes occurring. I don't see that there is ever a case where opacity is valuable; where the player should be blocked from knowing what the probability distribution will be. The case you made was for uncertainty, not for opacity.

Oh, and probably you guys remember Master of Orion 2 research, where after a set number of "beakers" was achieved, you started to have a probability of inventing the tech you were shooting for.

Click to expand...

I hate that kind of mechanic. I don't remember Moo2 working like that... (Moo3 maybe?) but it has been a long time since I've played it.

So in research you don't need to know exact number of turns

Click to expand...

Actually, yes, sometimes I really really do want to know. I want to know when exactly I'll be able to upgrade to a new unit, or when a wonder or building will become available for construction.

Ahriman said:

I don't see that there is ever a case where opacity is valuable

Click to expand...

Here we differ greatly in opinion. Incomplete information all round is a crucial part of the "fog of war". It makes strategy that much more interesting, varied, and less like playing Excel, when it extends to "Know yourself" as well as "Know your enemy". Obviously Sun Tzu didn't think you could ever have perfect knowledge of yourself when he advocated those two points

Ahriman said:

Actually, yes, sometimes I really really do want to know. I want to know when exactly I'll be able to upgrade to a new unit, or when a wonder or building will become available for construction.

Click to expand...

It's certainly beneficial for the player in an optimization sense to know. So that you can tweak your specialists just so that you finish the building you're doing in the exact turn you will have that new wonder available. To me this is precisely the kind of "Excel sheet" optimization and (micro)management that has very little to do with strategy, nothing to do with fun, and finally gives an unnecessary edge to the human vs the AI.

But eh, "fun" is greatly subjective.

Incomplete information all round is a crucial part of the "fog of war".

Click to expand...

Incomplete information about what the enemy is doing? Sure.
But not incomplete information about how the game works. I just find that frustrating.

I don't even usually do a whole lot of MM optimization, but sometimes when I can upgrade units makes a big difference, in the middle of a war, or if I am considering whether to accept an AI's "declare war on X in 10 turns please" request. Knowing that I will have rifling in 9 turns rather than 11 turns makes a very big difference here.

tlaurila said:

Here we differ greatly in opinion. Incomplete information all round is a crucial part of the "fog of war". It makes strategy that much more interesting, varied, and less like playing Excel, when it extends to "Know yourself" as well as "Know your enemy". Obviously Sun Tzu didn't think you could ever have perfect knowledge of yourself when he advocated those two points


It's certainly beneficial for the player in an optimization sense to know. So that you can tweak your specialists just so that you finish the building you're doing in the exact turn you will have that new wonder available. To me this is precisely the kind of "Excel sheet" optimization and (micro)management that has very little to do with strategy, nothing to do with fun, and finally gives an unnecessary edge to the human vs the AI.

But eh, "fun" is greatly subjective.

Click to expand...

I feel that any mechanic that I could theoretically work out on paper (I allied on this turn so I'll get a new unit this turn) should just be totally transparent in the game. I absolutely think that knowing when the next MCS unit is available is a great step forwards. Otherwise I'm either guessing or doing TONS of busywork to keep track manually.

tlaurila said:

Here we differ greatly in opinion. Incomplete information all round is a crucial part of the "fog of war". It makes strategy that much more interesting, varied, and less like playing Excel, when it extends to "Know yourself" as well as "Know your enemy". Obviously Sun Tzu didn't think you could ever have perfect knowledge of yourself when he advocated those two points


It's certainly beneficial for the player in an optimization sense to know. So that you can tweak your specialists just so that you finish the building you're doing in the exact turn you will have that new wonder available. To me this is precisely the kind of "Excel sheet" optimization and (micro)management that has very little to do with strategy, nothing to do with fun, and finally gives an unnecessary edge to the human vs the AI.

But eh, "fun" is greatly subjective.

Click to expand...

I agree with you completely, but that isn't CiV. This is why I also play other games that give me this type of thing that I enjoy.

Since you can calculate the exact number of turns until you get a unit, you should have it in the UI so you don't have to calculate it out by hand. The real question is: should we be able to calculate the exact number of turns until you get a unit?

My opinion would be no. I like the MTTH (mean time to happen) mechanics. However, I know other people do not. Furthermore, that would change the game to something that is not CiV, and that is not the objective of this mod. If you want that type of stuff, somebody will have to make another mod for it.

In v108.1 beta I implimented a nearly-final form of the open border / research agreement / declaration of friendship relationship.

Old version of mod:

• Mutual Open Borders: 1% of the combined gross of both players adds to your gold rate.
• Research Agreements: 5% of the combined gross of both players adds to your science rate.

New version:

• Mutual Open Borders: 1% of the combined gross of both players adds to your gold rate.
• Research Agreements: 3% of the combined gross of both players adds to your science rate.
• Declaration of Friendship: Increases income from Research Agreements and Mutual Open Borders with this leader by +50%.

This makes the research component depend less on the random luck of getting friendly neighbors.

Out of curiosity, what does this do to Korea's tech boost?