Beta 7.2.4 Comments

[Ahriman]

My prediction is that this will lead to a net gain for the peaceful approach, where one wasn't needed.

To be clear, I think that the peaceful player should potentially be more profitable than the warmongering one. But my preference would be to reduce the warmongering income further, rather than raise the amount of gold in the game.

Yes, this, exactly.

It also starts getting out of control on larger maps with more players, where you could easily have OB with 8+ other factions.

Also: if I have OB with 4 players, then aren't I actually getting more than a 10% bonus?
Assume that all 5 of us (A, B, C, D, E) have equal gold income.
If I get 2% of the combined income for each OB, then the combined income for each OB is 2X, so I am actually getting 0.04X per OB, so with 4 OBs I am getting 16% income bonus?
With 8 OBs I am getting 32% income bonus?
Which is much larger than a 32% increase in profits?

If you've coded it the way you model it here, then maybe you need to change the description:
"With mutual open borders, 2% of the combined gross gold income of both players adds to income per turn (income, not profit=income-expenses)."

You've also now made it a no-brainer to have open borders with every civ you encounter. It's just free money. Unlike DoF, there are no diplomatic downsides.

 

[SlightlyMad]

I'd argue that right now it's a "no-brainer" that you should never agree to open borders with anyone unless you absolutely must get through their territory. The biggest reason is that they use it to scout you out, and a stronger AI will actually try open borders before a DoW to see how well you're defended. They're also just annoying and block up your territory, preventing your workers from getting to tiles, and the like.

 

[Ahriman]

I'd argue that right now it's a "no-brainer" that you should never agree to open borders with anyone unless you absolutely must get through their territory.

I'd argue the opposite; at least until the late-game, it's a no brainer to sell your open borders for 40 gold to everyone. This is hugely valuable, especially in the early game.
The proposed fix would at least encourage you to trade OB for OB, which is something.

The biggest reason is that they use it to scout you out

I'm very skeptical that this makes much difference.

They're also just annoying and block up your territory, preventing your workers from getting to tiles, and the like.

I almost never encounter this.

 

[SlightlyMad]

So if you're already exploiting open borders for financial benefit, is your argument that you don't want the AI to be able to benefit from them as well?

edit: Sorry, I re-read that and it comes off as un-civil. What I mean to say is that I'd always prefer a system where the human and AI can benefit equally, over a one-sided exploit.

 

[Ahriman]

No, my ideal system would be one where open borders only provided open borders (and maybe a small diplomacy benefit), but could only be created as a mutual open borders treaty.
I think the whole idea of one-sided open borders was a mistake.

I don't like the status quo either. But I think the solution is to remove the ability to sell OB for gold, not to make it so that everyone trades OB for a gold bonus for everyone, all the time.

 

[Txurce]

So if you're already exploiting open borders for financial benefit, is your argument that you don't want the AI to be able to benefit from them as well?

You're not exploiting it - you're taking 40g in exchange for the advantages of traveling through someone else's lands.

No, my ideal system would be one where open borders only provided open borders (and maybe a small diplomacy benefit), but could only be created as a mutual open borders treaty.
I think the whole idea of one-sided open borders was a mistake.

To me, OB are like selling luxuries to an AI that doesn't need them - it's a part of the game that allows the human player to compete in an entertaining manner vs the AI's bonuses. While the better players don't need these bonuses, the game would be much less interesting if, for example, you couldn't sell luxuries if the AI didn't need them.

 

[Thalassicus]

In vanilla I always sign borders away for a free 30-40g, so that's a no brainer. Now I also request open borders with them, still a no-brainer. I haven't ever said this is designed to give strategic depth to border agreements -- we can't do that with the current modding tools.

I'd argue the opposite; at least until the late-game, it's a no brainer to sell your open borders for 40 gold to everyone. This is hugely valuable, especially in the early game.
The proposed fix would at least encourage you to trade OB for OB, which is something.

I think the whole idea of one-sided open borders was a mistake.

I agree with this entirely and it's the reason I did the change. Normally we wouldn't ever do AI border offers unless scouting territory for the first time, since it costs 30-60g and doesn't have much benefit. By giving a minor economic incentive to do the agreement both ways, it makes the game less annoying and exploitative.

The side benefit of a small passive gold bonus for peacetime is a nice plus too. Keep in mind that like DoF research, it's a way to help players who fall behind. A player far in the lead might get only a 2% bonus per agreement, while a person at the bottom could get much more. I offset the extra income with higher purchase costs for buildings and units. I also found it annoying that expensive and cheap units/buildings had such a huge variance in purchase cost modifier, so that solved another frustration too.

Spoils of War had no influence on these changes. I feel that policy is balanced well in the standard/emperor/continents+ conquest games I play.

Thanks for pointing out I was missing the "mutual" part from the ingame tooltip. This is the algorithm. It adds up all city, connection, and garrison gold for both players and multiplies by 0.02:

function GetOpenBordersGold(playerUs)
  local yield = 0
  local teamUs = Teams[playerUs:GetTeam()]
  local playerUsGold = playerUs:GetGoldFromCitiesTimes100() / 100 + playerUs:GetCityConnectionGoldTimes100() / 100 + GetFreeGarrisonMaintenance(playerUs)

  if playerUsGold > 0 then
    for playerThemID,playerThem in pairs(Players) do
      if IsValidPlayer(playerThem) and not playerThem:IsMinorCiv() and not (playerThem == playerUs) then
        if teamUs:IsAllowsOpenBordersToTeam(playerThem:GetTeam()) then
          local playerThemGold = math.max(0, playerThem:GetGoldFromCitiesTimes100() / 100 + playerThem:GetCityConnectionGoldTimes100() / 100 + GetFreeGarrisonMaintenance(playerThem))
          yield = yield + math.ceil((playerUsGold + playerThemGold) * GameInfo.UnofficialPatch["OPEN_BORDERS_GOLD_RATE_PERCENT"].Value / 100)
        end
      end
    end
  end
  return yield
end

You can see the income on the gold economy tooltip at the top-left of the screen. If you play a game and feel alters economics too much I'll increase purchase costs slightly more. I do want to maintain it as a decently powerful bonus to help players who fall behind.

 

[Ahriman]

You're not exploiting it - you're taking 40g in exchange for the advantages of traveling through someone else's lands.

The advantages of travelling through someone's lands are generally low, particularly in the early game when players do not control much land. Whereas getting extra gold to buy another building, worker, settler, CS alliance, etc. is very valuable indeed.

To me, OB are like selling luxuries to an AI that doesn't need them - it's a part of the game that allows the human player to compete in an entertaining manner vs the AI's bonuses.

I don't find it entertaining, it just feels gamey to me. Oooh, look, I can exploit this thing that the AI doesn't understand!

By giving a minor economic incentive to do the agreement both ways, it makes the game less annoying and exploitative.

I can understand that, but I worry a bit that the incentive is too large, particularly in the late game.

Keep in mind that like DoF research, it's a way to help players who fall behind.

Yes, I think this is an interesting effect, and a fairly decent anti-slippery slope mechanic.
I do worry a bit though in that everyone's research is somewhat proportional to their power, but gold isn't necessarily. It's possible to be very gold-heavy and to buy most of your stuff (eg getting commerce tree and trading post spamming), as opposed to being hammer heavy.

Spoils of War had no influence on these changes. I feel that policy is balanced well in the standard/emperor/continents+ conquest games I play.

The gold feels okish, its the culture gain that bothers me [but I haven't tested again lately.]

It adds up all city, connection, and garrison gold for both players and multiplies by 0.02:

If I am reading this right, both people get 2% of the sum.
So, if I have 4 OB agreements with people who have the same gold income as I do (call this X), then I am increasing my gold income by 16%. Because it takes 2% of 2X for each OB pairing. Right?
This is pretty significant in the late game, because it compounds with every other modifier.

But I will test first.

 

[Txurce]

By giving a minor economic incentive to do the agreement both ways, it makes the game less annoying and exploitative.

I doubt even 25% of TBC players found getting 40g for OB annoying and exploitative, but... whatever. More to the point is that the early reaction you're getting is that it's not a "minor economic incentive."

Spoils of War had no influence on these changes. I feel that policy is balanced well in the standard/emperor/continents+ conquest games I play.

The gold feels okish, its the culture gain that bothers me [but I haven't tested again lately.]

SOW may very well be balanced with 7.2.4's added building and unit maintenance costs. But it wasn't balanced before, and led to results like my getting 100+ happiness in consecutive games, tens of thousands of gold I wasn't spending and... to Ahriman's point... buying multiple instant-culture buildings in every one of my cities, thereby financing the sort of large-empire culture victory he has spoken out against.

 

[Txurce]

1. The advantages of travelling through someone's lands are generally low, particularly in the early game when players do not control much land. Whereas getting extra gold to buy another building, worker, settler, CS alliance, etc. is very valuable indeed.

2. I don't find it entertaining, it just feels gamey to me. Oooh, look, I can exploit this thing that the AI doesn't understand!

1. Yes, the general advantage is scouting. I usually trade OB's once for that purpose, then charge from then on. But the AI likes to send units through my territory, while I do not. Is that an "exploit," or just one of the myriad ways in which the AI plays an altogether different (and imperfect) game?

2. I found vanilla RA blocking gamey, and therefore chose not to do it. Do you find luxury sales in TBC gamey enough to not sell to already-happy AI?

 

[Ahriman]

SOW may very well be balanced with 7.2.4's added building and unit maintenance costs. But it wasn't balanced before, and led to results like my getting 100+ happiness in consecutive games, tens of thousands of gold I wasn't spending and... to Ahriman's point... buying multiple instant-culture buildings in every one of my cities, thereby financing the sort of large-empire culture victory he has spoken out against.

I haven't used SOW enough to be confident in my opinions here, so I will defer judgement for now. But if its really giving very big gold hauls, then yes, that's a problem.

Yes, the general advantage is scouting

I don't find strategic scouting to be that useful in Civ5. You can already get an idea of how strong the AI's military is, and its ability to instantly purchase units with gold means that the units you can see aren't a very good guide anyway. And maps are small enough and movement rates high enough that you can't really blitz-invade anyway. It takes a few turns to take down a city, and the enemy forces will be on you by then, so it's not like scouting lets you spring a lightning attack to capture cities while their units are away.
Also, maintenance costs are such that you are better off leaving your army as lower tech units and then upgrading them when war is declared, and the AI seems to do this too sometimes, so seeing what units they have isn't even a good guide to their military tech level.
[This is actually one of the things I really dislike about the TBC unit maintenance system; it feels gamey to deliberately keep my military obsolete so as to avoid maintenance costs.]

So I think it is less useful than in Civ4.

Tactical scouting has value (and arguably more than in previous civs, because of the added importance of unit positioning).
But OBs don't really help with this.

But the AI likes to send units through my territory

I don't really observe this.

Do you find luxury sales in TBC gamey enough to not sell to already-happy AI?

The AI at least gains some real advantage from additional happiness (faster golden ages). The AI gains almost nothing from open borders that it doesn't need.

 

[Txurce]

I don't find strategic scouting to be that useful in Civ5.

I don't really observe this.

The AI at least gains some real advantage from additional happiness (faster golden ages). The AI gains almost nothing from open borders that it doesn't need.

1. Me, neither. I was only saying that it has some use, and therefore some sort of OB-for-gold exchange isn't a flat-out exploit.

2. I observe it all the time, and some others have posted here that it's the main reason why they don't give OB's.

3. I agree. That's why I wouldn't call selling luxuries "gamey."

 

[Ahriman]

. I observe it all the time, and some others have posted here that it's the main reason why they don't give OB's.

I have seen the AI move units through my territory, but its mostly because they're moving somewhere else, and my territory has roads which makes the trip faster; they go through my territory because of the pathfinding system, not because of a scouting routine.
I have seen no evidence that the AI is really scouting, or that this affects AI behavior. I think that posters may be risking anthropomorphizing the AI; attributing human-like motives to it when it doesn't actually have them. I would be surprised if AI DoW decisions were based for example on what it had seen about military strength or whatever in your territory in previous turns.

That's why I wouldn't call selling luxuries "gamey."

I didn't. I called selling OBs gamey.

 

[Txurce]

I have seen the AI move units through my territory, but its mostly because they're moving somewhere else, and my territory has roads which makes the trip faster; they go through my territory because of the pathfinding system, not because of a scouting routine.
I have seen no evidence that the AI is really scouting, or that this affects AI behavior. I think that posters may be risking anthropomorphizing the AI; attributing human-like motives to it when it doesn't actually have them. I would be surprised if AI DoW decisions were based for example on what it had seen about military strength or whatever in your territory in previous turns.


I didn't. I called selling OBs gamey.

1. I didn't ascribe any motive - I just said that by definition, if they're there, they have one. It might be scouting, but could also be a projection of strength. I have no idea... I only know that, as others have commented, they are present enough tp sometimes screw up worker assignments and war positioning.

1. My mistake. A couple of posts up, you were referring to one thing, and I thought it was the other.

 

[Ahriman]

I just said that by definition, if they're there, they have one.

Their motive is very likely just that they want to use your road network to get where they are going. Nothing more. That isn't really much of a motive.
It is dangerous to assume that just because an AI does something that it has a good reason for doing so. This is what I meant by anthropomorphizing.

I only know that, as others have commented, they are present enough tp sometimes screw up worker assignments and war positioning.

The implication from the discussion above was that the AI gets some scouting benefit from OB, based on the observation that it sends units through your territory.
The fact that the human is sometimes inconvenienced by AI units doesn't mean that the AI is gaining any benefit.

Personally, I am really irritated by the fact that you can't send civilian units through non-hostile military units. This is one of my pet peeves (not a type of dog!) about how they implemented 1upt. I think it should be 1 military unit 1 civilian unit per tile, but that it shouldn't matter which faction they are as long as they are friendly. It's unlikely we can change this though.

 

[Txurce]

Their motive is very likely just that they want to use your road network to get where they are going. Nothing more. That isn't really much of a motive.
It is dangerous to assume that just because an AI does something that it has a good reason for doing so. This is what I meant by anthropomorphizing.


The implication from the discussion above was that the AI gets some scouting benefit from OB, based on the observation that it sends units through your territory.
The fact that the human is sometimes inconvenienced by AI units doesn't mean that the AI is gaining any benefit.

Personally, I am really irritated by the fact that you can't send civilian units through non-hostile military units. This is one of my pet peeves (not a type of dog!) about how they implemented 1upt. I think it should be 1 military unit 1 civilian unit per tile, but that it shouldn't matter which faction they are as long as they are friendly. It's unlikely we can change this though.

There's no way for us to know if they scout or not. (My guess is, not a whole lot.) But if they're passing through my territory, I would say they are gaining some benefit.

I also hate that kink, which is probably a bug.

 

[Ahriman]

But if they're passing through my territory, I would say they are gaining some benefit.

This is dangerous logic! If you think they are getting a benefit, identify the benefit. It is entirely possibly that they get a benefit because they think they will get to their end destination slightly faster. [Even if this is not actually true, because the pathfinding changes en route because they encounter your units on a road and so they idiotically step off the road.]
But this is pretty trivial.

I also hate that kink, which is probably a bug.

My guess is just design failure. They didn't come up with a way of having two factions have units on the same tile. Civ 1-3 were like this too IIRC, only in Civ4 were we allowed to coexist with allies.

 

[Txurce]

This is dangerous logic! If you think they are getting a benefit, identify the benefit. It is entirely possibly that they get a benefit because they think they will get to their end destination slightly faster. [Even if this is not actually true, because the pathfinding changes en route because they encounter your units on a road and so they idiotically step off the road.]
But this is pretty trivial.

You identified the benefit. I agree it's trivial. I don't think much of AI decision-making. But that doesn't make it utterly worthless. In fact, from a rich-as-hell AI perspective, 1.3g/turn for free passage is nothing. Speaking dispassionately, I would say that usually they are paying something they find inconsequential for something we find inconsequential.

 

[Ahriman]

40 gold in the early and early midgame is not inconsequential.

 

[Ahriman]

I had another thought; with all the gold cost increases in unit and building purchase costs, how has the balance shifted between spending gold on buying buildings/units vs investing in city states?

The more expensive you make units to buy, the more attractive you make (for example) military city state alliances. Do we risk getting back to the stage where investing gold in city states is a no-brainer?