Sooo... Happiness.

[BiteInTheMark]

@Gazebo, is the use of culture and science process also frequent in your test games? It does not feel right if the AI convert its bonus to production into culture or science because it has nothing else to do with its production.

@BiteInTheMark, If you have test time, you can maybe make a game where you deleted from existence all the process, to see if needs are more reasonnable without, or if this issue is unrelated to happiness.
I think you just need to add the following code at the end of Projects.sql:

DELETE FROM Processes
WHERE Type = 'PROCESS_CULTURE' OR Type = 'PROCESS_SCIENCE' OR Type = 'PROCESS_FOOD' OR Type = 'PROCESS_GOLD' OR Type = 'PROCESS_DEFENSE';

DELETE FROM Process_Flavors
WHERE ProcessType = 'PROCESS_CULTURE' OR ProcessType = 'PROCESS_SCIENCE' OR ProcessType = 'PROCESS_FOOD' OR ProcessType = 'PROCESS_GOLD' OR ProcessType = 'PROCESS_DEFENSE';

DELETE FROM Process_ProductionYield
WHERE ProcessType = 'PROCESS_CULTURE' OR ProcessType = 'PROCESS_SCIENCE' OR ProcessType = 'PROCESS_FOOD' OR ProcessType = 'PROCESS_GOLD' OR ProcessType = 'PROCESS_DEFENSE';

Iam talking about that now for 2 or 3 months..... processes are distorting the need calculations. Which is a direct translation from AI handicaps to player happiness.
I cant tell how big the influence is, maybe not that big, but its a part of it. Infixio also confirmed 2 or 3 months ago in his test runs the AI is using a lot of processes if everything is build and their cap is reached. He saw it in smaller extends even in king difficulty. But I think emperor is the first difficulty the AI can produce fast enough to stay up to date with tech even they are losig units in wars.
It would fit with the observations that some people get a lot of happiness issues later in the game. The handicaps for the AIs are getting bigger, they can finish the construction projects faster and run more often the processes.

In my atual game with netherlands, 3 of the 4 core cities of portugal are running processes. And its turn 180.

 

[Gazebo]

@Gazebo, is the use of culture and science process also frequent in your test games? It does not feel right if the AI convert its bonus to production into culture or science because it has nothing else to do with its production.

@BiteInTheMark, If you have test time, you can maybe make a game where you deleted from existence all the process, to see if needs are more reasonnable without, or if this issue is unrelated to happiness.
I think you just need to add the following code at the end of Projects.sql:

DELETE FROM Processes
WHERE Type = 'PROCESS_CULTURE' OR Type = 'PROCESS_SCIENCE' OR Type = 'PROCESS_FOOD' OR Type = 'PROCESS_GOLD' OR Type = 'PROCESS_DEFENSE';

DELETE FROM Process_Flavors
WHERE ProcessType = 'PROCESS_CULTURE' OR ProcessType = 'PROCESS_SCIENCE' OR ProcessType = 'PROCESS_FOOD' OR ProcessType = 'PROCESS_GOLD' OR ProcessType = 'PROCESS_DEFENSE';

DELETE FROM Process_ProductionYield
WHERE ProcessType = 'PROCESS_CULTURE' OR ProcessType = 'PROCESS_SCIENCE' OR ProcessType = 'PROCESS_FOOD' OR ProcessType = 'PROCESS_GOLD' OR ProcessType = 'PROCESS_DEFENSE';

I don’t see this, no. Never have. Not even on higher difficulties.

G

 

[BiteInTheMark]

I don’t see this, no. Never have. Not even on higher difficulties.

G

AI doing processes? You are kidding?
Ok.... I will make the game Iam playing to test lab and post my results.

 

[Gazebo]

AI doing processes? You are kidding?
Ok.... I will make the game Iam playing to test lab and post my results.

No, my point was not that they 'never' do processes, but rather that they never do them for extended periods (enough that it would grossly upset the median).

G

 

[Gazebo]

No, my point was not that they 'never' do processes, but rather that they never do them for extended periods (enough that it would grossly upset the median).

G

Edit: For the curious, a run I just did of a Deity AI-only game showed that, at any given time, around 6-7 cities in the world were running processes (it was intermittent, with the average process duration around 4 turns per city). Considering the lag-curve built into the median method, and the fact that the median looks for the 'middlest' value (not an average), you would need 40% or more cities in the entire world to run processes at the same time to have an appreciable effect on the median (and, even then, you'd be elevating the median by a just a few yields at most, given the conversion power of processes).

It's a non-issue, as I've stated before.

G

 

[Stalker0]

Yeah I haven't felt that the median calculation has been a big issue for a while, as you've said it remains fairly stable. I think the NEEDS inflation is what drives most of the unhappiness, because I do think that tends to hit a lot of of your cities at once, which causes some of the large swings in happiness.

 

[Gazebo]

Yeah I haven't felt that the median calculation has been a big issue for a while, as you've said it remains fairly stable. I think the NEEDS inflation is what drives most of the unhappiness, because I do think that tends to hit a lot of of your cities at once, which causes some of the large swings in happiness.

The biggest issue is that the tech % increase was simply scaling beyond the player's control in most situations, leading to situations where the median was outstripping the capacity of the player to deal with it with the infrastructure available to them. Taking that away, and replacing it with a tech average that penalizes tech leaders and benefits tech laggards, was the correct choice here. Getting the numbers right, however, takes time, as I removed a constant tech(t) scaler and replaced it with another variable scaler. I'm close, though.

I've also turned the city pop scaler into an exponential model, so that wide play is much more viable (whereas big cities hit a bit harder than they used to from a pop perspective). Ideally players will only dip into negative happiness from mistakes and/or overextension/bad wars, not average gameplay.

G

 

[Txurce]

Ideally players will only dip into negative happiness from mistakes and/or overextension/bad wars, not average gameplay.

Reach this ideal and you are effectively done, sir.

 

[chicorbeef]

I've also turned the city pop scaler into an exponential model, so that wide play is much more viable (whereas big cities hit a bit harder than they used to from a pop perspective).

G

Was this part needed? I thought Tall vs. Wide was already balanced.

 

[Gazebo]

Was this part needed? I thought Tall vs. Wide was already balanced.

Yes, because tech and city modifiers went hand in hand - without the change, tall civs were performing extraordinarily well.

G

 

[BiteInTheMark]

The biggest issue is that the tech % increase was simply scaling beyond the player's control in most situations, leading to situations where the median was outstripping the capacity of the player to deal with it with the infrastructure available to them. Taking that away, and replacing it with a tech average that penalizes tech leaders and benefits tech laggards, was the correct choice here. Getting the numbers right, however, takes time, as I removed a constant tech(t) scaler and replaced it with another variable scaler. I'm close, though.

I've also turned the city pop scaler into an exponential model, so that wide play is much more viable (whereas big cities hit a bit harder than they used to from a pop perspective). Ideally players will only dip into negative happiness from mistakes and/or overextension/bad wars, not average gameplay.

G

Wouldnt be reducing the tech modifier the easier solution? Or making it logarithmic?
Ive said the processes may do not have that much influence. But it has some and should be observerd.
But Moi Magnus take it to the point. A bigger question is, do we (the community) want this? AI is already getting additional yields by events. And it looks like the amount is not small, else I cant explain, how a civ with less culture output and more cities can outperform me in social policies like the portoguese in my game do. Should we allow the AI to transform their production handicap, which was made mainly to compensate losing more units in war and bad city placement, into a cultureal and science advantage?

 

[ryanmusante]

So, if the AI running processes could be an imbalance, what of the human player(s) doing the same thing. Won't my own doing cause the fluctuation of medians just the same?
Is running processes for extended periods of time for many cities a "bad" thing no matter the source or average length of time?

 

[Gazebo]

Wouldnt be reducing the tech modifier the easier solution? Or making it logarithmic?
Ive said the processes may do not have that much influence. But it has some and should be observerd.
But Moi Magnus take it to the point. A bigger question is, do we (the community) want this? AI is already getting additional yields by events. And it looks like the amount is not small, else I cant explain, how a civ with less culture output and more cities can outperform me in social policies like the portoguese in my game do. Should we allow the AI to transform their production handicap, which was made mainly to compensate losing more units in war and bad city placement, into a cultureal and science advantage?

For those keeping score this is reference #102 to the AI's production handicap.

I've already explained why the tech modifier needed to be changed.

AI handicaps are known, there are no secrets.

So, if the AI running processes could be an imbalance, what of the human player(s) doing the same thing. Won't my own doing cause the fluctuation of medians just the same?
Is running processes for extended periods of time for many cities a "bad" thing no matter the source or average length of time?

Processes have no appreciable impact on the median.

 

[tu_79]

For those keeping score this is reference #102 to the AI's production handicap.

I've already explained why the tech modifier needed to be changed.

AI handicaps are known, there are no secrets.



Processes have no appreciable impact on the median.

After you are done with this, please, make sure that city governor knows how to grow cities best. I find myself removing workers from food tiles constantly.

 

[Gazebo]

Iam talking about that now for 2 or 3 months..... processes are distorting the need calculations. Which is a direct translation from AI handicaps to player happiness.

Ive said the processes may do not have that much influence. But it has some and should be observerd.

Also...you can't have it both ways, bud. So which is it - they're distorting (a strong word, mind you) the need calculations or they don't have much influence? If you tell me that by 'distorting' you meant 'not that much' then I'm not sure what you're trying to say.Yields have effects on things? Yup, they do. Boy howdy.

This is what I'm talking about, by the way - you're more interested in winning points arguments than you are in discussing solutions and my responses to raised concerns. I've got my nose in the code, trust me on things, ok?

G

 

[Randomsheep]

I've got my nose in the code, trust me on things, ok?

G

We trust you. Next version out now please.

 

[Omen of Peace]

After you are done with this, please, make sure that city governor knows how to grow cities best. I find myself removing workers from food tiles constantly.

Interesting. In my games (Emperor, Standard speed), happiness is not an issue*, but the governor seems to grow my cities too little (perhaps due to overuse of specialists, which provide nice yields but consume a lot of food and in many cities will never lead to a GP). I end up having to select the Food governor orientation in some cities in the mid-game, to let them grow and better work specialists and good tiles.

*but I tend to lead or be 2nd, which most likely helps given the current algorithm

 

[ryanmusante]

We trust you. Next version out now please.

 

[Hinin]

Spoiler Worsened version :

 

[ryanmusante]

Spoiler Worsened version :

View attachment 497865

I'm kidding.

Spoiler :