Author Topic: Happiness Testing  (Read 4048 times)

0 Members and 1 Guest are viewing this topic.

Offline Nilla

  • Champion
  • Banished Expert
  • *****
  • Posts: 2418
  • Country: se
Re: Happiness Testing
« Reply #45 on: January 06, 2018, 03:10:41 PM »
Interesting. Small but significant effect. I have tried similar things but didn't see this, so I'm a bit confused. Maybe I tested each of these "happiness maker" separate, and you only see any effect if you combine them. It's been a while, so I don't remember exactly. I have one question. I don't know, if it matters; do everyone live in the happines circle of the church or are there also churchmembers living outside?

I guess the reason for the switch between 4˝ and 5 stars the first time is due to the "blunt instrument" stars. Say 5 stars means a happines of 97-100% and 4˝ 93-97% (I'm not sure of the exact numbers, just the principle). If you have 96,9% happiness it will be 4,5 stars, if it rises to 97,0 it will be 5.

(tried to post this before but it didn´t work, I´ll try now again.)

Offline Tom Sawyer

  • Modders
  • Architect
  • *****
  • Posts: 845
  • Country: is
Re: Happiness Testing
« Reply #46 on: January 06, 2018, 03:26:04 PM »
Thank you for your researching @Maldrick. I also made a test today to proof your find and to have absolute evidence about this issue.^^ I wanted to know if there is really any calculation in game and any noticeable effect on happiness by a church. I want to share it also if that is irrelevant for the game because of the very small effect.

My setup was a town created with debug options of about 200 people and provided with basic services: market, herbalist, schools, cemetery and various workplaces to not let them idle the whole time. That means a population with full graphs and no dropping happiness because of death which only makes it more complicated. This setting I let run for some years and followed the graphs. Happiness was at 95% all the time with a few steps up or down by 1%. It made a difference between 4.5 and 5 stars but this is pseudo-accuracy and not useful to look at.

The first step down was after 3 years due to a death of a mother and her sad child. They can become unhappy even with a cemetery. This drop recovered after a while and more of these fluctuations happened from time to time. I'm pretty sure they are such cases of death.

Then at a point with recognizable points in happiness graph (for orientation) I placed a chapel in the middle of the town and waited for any effect. I got a 1% increasing after a short time (picture 1). Question was, if it's a fluctuation or a real church effect. It only could be proofed by replication. I loaded the save game I made before building the chapel and let it run without this building.. no effect (picture 2).

Then I made a double test to proof it finally. 3. run again with chapel: Church effect (picture 3). 4. Run again without chapel: No effect (picture 4). I made even more runs and its totally clear. The chapel has indeed a very small effect on average happiness of about +1% in this setup. This effect is totally negligible and not worth to build a chapel (except as a decoration) but technically this function is included. So you are right. :)

http://www.banishedventures.com/images/churchtest1.png
Happiness Testing
http://www.banishedventures.com/images/churchtest2.png
Happiness Testing
http://www.banishedventures.com/images/churchtest3.png
Happiness Testing
http://www.banishedventures.com/images/churchtest4.png
Happiness Testing


« Last Edit: January 06, 2018, 03:37:32 PM by Tom Sawyer »

Offline Maldrick

  • Trader
  • ****
  • Posts: 496
  • Country: us
Re: Happiness Testing
« Reply #47 on: January 06, 2018, 03:48:59 PM »
@Nilla To answer your question, I don't really know.  Not using Happiness Radius on this one.  I would assume it's somewhere in the area of 75%.  This start was heavy on forestry with the chapels close to the main town areas, but one of those (and the side the chapels for that side of town are on) is butted up next to one of the forest clusters and they are the small Forest Outpost foresters, so they could be in the radius.  Would say the same for at least one on the other side of town.  But, again, not really sure. 60-80% would be a good guess.

Yeah, I'm thinking the shift in stars comes from Cemetery getting it right to the line or just beyond.  It's just weird because I had solid 5 stars with just a cemetery for about 20 years or more, then I noticed it drop to 4.5.  Wasn't watching the graph, though.  Only thing I can think of making the difference is the Cemetery is in one corner of town and I was expanding away from it.  Could it be that Cemetery happiness is a combo of things?  Like you don't get the massive unhappiness from not having one when someone dies if you do, but there's another level of happiness that comes from accessibility?  Or it could have been from something else.  My food diversity was getting stretched at the time I noticed it and turned on the chapels.  Is the only other thing I can think of.  Does health affect happiness?  I've tried to keep both up consistently for so long I've never noticed.  Or forgotten if I did.

@Tom Sawyer Yep, that's pretty much what I saw, too, with the first test.  If you still have your save, you might try it again with Chapels and Wells, together.  Noticed a lot bigger increase with the combo.  Up to capped health level.

And, you're right....How important this is will vary from player to player.  It's minimal, but not insignificant necessarily, depending on your personal preference.  Don't know that I have the patience to actually test it, right now at least, but I'd be interested to see how that 2-3% difference in the average translates to individual bannies.
"We are the architects of our actions and we must live with their consequences, whether glorious or tragic.”
― Ezio Auditore da Firenze

Online smurphys7

  • Trader
  • ****
  • Posts: 126
  • Country: 00
Re: Happiness Testing
« Reply #48 on: January 06, 2018, 04:58:58 PM »
I have acquired more data.

I loaded my previous simulation town with the Cemetery.  This was the resulting Citizens graphDownload the file here.  I built a Trading Post.  I traded excess Leather for stone.  I built 4 Chapels.  There were periods of time where the Chapels were all on with 4 Clerics.  There were periods of time where the Chapels were all off with 0 Clerics.  The only other thing I did was turn off the Builders and Traders.  Here is the town layout with Chapels.  I have recorded the entire simulation.

Can you identify when the Chapels were off and when the Chapels were on?

https://i.imgur.com/12Nygju.png
Happiness Testing


Are there any flaws for with loading that town, building chapels like I did, and turning them either on or off while correspondingly turning the Clerics on and off?
« Last Edit: January 06, 2018, 05:24:38 PM by smurphys7 »

Online smurphys7

  • Trader
  • ****
  • Posts: 126
  • Country: 00
Re: Happiness Testing
« Reply #49 on: January 07, 2018, 12:05:08 AM »
I loaded up my town with Cemetery + Chapel + Well + Tavern w/500 Alcohol + Physician + Health + Warm Coats + Steel tools.  I hadn't run it long enough before.  At all points I was at the clothing limit.

https://i.imgur.com/fOn0JWQ.png
Happiness Testing


It took quite some time for the Chapels + Everything to come into full affect.  The blip in the middle was when a disease hit and a few people died.  I built a physician at that point.  At the end is where I turned on and off the chapels.  When I turned them on and off is clear.

My previous graphs of JUST the Chapels and turning them on and off is unclear.  I don't think one can tell where the Chapels were turned on and off.  But, Chapels + Everything else did have a noticeable affect. 

So Chapels + Something else does affect happiness.  As others have stated, it does not appear to be significant.

Offline Nilla

  • Champion
  • Banished Expert
  • *****
  • Posts: 2418
  • Country: se
Re: Happiness Testing
« Reply #50 on: January 09, 2018, 04:18:41 AM »
I just want to tell, that I kept a little bit track on the happiness of some individuals in my present game. I looked at the recovery of happiness by some young people. It's in my blog, the last two pictures. http://worldofbanished.com/index.php?topic=2125.msg43753#msg43753

Summery

If there's a graveyard young people lose 1-3 stars, if they lose a parent, but recover as they move out from home. If there's no graveyard (or like in my game, they are full) they lose all but ˝ stars and don't recover any happiness.


Offline Tom Sawyer

  • Modders
  • Architect
  • *****
  • Posts: 845
  • Country: is
Re: Happiness Testing
« Reply #51 on: January 09, 2018, 09:25:36 AM »
Yep, this reaction on death of family members I also noticed. Without cemetery they are defined to be depressive, almost not working anymore and no chance to recover or to counter with happiness buildings. The only solution is their own death, not nice. People really should be able to recover after a year of mourning or so.

I made more tests with happiness and tried to increase this effect we found in the graphs. It's difficult because we have only a few values and no insight into sources how the game calculates. But I could get this in my save game from last post after altering the citizens.rsc. Default happiness is set down from 95 to 50% and the impact of buildings with different happiness aspects scales up to 100. These steps are noticeable in individuell stars of citizens and now relevant in game. I will try to balance this and to use it in some way. But it's hard to reproduce these effects. I have test games where it works and in others the happiness calculation seems to not work at all. No idea why.

http://www.banishedventures.com/images/happinesstest.png
Happiness Testing

Offline Nilla

  • Champion
  • Banished Expert
  • *****
  • Posts: 2418
  • Country: se
Re: Happiness Testing
« Reply #52 on: January 09, 2018, 10:32:08 AM »
............. almost not working anymore ........

Have you really seen this? As I looked at this (a long time ago), they did idle more than happy people, but not that terribly much.

.................. I will try to balance this and to use it in some way. But it's hard to reproduce these effects. I have test games where it works and in others the happiness calculation seems to not work at all. No idea why.

That would be interesting. Maybe also, if you like to look at a larger effect (=more idling than today), if they are unhappy.

Maybe there's some "symbiotic" effect by these happiness maker. You need several to get effect or the opposite; if you already have happinessmaker A, happinessmaker B doesn't change anything.

Let us know your findings!

Offline Maldrick

  • Trader
  • ****
  • Posts: 496
  • Country: us
Re: Happiness Testing
« Reply #53 on: January 09, 2018, 04:45:22 PM »
Very interesting.  And consistent with what we were seeing.  Including sometimes having no apparent effects at all.
"We are the architects of our actions and we must live with their consequences, whether glorious or tragic.”
― Ezio Auditore da Firenze

Offline kid1293

  • Modders
  • Banished Expert
  • *****
  • Posts: 2602
  • Country: se
  • Never confuse movement with progress
Re: Happiness Testing
« Reply #54 on: January 26, 2018, 03:46:01 PM »
Sorry to revive this thread. But I have had it spinning for some time now.

Do you people remember Anno-series? And Caesar game?
People needed to have the circles overlapping to be fully functional.
Like in picture.

Only the white area was covered. To get full happiness I think Luke
was intending to actually use the radius in each happiness building.
Else there is no other explanation for using a radius.

We have all seen cemetery effects on unhappiness after a death.
Sometimes not. Maybe outside the radius.
Maybe the other sorts of happiness were not fully implemented.
Or are they?
Have you tested the 'safety' happiness when there is a disaster?
Have you tested 'goods' when people are starving/naked?
Have you tested 'entertainment' when... don't know... unemployed?  :)

It looks too much something I have seen in other games. I do not think
it is just a coincidence.

Sorry for disturbing the peace and quiet.
« Last Edit: January 26, 2018, 04:27:12 PM by kid1293 »

Offline Tom Sawyer

  • Modders
  • Architect
  • *****
  • Posts: 845
  • Country: is
Re: Happiness Testing
« Reply #55 on: January 26, 2018, 05:58:35 PM »
Yes @kid1293 that's the way it works here too. In debug options you can show these overlapping areas of all five aspects. They are fully implemented and stacking. In my tests I noticed an effect only if at least 3 aspects are overlapping and it reaches up to 100% happiness with 4 or 5 stacks. Buildings with the same kind of happiness are not stacking of course. There is not any necessary order or something like that. Also, every detracting object removes one of these areas. So with 5 mines they are fully removed. That's all working fine. The problem is the default happiness of 95% and the not noticeable increase (except of staring at the happiness graph with a magnifying glass). But I'm sure it can be solved by re-scaling and balancing. :)

http://www.banishedventures.com/images/happiness-radius.jpg
Happiness Testing

Offline Discrepancy

  • Modders
  • Architect
  • *****
  • Posts: 912
  • Country: au
Re: Happiness Testing
« Reply #56 on: January 26, 2018, 06:05:23 PM »
so what kind of effect does changing the Citizen.rsc template file give?
this section:

Code: [Select]
float _maxHappiness = 10.0;
int _minFoodForHappiness = 4;
int _maxFoodForHappiness = 12;
float _happinessUpdateInMonths = 3.0;
int _workForHappiness = 10;
float _unhappyLevel = 0.6;
float _depressedLevel = 0.25;
int _idleRange = 25;


Can we make them more unhappy?

that way the building happiness code should be more visible to the players?

Offline Tom Sawyer

  • Modders
  • Architect
  • *****
  • Posts: 845
  • Country: is
Re: Happiness Testing
« Reply #57 on: January 26, 2018, 06:30:58 PM »
To get the graph in this post from Jan. 9 I changed maxHappiness to 1. It was setting the default down to 50% but the depression effect by death events were not working properly anymore. I'm currently testing 1.25 for a better result where bannies start with 3 stars and get up to 5 with all buildings. This is noticeable in citizen UI and the death event also works. What needs to be tested is the effect on work efficiency / idling time.

Offline kid1293

  • Modders
  • Banished Expert
  • *****
  • Posts: 2602
  • Country: se
  • Never confuse movement with progress
Re: Happiness Testing
« Reply #58 on: January 26, 2018, 08:12:24 PM »
Thanks @Tom Sawyer , that was what I suspected.
So the 'missing' effects are actually not finished game stubs?
Luke didn't make anything of them? Like soil degradation, I mean.

Offline Maldrick

  • Trader
  • ****
  • Posts: 496
  • Country: us
Re: Happiness Testing
« Reply #59 on: January 27, 2018, 09:35:50 AM »
Just to add...When I was new to the game and learning I read somewhere that changing jobs affects happiness.  Never noticed that to be the case but never particularly looked either.

Modding base happiness to make the effects more significant is certainly interesting.  You guys might be on to something really cool with this.
"We are the architects of our actions and we must live with their consequences, whether glorious or tragic.”
― Ezio Auditore da Firenze