District Costs

So todays run I pushed tech and civics to see if I could get an early discount... it is looking more like you need to meet both blue criteria
I'll run another test pushing districts rather than techs... gonna be hard ignoring eurekas, you get so used to it

upload_2017-6-16_10-15-50.png


What happens, if you change on turn 97 'shipbuilding' to another tech ... is _the discount_ exactly delayed until the later completion of that tech? What if any civic is completed before that tech?
Well the save is there if you want to try... I'll see on the next test as that will rely on the techs.... I might load it and see but I imagine any will do

The 25 is just the multiplier (25% of 60 per however many techs and civics you need for that)
I'm not following, can you be more specific?
 
Last edited by a moderator:
I'm not following, can you be more specific?

The "=25" as last part of the formula only refers to the speed at which the district cost increases. If you compare it to the scaling on previous copies (builders, settlers), for example, an "=15" there means that every next copy is 15% more expensive than the last. If the base cost is 80, like a settler, then every next one is .15*80=12 more expensive, so the next would be 92, then 104, and so on (settler actually has a higher scaling than this I think, I believe 25). While I don't know how what equals to one increase of 25% in districts (=15 production), I do know that if you reduce the 25 to, say, 20, the cheaper/more expensive division between districts remains just as was, but the costs increase less than they otherwise would. In fact, in the one game I actually finished neighborhoods and aquaducts (which scale on game time) became more expensive than the expensive "normal" districts.
 
The "=25" as last part of the formula only refers to the speed at which the district cost increases.
I can get it right 100% of the time using the formula =FLOOR(30*(1+9*FLOOR(100*MAX(TECHS/67,CIVICS/51),1)/100),1) which does not use the 25
However the 25 likely means a 25% discount rather than to do with techs... I guess I can try this on another run sometime.
I have just done a longer run with more challenging results..... This time holding of on the techs and spamming as many districts as possible.... What gets me is I have built 3 cheap ED and the 4th still shows as cheap... I have just run out of places to build in the test.

So it looks like you will get faster discount with more techs... perhaps?
It certainly seems like you get more discount with lower techs ... as in I can gain around 200 cogs from at least 4 ED

Mayybe its base don the average number fo techs which was why I was spamming techs at the end of the test.

upload_2017-6-16_13-27-4.png


I thought I would just run a few more techs just to ensure techs would wear it down... sure enough 18 techs in the ED got more expensive... so definately looking like techs and completed districts but exactly what formula is a jolly good question.

So possibly keeping techs down not only gives you cheaper districts but with district bloating more cheaper districts

upload_2017-6-16_13-40-40.png
 
Last edited by a moderator:
You don't have VI yet? :( gutted
Why? I don't have VI yet, but foremost I don't have ENOUGH time to play much while I'm busy working on a project (real life etc.) ... I'm not philandering :D[...]
Depending on what I thought many months ago about how long civ6 would need to mature to a degree I will love, I committed myself to work on a project which will take many months more to come ...
Well the save is there if you want to try... I'll see on the next test as that will rely on the techs.... I might load it and see but I imagine any will do
I still don't have civ6 ... and won't anytime soon. (I expected it to be hard to not play civ6, but ok. Though I was completely unaware it also messes up forum participation.)
No problem. :)
 
So I suppose, the answer to my question is 'yes' & 'the same' and summarise temporarily:

Preliminary draft
(The effective cost is calculated & locked in upon PLACING DOWN a district, 67 Techs & 51 Civics in game, rounding down)
District cost = 60* (1 + 9*[ 100*Larger of{ Number of Techs/67 OR Number of Civics/51 } ]/100)
- Aqueduct 50, max 500
- normal districts max 600
- district types rarely used max 450 (=75%)
Aerodrome, Aqueduct and Neighborhood scale on game time, Spaceport doesn't scale at all (fix 2000). So Campus, Commercial Hub, Encampment, Entertainment Complex, Harbor, Holy Site, Industrial Zone & Theater Square can receive the 25% discount. "less than average" is calculated by adding all districts of this types _you_ have together and then divide by 8.

To avoid, that if you have just one or a few districts all other would be at reduced price, at least 8 districts are needed to get discounts.
Discounts are turned ON based on completed districts and triggered by the completion of techs / civics or finishing districts.
Discounts are turned OFF based on completed+uncompleted districts and triggered by the completion of techs / civics or placing districts.
 
Last edited:
is calculated by adding all districts of this types _you_ have together and then divide by 8.
It's this bit that does not work?.. in my last test I could get 3 districts discounted with only 20 down... it seems there is a science inclusion in there as the 3rd stopped being discounted once I hit a certain number of techs.

We have something that is roughly good enough and a lot better than when @CPL_Yoshi asked the question but it's not accurate.... busy weekend for me but I will ponder

This is important for something cpl_yoshi is doing
 
You are right, it's not accurate ... Though I'd say it gives a normal player useful hints on which districts are to expect discounts and how to use that moderately - of course it is still not exact enough for the MinMax-Player, who wants to exploit discounts ...
in my last test I could get 3 districts discounted with only 20 down... it seems there is a science inclusion in there as the 3rd stopped being discounted once I hit a certain number of techs.
I saw this and I would be very interested in this last test, if you could run part of it in another direction of finding limits: from turn 118 on could eventually be built more than 3 Entertainment Complexes ... How many? (5 like Campuses, 6 like Commercial Hubs and Industrial Zones or even more???) ... eventually delaying techs, so they won't wear the discount down ...
 
Last edited:
from turn 118 on could eventually be built more than 3 Entertainment Complexes
The issue with these tests are they are sort of pre-built changing direction in a short number of turns would be hard.

Would it not be better to come up with a theory and disprove it... like (districts+techs)/8
Or where I spam science before building districts.
 
The issue with these tests are they are sort of pre-built changing direction in a short number of turns would be hard.
I am aware that building additionally Entertainment Complexes beyond those 3 built between turn 111 and 118 would have to be set up earlier (along with the 3 already built), but thought this would be feasible.
Maybe that can be sped up by setting the districts directly via the index WorldBuilderCity or by setting a lot chopable woods via WorldBuilder in FireTuner:
You can actually see the exact Diplomacy score in Firetuner. [...]View attachment 468658

---

Would it not be better to come up with a theory and disprove it...
You already falsified "less than average" is calculated by adding all districts of this types _you_ have together and then divide by 8 ... as you pointed out, 20/8 = 2.5 would only allow 2 Entertainment Complexes on discount, you managed already 3.

So the New theory is that the number of districts is ranked (at the end of your last test: 0-0-0-0-3-5-6-6) and as long as the 4th slot is 0, the number of the 5th - 8th is unlimited: the 1st - 4th remains discounted.

The next step (if 0-0-0-0-6-6-6-6 confirms '1st - 4th remains discounted') would look for the number on which 0-0-0-x-6-6-6-6 looses the discount.
 
Last edited:
Without having followed the experiments, upon reading the last post by c4c6, I just want to say: It may also pay to think about what's realistic in terms of making sense and being something someone would actually come up with. The game designers must have had some thought with whatever mechanism they implemented. The suggested explanation might fit the data, but to me, it seems completely random, and hence unlikely to be the actual mechanism. But that's just my not-very-helpful 5 cents.
 
ould look for the number on which 0-0-0-x-6-6-6-6 looses the discount.
OK but the science bit counts also... the last test I was avoiding eurekas.. its was not at all easy ... to get to 6666 with low science is tough because I need a high pop to do so. Not so easy and you do not mention the tech 2.

I Will see what I can do tonight to get more cities earlier by just using firetuner to give me lots of gold... hmmm it can remove techs as well and I can ty that if I get in trouble but whether the function references the same place is a good question.
All I know is do too much like add cities and districts manuall and hings start going odd and just what can you trust... I guess messing with a gold counter will be safe.

But that's just my not-very-helpful 5 cents.
No @kaspergm everything is helpful, thank you... you spend many hours doing this and when you get few replies it gets lonely

I suspect yu ar right with the devs and the last test was quite unreal because i forcably avoided a lot of eurekas.
I answer is likely close to the name of the function call

COST_PROGRESSION_NUM_UNDER_AVG_PLUS_TECH

So the progression threshold could be the COST_PROGRESSION_NUM
and it must be under the value of the average disticts + techs

its just working out what it is2

What I do see from the last 2 runs is that the discount starts happening when the number of techs and the number of districts when added together is greater than 25.. 2 runs does not make a formula so I am going to try a 3rd with this theory in mind.

If we can sort out when the discount starts happening first... I'll concentrate on that and lay the stats in here and with enough theories/data we can hopefully get an answer to how many can be discounted... all ideas comments appreciated
 
Last edited by a moderator:
The game designers must have had some thought with whatever mechanism they implemented. The suggested explanation might fit the data, but to me, it seems completely random, and hence unlikely to be the actual mechanism.
"the New theory" describes just 2 testcases - "the suggested explanation" will fit better to "game design mechanisms" (keyword Median).

Firetuner is erratic at best, often the district does not get added into the equation, a created city disappears but is still there... it's just too buggy to trust
Setting woods & giving builders with Firetuner seems to me simpler and probably is on a relative low risk of errors, maybe this can be used to restart on turn 118 and then chop them to get 1 Campus and 3 Entertainment Complexes?
OK but the science bit counts also... the last test I was avoiding eurekas.. [...] Not so easy and you do not mention the tech 2.
I'd avoid new techs&civics by changing them to others before completion.
In general I'd try to keep most the same (ceteris paribus).
hmmm it can remove techs as well and I can try that if I get in trouble but whether the function references the same place is a good question. [...] I guess messing with a gold counter will be safe.
I wouldn't touch the techs with Firetuner, because you can never know what that means for the calculations we are examining.
What I do see from the last 2 runs is that the discount starts happening when the number of techs and the number of districts when added together is greater than 25..
Your post #54 gives discount with 15 techs and 9 completed districts though.
I suspect the 25 has another meaning.
 
OK but the science bit counts also... the last test I was avoiding eurekas.. its was not at all easy ... to get to 6666 with low science is tough because I need a high pop to do so. Not so easy and you do not mention the tech 2.

You know, with the Time x3 mod, your above statement makes a ton of sense. Districts are so much easier to build because the science and culture rates are slowed (but not production). Maybe that's one of the reasons it is more fun to play?
 
Setting woods & giving builders with Firetuner seems to me simpler and probably
I'm using Japan this time as they have 3 cheap districts... population means I'm gonna have to get ED at some time. turn 40 and 8 cities and 6 districts so game a lot faster to test

I have given myself 50K to start so buy builders and settlers and chop in, limiting it to just cash as I have had some odd Firetuner bugs so just being cautious

Your post #54 gives discount with 15 techs and 9 completed districts though.
Yeah ... 15 techs and 8 districts the 9 was civics but thats 23.. :( Some combination of techs and districts does seem to be giving it

Utter fail ... turn 38. I have built my encampments too fast... I have 6 online and my Theatre districts come up to make as my second district and show a discount... no way of hoing back to previous turs to see when it started ... however to be able to get a discount after 6 districts and 5 techs is something of use I guess. Just gotta tweak what I get.
 
Last edited by a moderator:
Your post #37 has also a first discount at 6 completed districts! I thought this was a glitch ... but now 6 is confirmed as Minimum number of districts for discount.

Are all other district types discounted or just Theater districts?
 
... but now 6 is confirmed as Minimum number of districts for discount
Incorrect.. new testing is showing it lower! ... and it disappears with either tech or era or new district change (the new district is a commercial, CBA updating the spreadsheet, I will post whole presdsheet when done.
This certainly confirms #37 was not a glitch ... I still have those saves but its just easier with a better designed scenario to restore and change than go back to that
If on turn 22 I change it and complete a tech rather than a district, the discount goes then not on turn 23.
This test is showing you can get a discount very early timed right. (not turn 22 in a normal game though)
@CPL_Yoshi this seems strong combined with what we discussed.

upload_2017-6-18_1-6-12.png
 
Last edited:
Yeah, great work @c4c6 and @Victoria. I've been quite busy lately, and you've done so much work on this. If you'd like to make a post on the topic that we've discussed, go for it. I'll try to contribute where I can.
 
Top Bottom