Fixing Influence - any ideas welcome

Describe your experience with the latest version of FreeOrion to help us improve it.

Moderator: Oberlus

Forum rules
Always mention the exact version of FreeOrion you are testing.

When reporting an issue regarding the AI, if possible provide the relevant AI log file and a save game file that demonstrates the issue.
Message
Author
User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Fixing Influence - any ideas welcome

#46 Post by Oberlus »

My experience is that empires with good research resources can get to late techs like Death Rays and Titanic hulls 50 or 100 turns sooner than other empires. And having so much time to milk the technological advantage is great. Plus they can then switch to more production and keep a moderate tech progression after that and still keep a good technological lead.

Daybreak
Vacuum Dragon
Posts: 641
Joined: Mon Aug 13, 2018 10:14 pm

Re: Fixing Influence - any ideas welcome

#47 Post by Daybreak »

Oberlus wrote: Fri Aug 12, 2022 10:18 am My experience is that empires with good research resources can get to late techs like Death Rays and Titanic hulls 50 or 100 turns sooner than other empires. And having so much time to milk the technological advantage is great. Plus they can then switch to more production and keep a moderate tech progression after that and still keep a good technological lead.
I wonder if that is as true a it used to be. I am finding that with a much lower PP now being experienced overall, it seems spamming as many low cost ships as possible with lower costing tech/weapons is better than only a few higher tech/weapon ships, which just get overwhelmed. It used to be that did not happen as much due to superior shields, but now they don't matter as much with cheap spammable carriers/fighters.

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Fixing Influence - any ideas welcome

#48 Post by Oberlus »

True for 20th MP.

wobbly
Cosmic Dragon
Posts: 1874
Joined: Thu Oct 10, 2013 6:48 pm

Re: Fixing Influence - any ideas welcome

#49 Post by wobbly »

Ophiuchus wrote: Fri Aug 12, 2022 9:36 am
BlueAward wrote: Wed Aug 10, 2022 8:39 pm Of course, in Scylior case, you only need so much research; production looks to be more important in grand scheme of things. So maybe the science effect is similar, but it eventually trails off. Meanwhile, influence stays with you, specifically to put a soft cap on the size of your empire, and I think the difference on the cap is indeed substantial depending on how much influence you get, but don't have numbers to back it up.
that is in my opinion mostly because we did not balance research costs when we introducing influence.

before influence we had unhindered exponential growth in number of colonies. so also research points were increasing in an exponential way.
That is what the tech costs are based on, so of course late tech can be waaay to expensive, shifting power balance away from research.
And it was insanely broken for late game Scyllor or Chato. If you reached end-game and you had a large mass of either of these species and your opponents didn't, there was just no competition.

BlueAward
Vacuum Dragon
Posts: 646
Joined: Mon Aug 08, 2022 3:15 am

Re: Fixing Influence - any ideas welcome

#50 Post by BlueAward »

I amused myself with some math. Please review.

I gather influence upkeep per colony is what is set to soft cap your empire growth, as it grows exponentially. From other thread I saw it is 0.4*sqrt(x) per colony, though 4 outposts count as 1 colony. Let's not go into those intricacies and just call total colony upkeep cost as U

Let's call average influence point output per planet I

So in normal case you want x planets set to influence to meet the U demand: U = x * I

Now if you have artistic species and go artisan workshops, you need y planets instead that each produces I + 4 IP

If you fill in other planets with the artistic species it would have additional effect if stability >= 10.0 making things even better but for know let's assume you want to put artistic species only on the planets you want influence focus on. So you now need to meet demand with, let's say, y artistic planets: U = y * (I + 4)

U = x * I
U = y * (I + 4)

x * I = y * (I + 4)

x = y * (I + 4) ÷ I

From example by Oberlus, assuming planet population 30, average IP would be sqrt(30) right? So I mean:
x = y * (sqrt(30) + 4) ÷ sqrt(30)
y ≈ 0.577936 x

So having artistic species means needing 57.8% as few planets on influence focus as when not having artistic species assuming 30 pop per planet on average and ignoring the other effect of artisan workshops that makes it even better (though you go artisan workshops policy which excludes some other stuff/alternatives of getting extra influence points)

Following similar reasoning, If you get species that produces 50% of I, then you need twice as many planets on influence focus. If a species produces 200% of I, then you need half as many. Scylior, Human or Cray are 1.5 times the influence and artistic, so you can combine the two effects so it's more like 38.5% initial requirement of planets put on influence focus. Even better for Scylior thanks to population bonus, though. And if you put artistic species on other planets than influence focused - even better.

Also real life case will probably mean you'll concentrate influence on planets somehow especially suited for it (having some extra bonuses) so that again goes slightly further.

Note all those considerations afford a look that takes actual upkeep formula literally out of equation. The amount of needed planets will still grow exponentially, O(n^3/2), but with 200% influence output you'd need half as many as 100% influence etc. I mean maybe those finding are not surprising, but maybe there are some insights there

Oberlus in his nice thread goes with 30 planets empire, 30 pop on average each, in which case you'd need 12 planets on influence output (comes up between 11 and 12). So 6 planets less if you have a 200% influence species on those. With Artisan Workshops that would be like 7 planets (again integer math, between 6 and 7). He did not yet add a chapter on artisan workshops, but I guess he'd agree with me here, though again, putting artisan species on other planets not set on influence improves things... Of course other compatible policies if relevant like Diversity or Terraforming also could improve things (edit: and if it is both artistic and higher influence, even "more better")
Last edited by BlueAward on Fri Aug 12, 2022 6:00 pm, edited 1 time in total.

wobbly
Cosmic Dragon
Posts: 1874
Joined: Thu Oct 10, 2013 6:48 pm

Re: Fixing Influence - any ideas welcome

#51 Post by wobbly »

Artisians workshop is also crazy if you have Egassem. It's only +0.5 influence for a planet on production, but when you start piling on Egassem for all your production it adds up.

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Fixing Influence - any ideas welcome

#52 Post by Oberlus »

50% improvement is too much. Needs a huge nerf.

The gains from most stuff (techs, policies, species traits, etc. compared to the bonuses without the techs/policies/etc.) should not be greater than 25% on good conditions. Gains in the order of 10% are more reasonable for stuff that doesn't require much investment (and having a species and a policy is not much investment).

BlueAward
Vacuum Dragon
Posts: 646
Joined: Mon Aug 08, 2022 3:15 am

Re: Fixing Influence - any ideas welcome

#53 Post by BlueAward »

wobbly wrote: Fri Aug 12, 2022 5:06 pm Artisians workshop is also crazy if you have Egassem. It's only +0.5 influence for a planet on production, but when you start piling on Egassem for all your production it adds up.
Yeah that's the part I was ignoring, and just saying it makes things even better. Indeed it works extra towards offsetting upkeep costs themselves and scales with number of planets you have, directly offsetting a bit of U (though that still remains exponential). And in Scylior case let's not forget they are also above average production, having 125% population hence roughly 125% production as average, no? (egassem roughly 150% production)

And note diverse empire with many different species and relevant policy has that part even better covered, another thing one could try to muse about in terms of numbers (I think Oberlus plans it, looking at his thread?)

BlueAward
Vacuum Dragon
Posts: 646
Joined: Mon Aug 08, 2022 3:15 am

Re: Fixing Influence - any ideas welcome

#54 Post by BlueAward »

Oberlus wrote: Fri Aug 12, 2022 5:13 pm 50% improvement is too much. Needs a huge nerf.

The gains from most stuff (techs, policies, species traits, etc. compared to the bonuses without the techs/policies/etc.) should not be greater than 25% on good conditions. Gains in the order of 10% are more reasonable for stuff that doesn't require much investment (and having a species and a policy is not much investment).
I mean there's still this side of the coin that with species of 200% science you'd need half as many planets to get same output in science, remaining planets doing production or influence as needed. Same can be said about production. So I'm perhaps just showing this is the same in case of influence in that regard. It just depends on how one wants / needs to treat science vs influence vs production

Edit: OK I guess I'm not sure if by that 50% you mean my calculation on artists and artisan policy or the 200% bit

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Fixing Influence - any ideas welcome

#55 Post by Oberlus »

BlueAward wrote: Fri Aug 12, 2022 6:33 pm I mean there's still this side of the coin that with species of 200% science you'd need half as many planets to get same output in science, remaining planets doing production or influence as needed. Same can be said about production. So I'm perhaps just showing this is the same in case of influence in that regard. It just depends on how one wants / needs to treat science vs influence vs production
The species traits are tricky because they do not apply to all the bonuses from other content. Check out priority of each effect (in the scripting code): those with priority after first scaling priority are not affected by the traits. These include around half the pop-based bonuses and all or most of the flat bonuses. In the end, that means that Great Researc (200%) add up to an overall 130% research compared to Average Research.
I'm searching for an old thread with those calculations...

wobbly
Cosmic Dragon
Posts: 1874
Joined: Thu Oct 10, 2013 6:48 pm

Re: Fixing Influence - any ideas welcome

#56 Post by wobbly »

The bonus for artisan's workshop could be reduced to 1 or 2 and it'd still be quite strong on an empire with all artistic species. Maybe 2 is the starting pt to test?

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Fixing Influence - any ideas welcome

#57 Post by Oberlus »

+1 for that "+2"

BlueAward
Vacuum Dragon
Posts: 646
Joined: Mon Aug 08, 2022 3:15 am

Re: Fixing Influence - any ideas welcome

#58 Post by BlueAward »

wobbly wrote: Sat Aug 13, 2022 7:42 am The bonus for artisan's workshop could be reduced to 1 or 2 and it'd still be quite strong on an empire with all artistic species. Maybe 2 is the starting pt to test?
Looks like I made "fix influence" into "nerf artisan workshops" :) I mean not like you were not aware they're strong. Guess I'd like to see more ways to get higher influence, not fewer, and it still makes stronger influence generating species even better regardless if artists or not

Anyway... If y - amount of planets on influence focus under artisan workshops (having artistic people), x - amount of planets on influence focus without artisan workshops:

y = a * x

So you're after the factor a of how much artisan workshops improve things. Going by my earlier calculations

a = I / (I + A)

Where I - average influence per planet, A - artisan workshop bonus. We put I as sqrt(30) going by Oberlus's judgement. Not sure how close to reality it is, but I'm going with it (my initial thought though is that at least you'd actually want to put influence focus on planets that are above average influence producers, so sqrt(30) may be too little in that regard, but we have what we have now)

This comes to the 57.8% I mentioned earlier in current case. If you plug A = 2, you get 73.3%. For A = 1, it is 84.6%. Can also go other way around and if you want a = 75%, then A = 1.82574.

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Fixing Influence - any ideas welcome

#59 Post by Oberlus »

Nerfing Artisan Works. is necessary because it's too powerful compared to other choices (or other sources of influence should be boosted). If left as it is now, all empires with artisan species would abuse them and all empires without artisan species would lose their games.

Currently, it is easy to have abundance of influence early game and difficult late game. Moderate galaxy sizes are enough to put lots of strain in the expanding empires.

For 100 planets, average pop 36 (which is high, only possible if skipping the smaller planets), upkeep per colony is 4 and base influence-focused output is 6 IP. So 66% of the planets needs to be set to influence for 0 income (not ideal). If counting with great (+50% after certain PR is merged) influence species would produce 9 IP, so 44% planets set to influence. With +2 from Artisan Works instead of great influence, producing 8 IP per colony, 50% of the colonies needs to be influence-focused (notice that early game, when populations are smaller, the bonus from Artisan Works. is much stronger than these figures).
I think that can be an acceptable balance between the advantage from good/great influence species trait and the advantage from artistics species + Artisan Works.

For 400 planets (I'd like to allow players to have 400 owned planets...), upkeep per colony is 8. This would force 100% colonies set to influence for artistic or 90% for great influence species, with most of the PP and RP coming from unfocused bonuses.

If we boost current sources of influence to ensure that big empires (100 planets) can have more planets set to PP/RP, or that huge empires (400+) are possible, then we get overabundance of influence early game.

I think a good solution for this is to make influence upkeep grow slower. 0.5*N^1.33 instead of 0.4*N^1.5. But Geoff expressed his preference for simple exponents (it was already a concession to use 1.5).

Or, ideally, make influence upkeep consider the size of the galaxy, so that upkeep becomes an issue only when there are less galaxy to seize. But that solution was vetoed by Vezzra and Geoff because (IIRC) that would mean that the upkeep for (say) 30 planets would not be the same depending on the size of the galaxy, and that would confuse or annoy players. I still think their argument was fallacious, because players don't really know what is the upkeep depending on the number of planets (they know if they need more or less influence), and to get actual numbers they still need to make calculations, the same calculations they could make with an equation that accounts for systems in the galaxy.

About considering galaxy size:
https://www.freeorion.org/forum/viewtop ... =6&t=11363

Trying to not consider galaxy size to tackle the same problems (and here we are):
https://www.freeorion.org/forum/viewtop ... =6&t=11819

Anyone interested on trying the "galaxy-size" version of the influence upkeep?

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Fixing Influence - any ideas welcome

#60 Post by Ophiuchus »

How about make a game rule setting for "number of empire-owned planets where you should stop consider growing" which might equal 100 according to your reasoning with current balance.

So if one doubles the number of systems one could double the new setting for a similar feeling of influence cost based on percentage of colonized universe.
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

Post Reply