The Cohesive Values virtue is possibly nearly useless

[KrikkitTwo]

Well, in ciV -10% discount to policies were useless itself and I won't be suprised it's still the case in BE. Although we don't know actual formula for virtue cost.

Biggest difference is the wide/deep synergy bonuses.

In civBE an extra 'useless' virtue may get you one or two synergy bonuses.

 

[Acken]

Maybe it's useless but it's a prerequisite to many attractive knowledge virtues. Maybe you just break even, in that case see it as a road bump for something more juicy.

 

[joncnunn]

Maybe it's useless but it's a prerequisite to many attractive knowledge virtues.

If the policy is already a required prereq to the more attractive knowledge ones, then the 10% bonus will benefit those players. (They'd have taken it even if was a policy which did absolutely nothing at all)

But if you weren't planning on taking anything its a required prereq to nor using it as part of a synergy, then it would be a waste of a policy much like the Vanilla / G&K policy if not planning on completing the tree. (This also may have been in BNW initial release prior to the fall patch as well.)

 

[Olodune]

As additional benefits:

Doesn't Franco-Iberia have a UA that this would synergize with?

If there is a mechanic like Civ 5's Prora, etc, then there could be additional value.

Also, it means knowledge gets deep into its tree earlier, some of the tier 3 stuff looks decent. (If you think in Civ5 terms: I would take a "useless" policy in an opening tree to get me through rationalism faster in the mid-game).

 

[happybjorn]

Wonders could add many complications. I'm somewhat interested to find out if any have Virtue requirements (to force more diversity into the game and discourage hoarding).

 

[Acken]

I still think new cities changes the situation. The costs for second and third virtues with one city were: 24/36 and with two cities they were 26/39. So when the Virtue costs with new cities are going higher and higher the -10% saves you more and more culture. Or is there something Im missing?

I'm guessing that the formula that defines the virtue cost increase for each extra city returns a factor which is used to multiply the virtue cost. So the number of cities would make no difference in a comparison.

They always factor in. There's even a policy that reduces the cost increase for number of cities.

Just because you think it's a boring policy doesn't make it useless. Bias is bias of course.

It is purely a numbers game ofc, but a 10% reduction on a 10 city empire is worth more than a 10% reduction on a 4 city empire.

Regarding the debate for whether number of cities intervene, instead of saying "I'm right" and giving a very vague explanation here is the framework to work with, if people want to extrapolate. Edit: due to it being a longer post than intended I've made a simple tldr, for details see in the spoilers.

Tldr: Zet's argument is a bit weak and may confuse people, the factor due to city penalty isn't applied to all virtue (past and future) and varies across the length of the game so you cannot simply remove it from both sides of the equation and conclude the comparison to be the exact same. However, the effect of really taking care of the number of cities and when they were settled has a VERY limited effect on the breaking point. So Zet's conclusion still holds, whether you make 10 or 2 cities, the breakpoint will still be in the 30-31 virtues area. By breakpoint we mean the virtue number where you "refund" the cost of spending one in Cohesive values (the time where it finally pays for itself).
But like I said in an earlier post the intrinsic value of this comes from the fact that it's a prerequisite and 30virtue appear to be possible in that game according to devs, so you can consider it as some sort of a road bump to better virtues. You also get 1 point toward synergies.

Please feel free to criticize or prove me wrong.

Spoiler :

Notations:
C_k is the virtue cost of the k-th policy, before modifiers from cities and virtues
V_k is the multiplicative factor due to additional cities, for the k-th policy
q_k is the reduction for k_th policy, here q_k will be equal to 0.9 once the player takes the cohesive values virtue (after the 4th).
N_1 is the number of virtues invested taking Cohessive values, after T turns
N_2 is the number of virtues without that virtue after T turns (same time frame)

Note: Civ5 simply multiplies the base cost by an additive factor: V_k = (1 + p * (n_k - 1 ) ) where p is the penalty for additional cities and n the number of cities. By default p = 10% = 0.1.

Therefore the k-th policy always cost q_k*C_k * V_k if you have the virtue or C_k * V_k without (since q_k would always be equal to 1 here).

After a specific amount of time the total cost is:
With the virtue:
Sum(from 1 to N_1) q_k*C_k*V_k
Without the virtue:
Sum(from 1 to N_2) C_k*V_k

The "breaking point" would be the moment where N_1 = N_2 + 1 (so that you have "refunded" your investment) and the first sum to be smaller than the second but having 1 more policy.

What does this mean ?

This mean that the following argument:

the virtue cost increase for each extra city returns a factor which is used to multiply the virtue cost. So the number of cities would make no difference in a comparison.

is wrong because it assumes V_k are the same for all k, so that they do not intervene in the comparison. Or in other words it assumes that the increase in policy cost from the number of cities is retroactive. Which is not right. Or it assumes you have planted all your cities before taking the virtue, which is not right either.

However, does it mean Maddjinn is right ? Yes the number of cities intervene but how exactly, does it really make a big difference ?The question would be: does it really change the OP conclusion.

I cannot provide an analytic response to it without a specific pattern for when cities are planted. But if you put these into Excell and then play around with varrying numbers of cities at different timings you will see the actual effect of cities.

The other question would be whether or not the 20 + (2n)^2 is the right formula. I used a general notation so my argumentation shouldn't be modified a lot if it is another formula, but values given by examples will have to be changed.

To provide an example, if I settle 1 city before virtue 5, 1 more before virtue 7, 1 more before 8 and 1 more before 9 I get the exact same breakpoints as OP (N_1 = 30). However if I settle later, say at virtue 7,10,12,14 then the breaking point is actually only one virtue later (N_1 = 31). More tests reveal that a breaking point at virtue 30 is the best scenario and at most you will change the breaking point by 1 virtue.

So to go back on what Maddjinn said. City number intervene yes. Is it better for 10 cities rather than OCC ? No, breakpoint is always in the 30-31 virtues area. If by that we assume the control would of course have produced as much culture and played the exact same of course.

 

[Aeson]

What the cohesive values player will be getting earlier is the Synergy bonuses.

ie CV player Always has X total Virtues before the control player does.. so they get the X Virtue Synergies (either wide or deep)

Also ~30 Virtues is said to be standard (45 in a long game..not slow game speed but one that you win soon)

1. The CV player will get just as many 'useful' Virtues (although some will come later)
2. The CV player will get earlier (or more) Synergy Virtues

u
They could increase it to make the math slightly better, but it is still worthwhile (if you plan on the game going 350-400 turns OR you are going culture Heavy)



(New cities don't really change the comparison....at whatever time you build the new city, you can model it as slowing culture production)

The big issue is the
20+(2n)^2 v. 0.9 *(20+(2(n+1))^2)

As n goes up it becomes a better and better comparison

ie all of them cost more because you have one more.

Basically if you are going to get ~40 Virtues in the game it is Definitely worth getting.... but it is a long term investment.

So Firaxis have hinted that you can probably get around 45 virtues in a game, and that likely means top players will be able to get more once strategies become well fleshed out. So hardly useless. Its just a matter of fitting your path to the circumstance. If its going to be a long game you almost surely want it, its only a question of when (likely rather early)

 

[ggmoyang]

The break-even point is where a player taking Cohesive Values would have 1 extra virtue over a player that doesn't. That assumes the first player going straight for Cohesive Values.

The way I see this, that's a period of time corresponding to 25 virtues where the player taking Cohesive Values is at a disadvantage.

When ciV vaniila came out, there was Free Speech policy that gave -25% discount to policies.

I thought same as OP, and tried to find the point where the player can get 2 more policies. There wasn't such point even with -25% discount, due to exponential cost of policy.
Back then the formula was (25+(6n)^1.7), but it was changed to (25+(3k)^2). I don't think this would change the outcome, since 18th and later policies cost more with latter.


Now in BE, it's only -10%. Good thing with we got some synergy bonus, among them 8 tier 2 virtue gives a free virtue. If we see synergy bonus as an additional virtue from Cohesive Values, it could be good... but we don't know. We need exact formula for virtue costs. At least it's quite cheaper than policies, which is good.


EDIT: Checked virtue costs in Solargamer's video, it was 20, 24, 36, 56, 84. I think OP made right assumption there.

 

[Zet]

What does this mean ?

This mean that the following argument:
is wrong because it assumes V_k are the same for all k, so that they do not intervene in the comparison. Or in other words it assumes that the increase in policy cost from the number of cities is retroactive. Which is not right. Or it assumes you have planted all your cities before taking the virtue, which is not right either.

However, does it mean Maddjinn is right ? Yes the number of cities intervene but how exactly, does it really make a big difference ?The question would be: does it really change the OP conclusion.

I cannot provide an analytic response to it without a specific pattern for when cities are planted. But if you put these into Excell and then play around with varrying numbers of cities at different timings you will see the actual effect of cities.

Yes I was a bit hasty to respond to Maddjinn. I updated my program so that it can simulate cities being founded.

The results are that founding cities always pushes the break-even point back. Or in other words, Cohesive Values is worse the more cities you found.

A city or two hardly make a difference, but with cities being founded at 7, 11, 18, 20, 25 virtues for example, the break-even point is at 34.

The formula I'm using is 1.2^(k-1) * c where k is the number of cities (1.0 for k=1, 1.2 for k=2, 1.44 for k=3, etc), and c is the aforementioned virtue cost formula. This is just a guess.

The raw simulation output is here.

Spoiler :

1. Control: cost 20, sum 20; CV: cost 20, sum 20
2. Control: cost 24, sum 44; CV: cost 24, sum 44
3. Control: cost 36, sum 80; CV: cost 36, sum 80
4. Control: cost 56, sum 136; CV: cost 56, sum 136
5. Control: cost 84, sum 220; CV: cost 76, sum 212
6. Control: cost 120, sum 340; CV: cost 108, sum 320
7. Control: cost 164, sum 504; CV: cost 148, sum 467
City founded. Culture cost increase due to number of cities is now 1.20
8. Control: cost 259, sum 763; CV: cost 233, sum 700
9. Control: cost 331, sum 1094; CV: cost 298, sum 999
10. Control: cost 413, sum 1507; CV: cost 372, sum 1370
11. Control: cost 504, sum 2011; CV: cost 454, sum 1824
City founded. Culture cost increase due to number of cities is now 1.44
12. Control: cost 726, sum 2737; CV: cost 653, sum 2477
13. Control: cost 858, sum 3595; CV: cost 772, sum 3249
14. Control: cost 1002, sum 4597; CV: cost 902, sum 4151
15. Control: cost 1158, sum 5755; CV: cost 1042, sum 5193
16. Control: cost 1325, sum 7080; CV: cost 1192, sum 6386
17. Control: cost 1503, sum 8583; CV: cost 1353, sum 7739
18. Control: cost 1693, sum 10277; CV: cost 1524, sum 9263
City founded. Culture cost increase due to number of cities is now 1.73
19. Control: cost 2274, sum 12551; CV: cost 2047, sum 11309
20. Control: cost 2530, sum 15081; CV: cost 2277, sum 13586
City founded. Culture cost increase due to number of cities is now 2.07
21. Control: cost 3359, sum 18440; CV: cost 3023, sum 16609
22. Control: cost 3699, sum 22139; CV: cost 3329, sum 19939
23. Control: cost 4056, sum 26195; CV: cost 3650, sum 23589
24. Control: cost 4429, sum 30624; CV: cost 3986, sum 27576
25. Control: cost 4819, sum 35443; CV: cost 4337, sum 31913
City founded. Culture cost increase due to number of cities is now 2.49
26. Control: cost 6271, sum 41714; CV: cost 5644, sum 37556
27. Control: cost 6778, sum 48492; CV: cost 6100, sum 43657
28. Control: cost 7306, sum 55798; CV: cost 6575, sum 50232
29. Control: cost 7853, sum 63651; CV: cost 7068, sum 57299
30. Control: cost 8420, sum 72071; CV: cost 7578, sum 64878
31. Control: cost 9008, sum 81079; CV: cost 8107, sum 72985
32. Control: cost 9615, sum 90694; CV: cost 8653, sum 81638
33. Control: cost 10242, sum 100936; CV: cost 9218, sum 90856
Break-even reached. The Cohesive Values player has spent 100656 culture to gain 34 virtues, but the control player has spent 100936 culture to gain 33 virtues

 

[ggmoyang]

Virtue cost increment for number of cities is additive. ( +10% per city, rounded down)

 

[Acken]

The formula I'm using is 1.2^(k-1) * c where k is the number of cities (1.0 for k=1, 1.2 for k=2, 1.44 for k=3, etc), and c is the aforementioned virtue cost formula. This is just a guess.

You can look into my post again, you'll find the Civ5 formula for cities penalty. Your formula is a lot more than what it really is (since yours is a compounded penalty, while in reality it's only additive if it follows civ5 design). Not that it would really change your argument though.

 

[Zet]

Virtue cost increment for number of cities is additive. ( +10% per city)

How do you know?

 

[Acken]

How do you know?

Someone claims to have seen the 10% in a tooltip so it's reasonable to assume it follows civ5 formula.

 

[Knut_Are_M]

20-24-36-56-84 has a less steep curve then civ 5 had. where is was 25-30-60-105-170
standard speed. or 15-20-40-70-115.(quick)
or in persentage increase per policy/virtue: 20-50-55-50/20-100-75-61 / 33-100-75-64.

 

[Inhalaattori]

There is also Virtue called Memeweb: -50% Culture penalty from number of Cities for new Virtues. Is that Virtue more or less useful then?

With 2 cities build after first virtue the costs were: 20 - (new city) - 26 - 39.

I assume Virtue cost increment for number of cities is additive +10% per city.

If you conquer/build huge empire of 30 cities that would theoretically make your Virtue costs to rise more than 17-fold (1,10`30=17,45). The saving of culture for each new virtue would also go up, because this virtue would help to cut up the rising cost of new virtues from building/conquering new cities. Because you would get these new cities in random time between new virtues its practically impossible to say what the actual costs would be.

Btw They are still clearly working with the conquer mechanic so conquering city in Solargamers video did not change health or culture/science costs. It didnt even give option to puppet/annex/raze.

People have to also remember that you get free virtue from Tier 2 synergy. "Cohesive Values" gets you closer to that synergy bonus.

 

[3of5]

Thanks to Zet for starting this thread, since this is also a concern of mine. I did not make a such thorough calculation as Zet and Acken do, and it seems i even underestimated how useless a virtue this is. In Civ5 vanilla and G&K they had still limited use, because one needed a tot number of virtues for a winstate (culture victory), or they increased culture output by % which had other uses too, but this one just seems to be a rather annoying handicap to take when going through the knowledge tree. Even people who know how bad a virtue this is, will be taking it, because you need it to go farther down and if you have to take it, you better take it early. So playtesting would actually show it as a virtue that people beeline to, which only shows how one could misread that data. Not that other trees don't have useless virtues to bybass (if you took a worker with you at the seeding, you'd be hard-pressed on which one would be more useless: getting a free worker or a smaller KV bonus on Outposts). Just take it and assume that one outpost might have fallen although one should have 2-3 units at that time in the game (or could just rerout one trade route if the target reagion would be unstable). Same goes for +2 production from manufactories (-2 health). The health malus makes factories kinda meh, unless you have health to spend (in which case you should have build an additional city, not manufactories). So i guess every three might have some useless virtues one can't bypass, this one at least lets you recover part of its cost.

What really does bothers me though is that the devs did not see at first glance how lackluster the virtue is, which means that they seem to be oblivious to the mechanics of the game that they are "modding" and balancing (i think BE deserves to stand as its own game, but the type of work put into it is more comparable to a total conversion than a new game. This does not mean that BE is a ripoff to get your money, it means that total conversions are labor intensive and wonderful thinks that should get more recognition). Think about it: would you trust an engineer that seems to not be aware of how the nonlinear formulas he applies work mathematically in the complex model of say a wastewater treatment plant or a bridge?

 

[happybjorn]

Opening with knowledge 1-3-5 (5 is Creative Class, earn extra culture equal to 50% net positive health), then going for Cohesive Value should be superior to beelining Cohesive Values in most situations.

This may shift the breakeven point from the original argument, but also makes maintaining the other things being equal assumption more difficult.

 

[Inhalaattori]

You need this virtue to get to the one of best virtues of the game: Metaresearch Methods (Leaf technologies cost 20% less science)

Maybe they should change them both to be at 15%. Would make things more balanced.

 

[Zet]

You need this virtue to get to the one of best virtues of the game: Metaresearch Methods (Leaf technologies cost 20% less science)

Maybe they should change them both to be at 15%. Would make things more balanced.

With a 15% reduction, the break-even point is at 20 virtues with 1 city. With multiple cities, it will come a bit later.

I still think that this virtue is a bad concept. The break-even point assumes that the player goes straight for Cohesive Values. A player who takes it about halfway through the game (lasting about 45 virtues) will gain next to nothing from it even at 15% reduction if he founds some other cities.

This virtue should at least be optional, and not a prerequisite for later ones.
Even better if it was replaced with something that doesn't have this type of problem.

 

[3of5]

There is also Virtue called Memeweb: -50% Culture penalty from number of Cities for new Virtues. Is that Virtue more or less useful then?

That depends on how many wonders you have built. Honestly, with the tall playstyles i favor, research becomes a non-issue in the late game and i never bother much with Akademies and Manufactories too (even though i have tons of great people usually). So i don't see me building many of those in BE either. Therefore the whole right arm in Knowlede does not feel that rewarding to me, and i would only take it if i had so many wonders the 7 culture would ammount to something useful (and with the nature of the tech web this is less likely than it was in Civ5, where you'd go upper route or lower route depending on military or wonder focus and just built everything once you gained enough of a tech lead).

Also: Happybjorn, you don't want to take "Creative Class", because its even worse than cohesive values if you have 6 health lying around, please found another city, after 50 turns it will have you repayed the culture you missed out on and hey: a new addional city is all kinds of beneficial. And no you will probably not be building biowells to increase your culture through health (that would be 4 energy per culture). When you build biowells then you're planning on filling those health with actual population that is working tiles and buildings (remember that 4 pop give one culture), so you're only missing out on 1 culture every 4 biowells. Generally you will just want to hover around +1-5 health and be fine with it, which is a flat +2 culture bonus throughout the whole game due to "Creative Class". Its not that much, it scales horribly and it disincentivises you further for going negative health, even when that would be a good strategy (mid game expansion rush).