Variable difficulty

[void_genesis]

I had an idea for a mod after realising the most fun games are the ones where you are constantly on edge as to whether or not you are going to make it. Cake walks and being stomped aren't nearly as much fun of course.

Would it be possible for a mod to shift your game between difficulty levels depending on your performance? (maybe based on the data coming out of demographics or info-addict). You could give the AI the higher level difficulty bonuses if you were about to steam-roll them. On the other hand if you bit off more than you could chew or made a major tactical error it would make finishing the game and fixing your mistakes more enjoyable than rage quitting. Apart from that all you would need is a notification when you have shifted difficulty levels (or better yet a tracking info-addict style metric that gives you an indication of impending level changes). This gives you the dual aims of winning/finishing the game but also getting there at the highest possible difficulty level for your ability.

Doable? Enjoyable??

 

[Sporhund]

Was a mod component for Civ4, where depending on your position in the overall scoreboard, it would increase or decrease the difficulty every 50 turns.

 

[SlightlyMad]

I'd prefer to still keep the same difficulty level setting as a baseline, but give a boost to any civs falling behind. A penalty could be given to the front-runner as well, but that might have the impression of being un-fun.

It could really be something as simple as granting free great people to any civ that meets the criteria of "falling behind," possibly even in the category in which they're doing so. Maybe match the timing with those ranking pop-ups you get. Each time one of those pops up, the civ with the least production gets an engineer, the civ with the fewest techs gets a scientist, the civ with the fewest culture gets an artist, etc.

 

[Dunkah]

Perhaps simply sending them into a Golden Age, call it re-birth or re-juvination or something like that.

 

[SlightlyMad]

Well, the cool thing about the boost being a GP is that they could use it for a GA or free tech or whatever. If they were really behind (in multiple categories), they'd even get commensurate benefit in the form of multiple great people.

 

[Thalassicus]

There's two ways to auto-adjust difficulty:

• Build it into the basic concepts of the game.
• Layer a new effect artificially on top of the game.

For example, in a dungeon crawl I made, there were two difficulty adjustments for small teams (to balance the fact each person got a larger share of XP/loot):

• Less people on the team means less hero abilities, so the game is fundamentally harder.
• I artificially increased the damage of monsters when there were fewer people on the team.

Do you see the difference? Type-B changes are very easy to do, but usually feel contrived to me, so I try to figure out a way to make it type-A whenever possible.



The change I made to Declaration of Friendship research in January was an type-A effect. Since it combines the research of both players, the person further behind inherently gets a bigger bonus.

• Before DoF:
  • Player 1: 400/turn
  • Player 2: 100/turn
• After DoF:
  • (400 + 100)*5% = 500*5% = 25
  • Player 1: 425/turn (+6%)
  • Player 2: 125/turn (+25%)

With just four friends the player who's behind can double their research rate. The old vanilla free-tech system was about a +10% increase for comparison, and did not favor less-advanced players. The mod's version is type-A since it's built into existing effects, by replacing the exploitable free-tech approach and improving the low value of Declarations of Friendship. This was an extension of Afforess's Tech Diffusion (which was type-B). Tech diffusion layered a new effect onto the game by picking the least-advanced players and giving them research each turn.

This month I also made another type-A change to conquest by moving the +50% XP effect out of the Honor tree. This reduces the steamroller effect of successful warmongering.



Now let's apply this concept to Great People.... thinking... I have an idea!

It might be possible:

• Free exchange of ideas from mutual Open Borders increases Great Person generation rate by N% of your trade partner's total population.
  (I think it might be possible to code! )
• If gold/production/military is lowest in score, give a merchant/engineer/general every N turns.

Notice the difference between types A and B. The type-A one operates on the same principle as Declaration of Friendship research. It's beneficial to both, but since it's partially based off your trading partner, it helps the weaker partner more. I could put it on my todo list if it sounds interesting. Thoughts, opinions?

 

[Maxii]

I like it!

Gives a benefit to opening borders which I normally never do as my military weakness becomes readily apparent. btw do you know if the AI knows your military strength anyhow?

The second component adds a piece of the auto difficulty adjustment system. As another aside, do you still have a tech diffusion component in your balance mod? If so, that is another component of an auto difficulty adjustment system.

 

[GoUtes53]

Thal,

I love your "A" ideas. I would say implement them if it seems doable and the AI seems to "get" it.

 

[Seek]

Agree the A option looks great!

 

[SlightlyMad]

I love the idea Thal, and definitely prefer that more interesting approach! The only problem is that it doesn't help a pariah or civ with an isolated start, but other than that it's a great step in the right direction.

 

[Seek]

The only problem is that it doesn't help a pariah or civ with an isolated start, but other than that it's a great step in the right direction.

Maybe the GP idea could be implemented if a civ is extremely behind in certain aspects, like it has only 25% of the leading civ's score for 10+ turns or something along those lines..

 

[Thalassicus]

I've noticed isolated starts are more rare in Civ 5 than Civ 4, and start locations are overall very nicely balanced in Civ 5, so my concern over that has been gradually waning. The opposite of the pariah situation can be thought of as rewarding the player for good diplomatic relations. In vanilla diplomacy is not really important, so I feel anything to improve this is a bonus, not a detriment.


To explain something in more detail, the risk of tying things to something that's not a fundamental built-in gameplay effect is we can create a situation where players are rewarded for exploiting the system to get better bonuses.

This is what people did with the free techs of vanilla research agreements (type B). RA's give techs based on recent research, so smart players figured out we could put a turn into techs we did not want, and control the "random" free tech. There's no way I know of to exploit the type-A DoF research effect since it's built into basic parts of the game: generation and diplomacy.

Does this make sense? I sometimes don't explain things very well, dyslexic!

Score is an artificial construct that has no effect on gameplay, and isn't a perfectly accurate measure of a player's standing. Just like free techs, a clever player would exploit it by identifying weak points. Find underrated parts of the calculation, emphasize strategy on those, and intentionally keep score low while actually being very powerful.

This is the reason behind using type-A effects that are built into basic gameplay like open borders and great person point generation. I want to avoid type-B things like basing effects on score.

 

[SlightlyMad]

Yeah, that definitely makes sense and would be a clear improvement. I generally play on PW3 maps with it set to start everyone on the largest continent, so isolated starts aren't a big deal anyways, for me. The pariah detriment is more what concerns me. Unfortunately, diplomacy in Civ5 still feels mostly out of the player's direct control, and sometimes you can end up having everyone hate you just because you founded that fourth city too early, or because your army is too small and your neighbors go hostile. Like I said though, additional benefits along the lines you speak of would still be a clear improvement, and probably less "gamey" in their implementation. I understand what you mean about the score calculation; the GP idea I was stating above would be more based on concrete measurements like number of techs or policies, but I can still certainly see players trying to min/max and game the system to get free bonuses.

 

[SlightlyMad]

I just had a thought, particularly since reading your mathematical breakdown of policy costs, it's really been bothering me that more cities is still innately better than fewer in almost every way. So it occurs to me that we could kill two birds with one stone with this idea, by both helping out struggling empires and small or "tall" empires.

What if the calculations behind the science and GP bonuses were weighted somehow to more greatly reward empires with fewer cities? The most obvious and heavy-handed way I can think of is to have that 5% could have a multiplier tied to it, based on the relative difference in number of cities. So to use your 400/100 science example, let's assume civ1 has 12 cities and civ2 has 3 cities. That's a ratio of 4:1, so the total benefit of 10% (normally 5% each per side) would instead be split according to that ratio as 8:2, and then flipped, so civ1 would get a 2% boost and civ2 would get 8%.

So instead of 25/25 being their ultimate benefit, it would become 10/40, for final science outputs of 410 and 140. The end benefit for each would be 2.5% of science rate for civ1 and 40% for civ2.

That might be a bit too strong of a benefit, but that's the basic idea. It might seem at first that this situation is simply counting the same sort of benefit twice, but that's not necessarily true. If we take for example, civ1 with 200 science output from 12 underdeveloped cities paired with civ2 with 300 science from 3 huge core cities, it still becomes skewed in civ2's favor for having the smaller empire, even though they have a greater science output. They'd end up with the same 10/40 split of the pie, giving civ1 a 5% benefit and civ2 a 13% benefit.

Just a thought, but I've always thought small but tall empires should be balance with some sort of a science bonus, and it makes sense to reward them for peace and diplomacy as well, so this seems like it could work perfectly to that effect.

 

[Thalassicus]

More cities isn't necessarily better, each one does add to great person, golden age, and national wonder costs. Science favors these empires too due to how I set up AI personalities. Peaceful players are naturally friendly ones, but if you go to war, all the peaceful players will eventually hate you and refuse to sign DoF agreements, therefore significantly reducing the science rate of larger (typically conquest-oriented) empires.

If we want to shift things towards small empires, I'd prefer to do it with the existing effects of GP/GAs/NWs. I've buffed national wonders considerably so it might be time to start focusing on great people.

 

[Stalker0]

More cities isn't necessarily better, each one does add to great person, golden age, and national wonder costs.

I would still say it adds to policy costs as well, I've never seen good evidence that suggests that a large empire can acquire policies as readily as a small one can.

 

[Thalassicus]

With Representation and a few culture buildings, founding a new city has no net effect on policy generation. I posted about it on the strategy subforum:

Policy cost increases per city

For example, look at the third column of the first table. In that situation we could have 5 cities or 50, and we'd still get policies at the same rate.

 

[SlightlyMad]

I'd always suspected that but couldn't quite figure it out, and your number crunching matches my in-game experience. The big clue that made me think something was up in games before is when I've select the policy that reduces the additional city policy cost, it sometimes seemed to have made little or no difference.

Are these values moddable? It seems obvious that they intended to increase policy costs per city for larger empires, so it seems like the value should be increased if it's not sufficient. I'd also be happy to see the golden age and great person costs to be increased more significantly with empire size, along with maybe an overall decrease in cost so that medium-sized empires would see about the same rates as now, but smaller empires would get them quicker and larger empires get them slower. There should be a point in empire expansion that basically means you're not likely to see any policies, golden ages, or great people without significant effort. Getting an empire to that point though, should mean you're basically powerful enough that you don't need them to win the game anyways.

 

[Thalassicus]

I agree about golden ages and great people, those are ways I want to focus on for buffing small empires.

I disagree about policies though, as a matter of personal philosopy. Policies are an incredibly fun way to specialize our game, and all playstyles had equal access to social engineering in previous Civilization games. Significantly restricting any particular playstyle from such an exciting part of the game doesn't feel very fun to me. I don't believe they intend policy rate to be particularly low for large empires... otherwise they wouldn't have added the new Representation effect. That effect is what makes it feasible.

 

[void_genesis]

These deeply engineered mechanisms to make the game self balancing and to make a wider diversity of playstyles viable is what I love so much about Thals modding. I recently played another mod for diversity and while interesting, it simply doesn't compare.