1. We have added a Gift Upgrades feature that allows you to gift an account upgrade to another member, just in time for the holiday season. You can see the gift option when going to the Account Upgrades screen, or on any user profile screen.
    Dismiss Notice

Finding the proper difficulty setting

Discussion in 'Civ3 - General Discussions' started by justanick, Apr 19, 2015.

  1. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    So did i. Therefore i copied this discussion so we can continue it in civilised manner without flooding the 'Newbie' Questions thread with it. :)
     
  2. Buttercup

    Buttercup King

    Joined:
    Oct 20, 2011
    Messages:
    812
    Well, whenever I mention about how HoF only records one specific form of difficulty setting I tend to get set on like an extra in a horror movie.

    The point I make is that the game has lots of interchangeable difficulty settings before you even choose what level you wish to play at. You press New Game and you get choices:

    Size: Tiny, Small, Standard, Large, Huge, Random
    Barbarians: None, Sedentary, Roaming, Restless, Raging, Random
    Landmass: Pangaea, Continents, Archipelago
    Water Coverage: 60%, 70%, 80%
    Climate: Arid, Normal, Wet
    Temperature: Warm, Temperate, Cool
    Age: 3 Billion, 4 Billion, 5 Billion

    Which all effect the difficulty of the game you'll be playing and how you'll approach the game - possibly even making random the hardest setting as it prevents you from any kind of initial planning mindset.

    Then you get a whole new page of choices:

    Your Civilisation - One of 31 + Random
    Opposing Civilisation - Between one and fifteen rival either of your choice or Random
    AI Aggression - Least aggressive, Less Aggressive, Normal, More Aggressive, Most Aggressive.

    Then, and only then, do you get the level choice of:

    Chieftain, Warlord, Regent, Monarch, Emperor, Demigod, Deity, Sid.

    So if someone plays easy setting, easy setting, easy setting, easy setting, easy setting, easy setting, hard level

    it is different to someone playing hard setting, hard setting, easy setting, easy setting, hard setting, hard setting, easy level

    The game is quite specific/obvious about what base setting the game is designed to played at:

    Standard - Roaming - Pangaea - 70% - Normal - Temperate - 4 Billion - Any - Culturally Linked - Normal - Regent (no bonuses to either you or the AI)

    So, without changing the Regent setting, you can make your game 'harder' (in terms of completion date before 2050) by choosing:

    Huge - Raging - Archipelago - 80% - Arid - Cool - 3 Billion - A civ that is not suited to the environment - Selected good civs for the environment - Most Aggressive - Regent

    And you have a completely different set of expectations for finish date and recommended advice of how best to take advantage of your environment, tricks and techniques.

    And then, the final roll of the difficulty dice is start location. From isolated pure Tundra island to an island with a fair bit of Grassland and a river or two and all inbetween, to which, IMO, all HoF games should have a difficulty identifier next to the game to show how easy/hard the starting location was, some kind of 1-10 scoring system where 1 is easiest and 10 is hardest. This would be more subjective but, for the experienced, it should be fairly obvious how easy/hard a start is depending on all the factors involved - such as luxury/resource/freshwater availability, at capital, nearby, far, none/conquest/distant exploration, how close/far other civs start to you and who they are etc etc.

    My conclusion being that the common advice that "staying at the lower levels teaches you bad tricks" is relevant because the current HoF system encourages its own bad tricks by demanding people only ever train themselves in easy settings and reloaded, always the same starts.

    A good example of this was when I posted a an interesting Deity tiny map and both Spoonwood and VMXA had a go at it. Spoonwood, the HoF pro had a lot of difficulties initially, whereas VMXA, just a forum pro, completed it first and with the least grief. Because the HoF teaches a very narrow path, like cramming for a specific exam, whereas the whole universe of Civ, which covers every aspect of the game's potential, involves a lot more adaptiveness and a larger leeway for error than simply Map size, victory condition and level.

    For example, if you start on a pure Tundra island, then the age old adage of Workers, Workers and more Workers, would be entirely incorrect, as Tundra has some pretty steep limitations to Worker requirement.
     
  3. Lanzelot

    Lanzelot Moderator Moderator

    Joined:
    Jun 25, 2007
    Messages:
    5,769
    Location:
    Heidelberg
    Good idea! :goodjob:

    Buttercup really makes a good point here. (Which basically coincide with my reason for not playing for the HoF. And which is the reason why I like the GOTM competition so much: in the GOTM you will get good starts, bad starts, Pangea, Archipelago, powerful agricultural civs and weak outsider civs, and what ever the game designer throws at you, you will have to try and make the best out of it... (Anybody still remember the "Warlord" GOTM 102? :D That was much harder than the "Demigod" COTM 112 we just finished...)

    But I have two objections:
    a) Choosing Archipelago, 80% water instead of Pangea, 60% water doesn't necessarily make the game harder. ;) On the higher difficulty levels, Archipelago, 80% will be much easier for the human player, because the AI will be isolated until Navigation and will drown in unit upkeep on their small islands, while the human player can benefit from tech brokering. On the lower difficulty levels, however, Pangea, 60% is easier for the human player, as he can quickly swallow a couple of neighbors and expand that way.
    b) Not all HoF players are bad. In fact, I think there are some HoF players who are also pretty strong "alround players" that would no doubt manage to get the optimum form a tundra start as well.

    So what is the "right" difficulty level? Let me try and offer an answer:
    "The one where you win half your games and lose the other half."

    If you win more often, it's too easy, and if you lose more often, you'll get frustrated and de-motivated, so you will probably neither be able to improve nor enjoy your game.
     
  4. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    That sounds about right.

    One important setting Buttercup did not mention is cultural flips. If you deactivate those Sid will play like Deity etc.. If you donnot have to care about culture, can spend all shields on military instead and can overtake everything AI has generously build up for you without risking cultural flips, than things get a lot easier.
     
    Buttercup likes this.
  5. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    I thought I'd revive this thread rather than start a new one - given excellent post by Buttercup. I play with a friend and both of us find vanilla Emperor far too easy now, but the next difficulty level up way, way too hard. We are fairly casual players and don't seek to be grand masters knowing precise micro-strategies and all that. We just play for fun. Multiplayer, 6 AI, 2x humans trying to play games where either human or any of the AI have a chance of victory. Given the multiplayer limitation we play small and standard maps (usually archipelagos with 60% water and me customising the map to ensure fresh water supplies on every island, but really land type doesn't seem to make things easier or harder).

    I tried giving the AI more starting units (workers and warriors), but that isn't difficult enough. I then tried giving the AI an extra settler and it pummels us into submission. My friend also hates raging barbarians (which too me is the next most obvious difficulty tool) and if there is too much corruption or rubbish land the game isn't fun for us (so I want to avoid that). Even if I play inappropriate Civs for the map and give the AI the best Civs for the map, they still don't give us a challenge in vanilla Emperor.

    So my question would be, can anyone suggest the most suitable options for customising the game settings in the editor to create a big step up in Emperor difficulty that isn't:
    - raging barbarians
    - increased player corruption
    - deliberately rubbish land
    - an extra AI settler

    An alternative solution might be to give the AI an extra settler, but then make the AI weaker than Emperor through other rules (incase that is easier).

    Also the single player aggression slider doesn't seem to be available in multiplayer. Is it somewhere in the Editor, or would I need to manually adjust the aggression of all 31 Civs?

    And finally. Would 8x Civs on a smaller map be considered harder or easier than 8x Civs on a standard map? Or is that just not relevant?

    Thanks in advance for any suggestions.
     
  6. Lanzelot

    Lanzelot Moderator Moderator

    Joined:
    Jun 25, 2007
    Messages:
    5,769
    Location:
    Heidelberg
    You could assign extra starting techs to the AIs to give them a head-start.
    Or you could give them 3 traits instead of 2.
     
    Fergei likes this.
  7. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    In my opinion the best way to adjust the difficulty settings is to remove all the extra starting units for AI.

    -remove all extra starting units for Monarch to Sid.
    -reduce the cost factor for Monarch to Deity by 1. So that is 8 for Monarch, 7 for Emperor, 6 for Demigod, 5 for Deity and still 4 for Sid. No need to make Sid harder.

    This will give the AI no initial advantage, but only a long term one. Anything between Monarch and Deity should play reasonably well for reasonably good but not overly strong players.
     
    tjs282 likes this.
  8. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    I think we might have an early winner here! I didn't know what that Cost Factor meant until I looked it up just now. Something that makes the mid and late game harder has been the holy grail for me. I either get destroyed very early (too hard) or survive and become comfortable in the Middle Ages with almost no danger of being destroyed. Extra starting units increases the risk of being destroyed early, whereas your suggestion means it'll be a more even start, but that the AI will continue to have advantages throughout the entire game. This hopefully means I won't get to that relaxed feeling in the middle ages but will be kept on my toes.

    Plus I hated extra starting units because I felt it kind of undermines the raison d'etre of the Expansionist trait (already weak) if the non-expansionist Civs all run about with lots of units straight away (popping huts and making early contact with other Civs). Thank you, I'm going to test this today with my fussy friend!
     
  9. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    I may be going mad, but I've never really struggled on Emperor on small maps, and reducing the AI costpoint worked well and made it a very difficult challenge. So thanks again for that advice.

    But as soon as I go on standard size, even without raising the difficulty but reducing the cost point, it becomes a lot more difficult. My friend and I usually have to bail on the game early on or join forces to just barely survive for as long as we can.

    We used to only play on small but are transitioning to standard size, so I think I'm going to have to create two different difficulty levels.

    - Emperor minus one cost point for small maps.
    - Monarchy minus one cost point for standard maps.
     
  10. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    Fergei likes this.
  11. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    A very interesting formula, thanks. I think I may also be reducing difficulty on small maps by squeezing in 8 Civs rather than the default amount of 6. Which also doubtless messes up that research formula.
     
  12. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    No, the formula will still work fine.
     
  13. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    Having done testing on this, and at the risk of stating something obvious, I thought I'd mention that the way that I'm finding the most effective to test difficulty settings to find a suitable combination for me (up to Emperor difficulty at least). I host a multiplayer game with 7x AI in it and I load up a map with my custom difficulty settings. That way you see the Civ scores of all the AI from the first turn and can compare yourself against them (and better take into account any blips they might have that would otherwise skew your conclusions). After 30-60mins I can generally evaluate the difficulty as follows.

    i) If I'm in the top 4 scores, the difficulty will be too easy and the AI will be unable to have much influence on how I choose to play the game.
    ii) If I'm last and say 10% lower than the 7th placed AI, it is going to be too difficult and I'm probably not going to survive more than a couple of very unrewarding hours.
    iii) the sweet spot is being 7th or last but very close to 7th. This seems to result in an engaging challenge where I have a fight on my hand to secure sufficient territory and might have to resort to military means. Win or lose, I will be stimulated and entertained.

    I see, because the formula applies to both the AI and player. From testing, I think there is definitely a relationship between available land and difficulty. The lower the amount of available land per AI, the worse they seem to perform. I have no explanation for that (perhaps the AI struggles to work out what to do when it runs out of space and is too slow to transition to military action or colonising islands?), but its the rule of thumb I'll be using now when tweaking difficulty. All things being equal, a 60% water seems to be a tougher challenge than a 70% water, a standard map will be a tougher challenge than small. So that's when I'll start messing with cost factors, barbarian settings, or hand selecting weak/strong opponents.
     
  14. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    Those scores are not a good indicator.
     
  15. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    I have to disagree. The higher the difficulty level, the faster the AI scores rise. If they have extra starting units or a lower cost factor (as per higher difficulties) then the score will rise faster.

    A players performance relative to those AI scores is entirely an indicator of their likely relative performance; both later in that game and generally in a different game with identical game settings.
     
  16. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    Yes. In fact the higher the difficulty setting, the higher your scores rises, too.

    Early on score is mainly an indicator for accumulated territory weighted by the amount of turns. Later on population and happiness matter. But excessive happiness is not a good thing and a larger than useful territory is not that good either. It is quite possible to be clearly inferior in score, but building up a decisive advantage against the competition nonetheless. Productivity in F11 would be a more useful indicator as it is counting your yields past corruption and food consumption.

    https://forums.civfanatics.com/threads/maximizing-your-score.18729/
    https://forums.civfanatics.com/threads/how-the-demographics-works.37033/
     
    Fergei likes this.
  17. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    Very helpful as ever. I suppose to clarify, when I say I am running the game against the AI for 30-60 minutes, I am very blandly repeating the expansion phase only. I'm not amassing armies (power but no score), I'm not doing infinite city sprawl (amassing population without territory) and I'm not really doing anything different from game to game. I try to keep my performance as consistent as possible because the purpose is to evaluate difficulty after 30-60mins then stop. This combined with the relative consistency of AI behaviour in those first 30-60mins (i.e. focussing on their own expansion) means that the amount of variables at play in those initial 30-60mins feels a lot less than if I'd extended the game time to 60-90mins (when warfare would, as you point out, really undermine any conclusions to be drawn by comparing 1x human score against 7x AI scores).

    I appreciate that the level modifier for the human players scores makes drawing conclusions more challenging, but I can't think of a better way to go about it and I'm happy with the results so far (I had been really floundering over getting a well balanced challenge prior to reviving this thread). Despite that modifier, it is reasonable to say that the human player's score relative to the AI will, on average, decrease as the difficulty level rises. So it can be factored in to calculations of potential future difficulty by looking at the human score relative to the AI score. I would never propose comparing the human score to human scores (or the AI to AI score) across difficulty levels or even across games on the same difficulty level (as neither will tell you a whole lot and there are so many variables).

    If you are settled on a single difficulty level and simply modifying all of the other available variables, then the difficulty modifier in the 'maximising your score' thread equation will presumably not come into play.
     
  18. Lanzelot

    Lanzelot Moderator Moderator

    Joined:
    Jun 25, 2007
    Messages:
    5,769
    Location:
    Heidelberg
    Isn't infinite city sprawl "amassing territory without population"? :confused:
     
  19. Fergei

    Fergei Chieftain

    Joined:
    Mar 31, 2020
    Messages:
    59
    I will take your word for it as it isnt something I do. I am working on the assumption that most players doing infinite city sprawl are not creating cultural buildings (expanding borders) and are possibly using forced labour (and therefore may score lower in a 30-60 minute game start than using more AI-esque expansion rules in terms of city spacing and culture). Maybe I'm wrong on that and they score at a similar rate (even as ICS gains more unsecured power).
     
  20. justanick

    justanick Emperor

    Joined:
    Oct 6, 2010
    Messages:
    1,082
    Location:
    Germany
    I would disagree. The amount of worked tiles per covered tiles will be higher than if settling farther apart. So "amassing population without territory" is a valid if oversimplifiying way to summarize it.

    The idea is to amass economic output and ICS implies that rank corruption hits in when covering less territory. Hence one needs to get more out of this limited territory. Overlapping use of tiles with 3+ food is part of that as is maximizing the amount of towns at fresh water. The Maya are suited well for this approach, as are the Iroquois with their lowered rank corruption.

    Forced labour will hardly pay off unless you have reached the corruption limit. But any sensible player will not do that much unless going for 100k culture victory. At low corruption fored labour will avoid more output than it gains. At high corruption the amassed unhappiness is a serios problem that is reduced by only 1 per 20 turns, hence whipping only 1 citizen way per 20 turns is viable in the long run.

    I would not do a true ICS. But while still in despotism it can be reasonably close to the idea. Much later in the game a few abundant cities will be abandoned to get to something like 16 tiles per metropolis.

    The main idea is to get the economy going in the first 150 turns. Get effectivity even if that is at the expense of efficiency. After about turn 150 the focus changes towards efficiency, but working all tiles available will still be a paramount concern.
     

Share This Page