What Video Games Have You Been Playing #11: I should go

[Timsup2nothin]

Nah, Civ 5 (and by all I've been able to learn 6 is the same way) suffers from a much worse "coast to the end after clawing your way to parity with the AI" problem than 4 does.

That coast to the end is the greatest weakness in Civ. In most games, really. The AI should adapt to the fact it has been caught up to and start pressing.

 

[Lexicus]

That coast to the end is the greatest weakness in Civ. In most games, really.

Yeah, pretty much any 4x/strategy game where the AI is given resource bonuses (tech, gold, extra units, whatever) has this problem. Interestingly Medieval 2: Total War has a decent solution in the form of the Mongol and Timurid invasions. Admittedly it only functions as a solution if your empire is on the eastern side of the map, but it's better than nothing.

The AI should adapt to the fact it has been caught up to and start pressing.

Maybe some sort of dynamic AI bonus system where bonuses can be applied to the AI when the player reaches certain milestones? I'm not sure how well that would work though, because "parity with the AI" can mean a lot of different things depending on the exact context.

 

[Timsup2nothin]

I've thought about how an algorithm could be developed that would consider most advanced tech, total unit strength, and number of wonders possessed and assign a dynamic bonus to the AI so the further ahead (less behind) you get the more bonus they get. It also has the advantage that since each AI civ would be receiving an individually tailored bonus the weaker civs would tend to catch up to the stronger AI civs as well, since they would be getting a bigger bonus.

 

[hobbsyoyo]

I prefer AI to do a better job with the resources they have than to give them bonuses which amount to a kind of cheating. I know that's not trivial or else it would have been figured out already but I am not a huge fan of just throwing bonuses at the AI to level the playing field.

 

[Timsup2nothin]

I prefer AI to do a better job with the resources they have than to give them bonuses which amount to a kind of cheating. I know that's not trivial or else it would have been figured out already but I am not a huge fan of just throwing bonuses at the AI to level the playing field.

Problem is that "do a better job" is a function of strategy. In order to make better use of their resources the AI would have to be pursuing a specific strategy, and that strategy would have to be given to them in the programming. Chances of the programmers being able to provide the best strategy to the AI at the time of programing are really slim in the first place, plus that is totally non adaptive. Once you figure out that "the AI always plays this strategy" the game becomes about developing a specific counter to that strategy, not about freely developing your own.

 

[Synobun]

Something I liked in Civ 5 is that you could turn things around during the transition to the Industrial Era, if you were lucky. There were more than a few multiplayer games where I'd get a tundra/desert start, be at rock bottom for most of the game, and then quickly rise to the top 3 after modernization. Whenever I played Civ 4 and Civ 3, I found that the beginning mattered a whole lot more, and there was often no opportunity for redemption.

 

[EgonSpengler]

I've thought about how an algorithm could be developed that would consider most advanced tech, total unit strength, and number of wonders possessed and assign a dynamic bonus to the AI so the further ahead (less behind) you get the more bonus they get. It also has the advantage that since each AI civ would be receiving an individually tailored bonus the weaker civs would tend to catch up to the stronger AI civs as well, since they would be getting a bigger bonus.

I've imagined something similar, specifically for Civ, to somehow mimic or model the "rise and fall" of cultures and empires throughout human history. Civ is, theoretically, a game inspired by human history, but history's fundamental dynamism is entirely missing. I think the only reason the early eras are exciting and tense is because at that stage the game is unfolding on a blank canvas.

 

[Lexicus]

It's really hard to do a "fall" mechanism that doesn't just kneecap the player with a bunch of annoying stuff though.

 

[hobbsyoyo]

Once you figure out that "the AI always plays this strategy" the game becomes about developing a specific counter to that strategy, not about freely developing your own.

Giving the AI more bonuses at set phases is functionally the same as it using the same strategy in my opinion, just more ham-fisted. I get it that I'm wishing for the moon on this but I don't like dumping more resources on the AI is the best solution to the problem. It may be the only viable one at this moment in time but I don't think it's the best.

It's really hard to do a "fall" mechanism that doesn't just kneecap the player with a bunch of annoying stuff though.

I think they tried something in the development of Civ IV but found it to be pretty sucky and then they inverted it to make golden ages which are pretty awesome. Doesn't Civ VI have dark ages though?

 

[TheMeInTeam]

Civ 4 and Civ 6 are both usually won or lost by t100-t150 (with earlier moves having greater weight towards winning). However playing out the remaining turns in Civ 6 takes several times longer. This is due to a combination of factors:

• Gross lack of concern for end user experience with UI interactions/#inputs. Patches have shown minimal effort to address this lately, but Civ 6 is still behind 1990's games in basic UI conventions so I won't give them credit for this just yet.
  
• Design (Civ 4 had things like vassal spam and a different setup for diplo wins that could both rush the game to a conclusion more quickly)
• Turn times (newer civs games are slower/more poorly optimized)

All that said, even Civ 4 has some issues with games being over before they end. It's just less significant in both % of turns and in IRL time spent.

I've thought about how an algorithm could be developed that would consider most advanced tech, total unit strength, and number of wonders possessed and assign a dynamic bonus to the AI so the further ahead (less behind) you get the more bonus they get. It also has the advantage that since each AI civ would be receiving an individually tailored bonus the weaker civs would tend to catch up to the stronger AI civs as well, since they would be getting a bigger bonus.

I don't like rubber banding as a solution. It decreases the outcome difference between elite, good, average, and poor play. IMO it's better to just make the game end faster, and/or make it harder to secure that runaway position in the first place.

Whenever I played Civ 4 and Civ 3, I found that the beginning mattered a whole lot more, and there was often no opportunity for redemption.

Players would often break out from ~6 cities to 15+ in renaissance or industrial eras in Civ 4.

I think they tried something in the development of Civ IV but found it to be pretty sucky and then they inverted it to make golden ages which are pretty awesome. Doesn't Civ VI have dark ages though?

Civ 6 has dark/normal/golden and a special heroic (getting golden after dark). You can tangibly influence which you get through the choices you make, and while dark ages have some serious disadvantages (halved loyalty pressure being the most important) they also open policy cards that give a strong bonus + significant downside which can be very useful.

As such this is a legit mechanic with agency rather than some arbitrary "rise/fall" random factor. You EARN those ages, dark and golden alike. This is one of the areas I'd say Firaxis actually has improved on design and playing around era score is another resource to plan/consider/make tradeoffs.

 

[EgonSpengler]

It's really hard to do a "fall" mechanism that doesn't just kneecap the player with a bunch of annoying stuff though.

1. Buildings (or Districts, in Civ VI) that produce any kind of currency (e.g. Research "Beakers" or Production "Hammers" or Culture whatevers) could decline over time. So the decision is not to build a university or a theater, but when. The House of Wisdom in Baghdad is no longer even useful, nevermind the world's preeminent college. The Acropolis and the Globe Theater don't produce today's cultural milestones. Cities like Pittsburgh and Detroit are no longer engines of industrial power, and Venice and Tenochtitlan are no longer hubs of commercial trade.

2. Without introducing internal politics to Civ, something they've obviously steered hard away from, they could bring back the administrative costs from Civ IV that put a constraint on the size of a nation. In Civ VI, more is always better, and distance is no obstacle.

3. Cutting-edge technologies and cultural advancements should cost more, while cultures that are catching up should pay less.

4. Some technologies should be nearly "game-breaking"... until the next one... and the next one. Being the first one to 'unlock' a powerful tech should be a feat akin to building a World Wonder, but the advantage gained from it is on a timer, and if #1 and #3 above are both in effect, nobody should be getting two or three big advances in a row.

5. I know some people don't like "random disasters", but they wouldn't have to be completely random. A game like Civ presumes a certain amount of god-like knowledge and power on the part of the player, so it wouldn't be unreasonable or inconsistent for the player to have some ability to predict events that real people didn't have.

Ultimately, surviving the inevitable bad stretch of road would be part of the long-term strategy for every Civ. Everybody would have a Great Flood, or a religious crusade, or a Nationalist uprising, or a draining war, or a Great Depression, or whatever. Maybe each Civ could have an 'Achilles Heel' the same way they each have a special unit or a unique building, something characteristic of their real-world history, so when you choose to play a given Civ, you'd have some idea of which Devil is coming for you.

Anyway, I'm just thinking out loud here, but my point is that I don't believe it's an insurmountable problem.

 

[Kyriakos]

In Civ3 there are only AI bonuses (cheating) and the opposite for the human player (squares produce less than in easier settings).
Also, apparently, Civ3 AI defensive strategy is just to pile 20 units in their capital for all time, whereas their offensive strategy is to magically know which side (usually the human player) has less units in cities and then travel the entire map with a caravan of 50 attacking units so as to suddenly declare war upon entering your territory.

 

[cardgame]

Civ 4 had the absolutely amazing (imo) mod, Legends of Revolution, I believe, which could see half your empire split apart in a violent civil war. You even got the option to join them, IIRC. Not only that, but this applied to AI civs as well. It really upped the tension and dynamism of the game, if you ask me. This was its key feature but it also had a TON of other cool stuff that I forget.

 

[Chukchi Husky]

I've been playing Civilization IV for the past couple of days on Chieftain difficulty and ended up losing.

 

[hobbsyoyo]

Sword of the Stars has a rebellion AI mechanic that is a total blast to play. If you research AI, there is a chance every turn that the AI will turn on you and take a portion of your planets and ships with it. Fighting back the rebellion was almost always a joy for me. I believe they even have a scenario based around this where you start the game after the rebellion but it was less organic and fun than when it happened naturally in-game.

 

[Lexicus]

Stellaris has AI rebellions, they are fairly easy to deal with in my experience. Rome II Total War also has civil wars, they are quite easy if you make sure all your generals are loyal.

 

[hobbsyoyo]

SotS rebellions are not easy to deal with at all. They are usually massive slogs but somehow still fun.

 

[Lexicus]

Also, apparently, Civ3 AI defensive strategy is just to pile 20 units in their capital for all time, whereas their offensive strategy is to magically know which side (usually the human player) has less units in cities and then travel the entire map with a caravan of 50 attacking units so as to suddenly declare war upon entering your territory.

Yes, and in a nod to what @Timsup2nothin said a powerful strategy on higher difficulties was baiting the AI into attacking in a predictable place by leaving a city unprotected, then their stacks could be destroyed easily by your siege units.

 

[Synobun]

As a concept, I like rebellions and things that disrupt your empire.

But as an actual player, I just want to go directly from A -> B and win, and RNG-based "screw you" events are annoying to that aim.

In multiplayer it can be fun, though.

 

[TheMeInTeam]

As a concept, I like rebellions and things that disrupt your empire.

But as an actual player, I just want to go directly from A -> B and win, and RNG-based "screw you" events are annoying to that aim.

In multiplayer it can be fun, though.

Rebellions don't have to be entirely RNG, or even RNG at all. I've seen various degrees of RNG involving rebellions.

Probably the best model I've played so far is the one CK2 uses, where your subjects holding lower titles will conspire against you if they're both strong enough and dislike you, while you can block this to a high degree with force, diplomacy, and well placed marriages (creates alliances/NAP). This means that in most cases if you eat a giant rebellion you screwed up somewhere...but this is reasonably likely unless the player is very experienced.