Hacker News new | past | comments | ask | show | jobs | submit login
4X (wikipedia.org)
124 points by tosh on Feb 3, 2019 | hide | past | favorite | 25 comments

I start Civ VI games but never finish them. It always seems more fun to start over than continue a saved game. With the latest expansion, triggering a golden age is a good goal to shoot for, as well as a good stopping point to prevent "one more turn" syndrome.

That's due to the nature of economic engine games. When you're just starting out, it's an uphill battle where every decision matters a lot. When you get past a certain point, you enter what my friend and I like to call the "mop-up phase." This is the situation where your opponents are extremely far behind, with no hope of catching up, and your decisions matter very little. Not only do you have to make a lot of mistakes to even come close to falling behind but your good decisions are inconsequential as well.

Game designers have thought long and hard about this problem over the entire history of the genre. There have been numerous attempts at putting in mechanisms to slow down a player with a large empire but they often just make the game more tedious. I think this is the wrong approach. Once you reach an overwhelming advantage the game should allow you to close it out easily.

I think of chess at a high level. If no player ever resigned, instead forcing their opponents to perform extremely long, drawn-out checkmate sequences in theoretical win positions, people would hate it. Just resign so we can move on to the next game!

In my experience it's mostly just the novelty of figuring out new starting conditions, and crafting a small system that you can know intimately without too much mental burden. We like figuring out challenges, but starting from scratch tends to be a greener pasture than building up what you already have.

It's an issue I have with all of these infinitely replayable games, and a fundamental part of human psychology, I think. Most people tend to start lots of personal side projects, but not many can truly commit to finishing something worthwhile.

I've heard this referred to as the "snowball problem": http://www.big-game-theory.com/2015/02/the-snowball-and-stea...

... because like a snowball, once you start rolling downhill, you only ever get bigger, which makes you roll faster, which makes you get bigger, etc.

4X games have had the "snowball" problem for as long as I've been playing them. You hit a point where your victory is all but assured and all you're doing is hitting the next turn button over and over again until you win. With Civ you can get to a point where you're wiping out enemy musketmen with your attack helicopters. It's not an easy game design problem to solve.

I'd be interested in playing a 4X game where research more closely resembles how R&D works in the real world. Meaning that you don't always get what you want. Success is not guaranteed. And sometimes you figure out something else that works even better.

There are games which do something similar. Alpha Centauri for example gives you four category options to choose from (corresponding to the 4 x's, if my memory is correct), and then you get one of the next techs in the pool from your category choice.

More recently Stellaris gives you three topics to research at once, but you're limited to about 3-5 choices (there are ways to increase the number of choices) in each category. Those choices are weighted to appear based on how rare / expensive the tech is and the area of interest of the scientist researching that area. There are also gameplay elements which impact the choices you're presented with, like for example if you have a gateway in a solar system you own (think stargate), then you're more likely to get an option for the tech for building stargates. It works pretty well.

The only game I can think of that was similar to the model you suggest was the long ago "Spaceward Ho!" series, which was uncomplicated but fun. There was no tech tree; putting money into each of several categories of research yielded incremental improvements in that area over time but occasionally one would get a "breakthrough" that would boost that area dramatically or provide a new capability.

I thought it was an interesting idea but it seems fundamentally incompatible with having a tech tree at all.

That also made me think of Spaceward Ho! I know it may be too simple a game for a “serious” 4x game conversation, but I still in joy its style of gameplay. Sometimes a game like Civ can be too deep, or demand too much time for a satisfying play through, where Spaceward Ho! is fun, quick, but still hits some of those 4x notes.

Are there any modern 4x “lite” type games like this?

The first Master Of Orion was like this. You put resources on research fields, and once you filled them you had a small chance (increasing every turn) of discovering the next breakthrough. You couldn't choose specific techs to research, and you couldn't discover everything on your own, meaning that you needed to trade or steal tech from the other civilizations. This was essential, because sometimes you were left with big gaps in your knowledge (say, lack of fast engines for your spaceships, or weak terraforming tech), and keeping up with everyone else was essential.

Master of Orion II was slightly better in this regard - there were 2-4 technologies in each tech discipline per rank and you could only research one of them (except for early game tech where everyone would get everything on the list). This meant you could plan your tech strategies ahead of time - barring random breakthroughs and salvage tech - so you could design your empire around the tech tree you were going to pick.

The game also completely broke this approach by including a racial attribute that guaranteed you discovery of every technology at each level - "creative". You could stack population growth and the Creative stat and pay for them with minor downgrades in other nonessential characteristics (e.g. ship dodge % or spying effectiveness %). The resulting race would have stupidly advanced ships and the economy to support them very early in the game, which usually leads to a total stomp as you just roll over the local enemies and claim a large chunk of the galaxy map as your own, turtling until the endgame.

Sword of the Stars has a more or less fixed tech tree, but your scientists can completely fail to find techs at random, sometimes locking you out of important technologies. Turns out that making a core gameplay mechanic RNG based is kind of annoying.

I liked the one in the old Stars! game where you invest in broad categories (chemistry, biology, etc...) and tech would unlock at a combination of tech levels. A better laser might be electronics 4, metallurgy 2 for example.

In a pretty distant way this reminds me of Game of Life. People spend real research effort (and brute force computing power) to discover structures with certain capabilities (e.g. glider guns). I think the key is that the new tech to research is not even known by the game creators in advance. Something similar can be seen in machine building games like Besieged.

These grand strategy games are always my theoretical favorite genre, but I just don't have the patience and/or time to commit to them (beyond the relatively simple Civ entries). I am equally envious and dismayed by friends who tell me how many hours they've logged not to just the game in total, but a single play in Europa Universalis and its ilk.

I always preferred Colonization to the Civ series, and keep wondering why I am apparently in the minority.

Civ, and also SimCity, always felt as if they would progress along more or less the same trajectory of growth, with little actual choice for the player as long as they broadly stayed within the bounds of what the designers considered typical gameplay.

In Col, I loved building these intricate industrial networks and trade routes.

Now that I’m writing this, I’m wondering if Colonization possibly came too close to the Soviet idea of central economic planning for the taste of the audience.

Same. Though the AI since hexagonal tiles is so bad I've signed off Civilization series. Why they haven't contracted to that Google subsidiary to use machine learning to improve the AI of the game is a mystery.

If you play Civ V on Windows, you can get the Vox Populi mod for vastly improved AI.

I love Civ V (and loved Civ 2 back then), somehow I never played Col. Just installed Civ IV Colonization and the weekly build of FreeCol :)

You can get the original Colonization DRM-free from GOG for $6: https://www.gog.com/game/sid_meiers_colonization

It's worth it -- the Civ IV DLC version didn't really capture the magic of the original, and FreeCol has been a work-in-progress for a long time. The '90s graphics of the original can take a little getting used to, but the gameplay is still extremely solid. I keep a copy on my laptop to play when I'm stuck waiting around in airports.

My favorite game to restart constantly, second only to Civ itself.

I want to see a AlphaGo/AlphaStar version playing Civ VI.

I doubt that will happen, considering that civ is much more complicated (in terms of mechanics) than playing go or even starcraft.

Civ isn’t really that complicated compared to sc — “real-time” play is a big burden, and dramatically increases the search space.

I think the bigger blocker is that civ has no real tournaments to fuel AI popularity, that the game (and 4X in general) is likely buggier due to less online-play (meaning the AI is likelier to get trapped exploiting uninteresting bugs) and at least newer games are probably somewhat unoptimized and will have longer simulation times (thus slower training) (both valve and blizzard are quite good about making games playable on a toaster; Firaxis is.... not.)

Sure, Civ has a lot of buttons to press and things to choose but is it much more complicated than StarCraft II? Never played SC2 but it sure seems there's a lot of buttons to press and things to choose.

Can't wait to play AoW Planetfall!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact