Brad Wardell's views about technology, politics, religion, world affairs, and all sorts of politically incorrect topics.
Hard core vs. mainstream
Published on July 16, 2006 By Draginol In GalCiv Journals

Writing strategy games is a lot like playing strategy games. Lots of choices on where to put scarce resources.

As the primary developer of the computer AI for Galactic Civilizations II, I have to make a lot of tough decisions on who to focus effort on.

And by "who" I really mean the general population (95% of gamers) or the hard core 5% who are the most vocal but small as a percentage of buyers.

During development, we focused on mainstream gamers mostly.  We wanted to make a game that strategy gamers would enjoy and to do that, we had to be careful where we spent a lot of time on. 

For example, we put a lot of time on having really cool ship design. Players can literally create any kind of ship they want both functionally and visually. There's Babylon5 ships out there, Star destroyers, etc.  But at the same time, we didn't put in multiplayer because the cost of doing it would have meant the game would have to list for $50 instead of $40 and that was something we wanted to avoid.

Getting back to computer AI -- how tough the computer players are -- one of the game's claims to fames is that the AI provides a challenge to players without having to resort to cheating. The computer players play the same game as the human players do.  In most strategy games, the computer players have to be given very significant resource advantages to provide a challenge.  In many strategy games, the computer players aren't even playing the same game.  One popular RTS I've been playing has its computer opponents able to essentially conjure buildings and units out of thin air in order to provide a real challenge.

But a challenging AI isn't the same as an unbeatable AI.  So the result is that some of the hard core players -- as well as top level strategy gamers -- are able to defeat the game not at the highest levels but at the levels where the AI isn't cheating.  As a result, they demand ever increasing amounts of time to be spent on making the computer players smarter as opposed to just bumping up the difficulty higher to where the computer players do get some extra resources.

Those players are able to list a myriad of "dumb moves" that the computer players made during a game in question.  The problem is that all computer players will make "dumb" moves.  Sometimes dumb moves are the result in a logic flaw, other times it's merely a difference in opinion on what is a "good" move, but most of the time it has to do with large-scale, multi-turn coordination which humans are much better at doing than computer AIs.

Right now I'm working on v1.3 of Galactic Civilizations II.  In terms of computer AI work, I'm trying to provide something to both groups. I want casual users to be able to have more tools to beat the AI with if they are clever but want hard core players to see smarter AI.  I'm also going to put in some additional difficulty levels so that players have more graduated difficulties (the game has several already but more is good IMO).

The issue the developers of GalCiv II are running into is no different than other games.  Players of World of Warcraft have long seen Blizzard have to juggle time between the hard core players with their level 60 characters and massive guilds vs. casual gamers who just want to see more missions and more solo-oriented content.

So that's the conumdrum that developers regularly run into when deciding where to put updates.


Comments (Page 2)
2 Pages1 2 
on Jul 18, 2006
I think the AI is good as it is, and I only play on challenging, but that's just me.
I would definately like to see more diplomacy in general as well as espionage and breaking alliances with minor races.
Thats not to say I don't want better AI...
on Jul 18, 2006
That's a very interesting journal. It's good to be brought to see another point of view. Now obviously I have expressed my opinion on AI, so you don't need me to do it again. I'll just confine myself to saying that casual players won't be turned into fans unless they really care about playing the game, and good (and continually improving) AI is surely a major factor in that, at least as much as "candy" is.

Don't forget that "5%" (and I bet it's more than 5%) are also the ones that pre-order your game at full price, always get the expansion pack, persaude other people to buy the game, and will likely pre-order the next game too.

We're the best 5% you got. We're hardcore.


(Even if we're a pain in the ass too...)
on Jul 20, 2006
I will start by saying that I personally think you folks at SD have done a fantastic job with GC2. From release to v1.2 the game has gotten better and better each step of the way. No small feat in this industry. And for a small company it is even more impressive.

As for the AI I think at times what people see as *flaws* are actually intentional. Take the planet developement for instance. It has been stated in several posts that it is meant to be *good* not *perfect*. Also think about this: if your core planets are generating enough MP and RP why would you ever want to fully develope what are in effect frontier worlds that may be taken over by a hostile force? Basically you'd just be handing them completely upgraded and running planets. Sounds rather silly to me to do that don't you think?

Also I'd really like to know what map settings people use when playing on those levels above Tough. A large or huge galaxy with all AIs enabled on Tough gives me a very challeging game to say the least. Above Tough I've got to either reduce the number of players or reduce the size of the map to have a snowballs chance in hell of coming out of the first 100 turns with an empire let alone be in a position to win. Now maybe that's my style of play since I don't usually start out at turn 1 conquering the universe. But even so I wonder at times if people truely give themselves a challenge or just beat one AI into submission and then point out all the things it did wrong.

So I know its only one vote in thousands but I'd really like to see just more game, more ships, more AIs, more buildings, more planets. Stuff that it sounds like is coming in the expansion which I'm definitely looking forward to.
on Jul 20, 2006
Thanks for the opprotunity to forward some feedback.

To crank up the difficulty level of the Hard core types:

1)Use the exsiting fire power of all the races as a potential counter threat too a Human player that is steadily increasing his ability to dominate the universe. A option for any dificulty level is to Force the AIs to become increasingly intollerant of a human's ability to dominate. With AI colitions suddenly springing up against the human player, as well as increasing resistance to trade, negociaite, maintain alliances, and or levels of diplomatic confidence, IE dropping relations....

Where it would cost the human player more in $$ to function in an increasingly hostile universe.

More coordinated AI ambushed/attacks against a dominate humans. IE attacks agaisnt both key Tech, manufacturing planets and key star bases by multiple AIs.

More ship/techology trading to help AI local too human, or in serious need.

Use the Existing economies of the AIs to match human output in a timely manner....


2) Tune up AI use of Star bases. A huge increase in lethality of an AI is created if it simply created one maxed out econ SBase for each appropriate planet cluster. I do this as a standard move. Ive only see the AI do this once in many many games. and then the SBases where mostly un protected....

Star base use is critical to succes....

Also tune up the AIs to sneak attack human SBases with lethal force.... not in dribs and drabs... AI must take out key human Sbases while in a war.

3) Get AIs to stop with the bizzar planet building strategies of a class 20 AI planets with 12 farms or anti influence structures.... AIs need a few key maxed out manufacturing, tech planets.... as well make sure the AI uses the bonus tiles to effect by building appropraite strucutres their without exception.

4) Get AIs to quickly crank out/modify ships to adapt to human fire power weapon builds.... we can do this within 15 turns why cant a drengan AI with a huge economy not do this as well??

5) Get AIs to concentrate overwhleming force..... placing ships to protect planets that can be picked off one by one is a stupid strategy.... a weak troop ship stopper per planet is fine.... but AIs should be setting up huge well placed fleets...

6) Get AIs to target Human troop fleets at all costs......

More suggestions later

Yukon Jack
on Jul 22, 2006
Are these multi-turn-strategies for computer players really that difficult to implement? Isn't it possible to implement some kind of planning-tree for future turns, bind to that tree some objects that contain the goal (conquer planet xy, build starbase at...) and the resource-objects that should fullfill the plan (military, scout, settler, constructor, "SpyObjects" (test if planet population + enemy troop-tech is bigger/better then my invasion fleet, test if enemy could build more ships faster then I can, test if other ships are in range, test if he has more advanced ships/techs etc.) bind to that tree also some actions (move x to y, "spy population", "spy tech") and then parse through that tree every turn and see, if the plan goes at it should or change plan or resources - or - most important - cancel the plan, if it becomes impossible to reach the goal.

I paused playing the game at v1.1, because i couldn't see the AI doing/checking things like these (and in the upgrade-listing to 1.2 I found nothing, that this had changed). It was very boring to be permanently in war and kill tiny ships all the time, because AI didn't see that I could build the mightiest army in universe in 3-4 turns - because it only sees that at the moment, my military fleet is nearly at zero - and don't calculate that at the time it's fleet arrives, I would be well defended, or at least would have the power to take everything and much more back. So the AI started to send every 1 or 2 turns a bunch of troop transports, get them destroyed, send another bunch of troop transports and get them destroyed... attack a planet with a population of thousands with some hundred men.. and becomes weaker and weaker until I conquered it. So there was no strategic AI planing at all involved (to fight if I can't win is stupid), or at least it looked like this. And because I played on a slow machine, I didn't like all that boring fighting for nothing, that never stopped.

So my suggestion is, the AI should only act if it is sure it can reach a goal, try to check everything before it does anything, don't spend resources senseless on weak military units that got killed every turn, aboard stupid attacks and wars, and instead fully concentrate on reaseach/social until it is a power that really could win a war. AI should include checks for enemy potential and not simply go after the "right now" state, so a lot of stupid attacks could be avoided and gameplay would be much smoother, because a war would only be fought if it is a really challenge and not a boring counter-attack (I never delared war, but everytime the AI declared it, AI was conquered in no time).
And it should build only ships on purpose and not a huge fleet of ships without any power that fly through space without any benefit (but slow down my machine).

As for Hardcore vs. Casual Player, difficulty level could then adjust how much checks the AI does or if the AI sometimes attack even if it is sure, that it will lose - so a casual player could get that feeling of power at successfully defending against an attack. (...that he could had always defended against, but well, because he is a casual gamer he wouldn't had known this). At higher difficulty settings, AI would do more and more checks, so becoming more and more successfull on every AI-actions because they are more likely to be a success and the gamer has more and more to think for countermeasures. This would also give much more satisfaction to the hardcore-player if he knew that the AI will only do things with a high success rate, so defending against it, is an archivment (It's no archivment to conquer a weak enemy or to kill a thousand tiny/lowtech ships).

On the other side, different strategy styles enrich the playing experience on every difficult level - like hit and run (fight for a single planet and then try a peace negotiation), blockade (send a fleet to every planet and there, destroy every fleet and ship until own population growth allows invasion), tech&social power, gradually weaken (attack only to weaken not to win), find the single most weak spot (planet that is far away from every enemy production center, but near to own territory), grow bigger, diplomatic isolation/alliances. So, for every strategy, the player had to find its own countermeasures and difficulty only sets how well planned that choosen AI-strategy aproach is and how likely it will succeed.
2 Pages1 2