Frogboy Frogboy

GalCiv III: Weekend update

GalCiv III: Weekend update

It was a totally awesome weekend for GalCiv AI and perf progress.

There is a lot new in GalCiv III over GalCiv II but one of the most obvious is the number of players and the galaxy size the game supports.

GalCiv II topped out with 16 players. That was a lot for back then when you’re dealing with a non-cheating, sophisticated computer AI that has to design its own ships and planets.

But GalCiv III tops out with 128 players.  That’s 8 times more players than GalCiv II and the map sizes (or more importantly, the number of planets) tops out at around 8 times more as well if you’re really wanting an insane game.

Thankfully, our CPU power has increased by a factor of 6 since then. But that power is a bit deceptive because most of that power comes from multiple cores.  In 2006, the high end machines had 2 cores which GalCiv II made the most of.  Today’s machines have more and we do our best to utilize them. 

Turn Time focus

The most interesting way to get perf improvements is to throw everything we have at it.  In this case, play a game on an insane galaxy with 100+ players.  On Beta 5 (the build you have) each turn took on my monster box 95 seconds by turn 5.  That’s unplayable IMO.

So why was it so slow? Things that are fine with 10 players quickly break down as you add more.  But even at 10 players, those inefficiencies are there.  By the time we finished this evening, we had gotten that time to 24 seconds.  That’s still really long but we’re going to have to soon make tough choices between non-cheating, smart AI and performance. 

The good news is that there’s still a lot of room for improvement between now and release.  On a more reasonable map size, the next beta update should be a pretty spectacular improvement.

AI Focus

The analytics on strategy games show that most people don’t really appreciate good AI. But we know our core customers care about it and that helps motivate me to make sure the AI is as good as I can make it in the time available.  I look at GalCiv III as the starting point as I am sure I’ll get schooled by other players.  But this weekend saw some massive improvements to the way the AI fights wars and detects threats.

The AI improvements were one of the reasons I decided to dive into the performance issue so much.  The things I’m doing are expensive and while I am pretty familiar with how to limit the scope of an AI call, it’s still expensive to do a proper threat evaluation.

Not this week but next week I’ll get started on the diplomacy AI. I’ll be asking for your feedback on possible exploits and such to look out for.

Stability

The stability on the largest maps is still really tough. It’s made a lot of progress recently but it still has a ways to go.  It’s not particularly complicated it’s a matter of optimizing and compressing data to handle those really really really large maps. 

As a practical matter, if you have less than 4GB of memory you should probably not be going beyond large.  The large map size in GalCiv III is really big. The bigger ones are gratuitous but are also being made with the knowledge that in a few years, 16GB and 32GB will be a thing. 

That said: We are focusing a great deal of effort in optimizing memory use.

96,968 views 30 replies
Reply #26 Top


- For the range tech I think the number of squares covered should proportional to the map size.    This way if a Ion engine can cover 1% of a large map, it should cover 1% of a tiny map so that the feel of the game would be the same.  Crossing from one corner to the other would take several moves.

- With the A.I as it is now, if I have a planet that is about to rebel and join another civilization and I am unable to reverse this with approval/influence modules, I find I can sell this planet for a huge weekly sum to a civilization. I usually pick the most powerful one to crash their economy.

When a planet is about to rebel to join a civilization, then that civilization should be aware they are about to get it for free.  (Not the other civilizations), so I should still be able to sell them that rebel world.

 

- On a large map, sometimes when I click on a starbase, the "to build" list will not display what I have selected right away. It takes quite a while to display.

Reply #27 Top

Quoting Seilore, reply 11

That depends, how much memory is other programs using on your computer at that same time?  Do you have virtual memory enabled?  All that will play a roll in this.

Its just Windows 7, Steam and GC3, nothing else is running that is using notable memory.

Reply #28 Top

I think I may have missed something I was really interested in, this chat about the AI, but I'll ask my questions anyway.

One of the most enjoyable aspects of the game is war itself, designing your own ships and making changes to the weapons and defensive tech placed on your ships can make the difference between dominating the enemy or a long drawn out sluggfest, I prefer to counter the AI payload with as an example point defense on my ships if the enemy is using Missle attacks, and the use of Beam weapons to counter that tech.

What I miss is your ability to actually "See" the enemy when he's in your territory, I can have my area of effect completely covered with Starbases, yet they can sit right next to you and not be seen until they move, makes me nuts at times, you should be able to set up a surveillance net and see what's going on, the AI can certainly see everything.

The big issue I have is the frustration of not being able to set an offensive or defensive mode for my ships, the AI certainly has that capability, but ships are USELESS unless I give a direct order to attack, and god forbid that you have to complete a turn....you order attack.....the enemy moves one space and your ships sit there deaf dumb and blind....there is NO PLAYER AI for your ships, trying to War at this point is an excercise in frustration, your ships should be able to see the enemy and attack or defend thier territory, adding the weight of sensors to your ships is pointless other than from a "Moves available" point, they can see an enemy fleet but have no ability to do anything with the information, I have been so frustrated at times that I have thrown away a great game that I have 40 hours in because the ability to manage a war was useless.....that just turned enjoyable to throw this away and start over, not much of an incentive to keep playing a game if only half the game works.

That's my biggest gripe at this point, I keep asking myself why I keep starting games, I personally hate Online Gaming, I gave it up years ago because of the nonsense that starts after a while, unless I'm playing against a friend I know, I just don't do it anymore, I have better things to do with my time, it usually boils down to who can rush the quickest, I can't think of anything in gaming that's less enjoyable anymore....been there and done that, pointless and not what I find fun anymore.

A game with a good AI that "Thinks" is a lot more fun than a human with one pointless objective, end the game.

I've been building my own custom gaming systems since 1992, so I'm no dummy when it comes to gaming, I'm personally good at making battle maps that use AI and have won awards from gaming companies in the past, 1st place from EA games for the Command and Conquer Generals series, Zero Hour.

Just before Stormregion the developer was bought, I got to be good friends with the producer and we swapped ideas, I would turn out maps that were made for the Code Name Panzers Series I and II games, the producer enjoyed my mission maps because of the complexity of the scripts I wrote for the AI and the detail I put into the terrain and locations in my maps, I was offered a job by them to work on the Code Name Panzers III game that was just starting development at the time, then they were sold and the new developer destroyed the game by releasing their "own" version.....a total and complete flop, they broke the magic of the original game....turned out a total loser....a mistake I have seen repeated over and over by gaming companies, they break what works.

If it wasn't for the reputation you guy's have for "Sticking with it", I would never have purchased this version of the game, so far I love it, amazing in a lot of ways...but the main reason for buying the game is BROKEN right now....in fact it's not broken...it's missing.

PLEASE give me a battle AI....I'm pleading with you.

I'm sure you'll enrich the other aspects of the game, so I'm not worried about that part, that part of the game is working well enough for now, but that's where my fun stops.

I have yet to complete a game that I have over 100 hours invested in, all because I just stop playing after a while, it stops being fun to play past the Ascension Phase of the game, and that's where it should be cranking up the fun for the time I have invested.

Please give me a reason to keep playing, Galactic Civ 1 and II are still a blast, so far Civ III is amazing, but it's turning into a "Why Play" this game for me, a time burner that brings no satisfaction for the time invested....I'm past the point of this is fun now.

Please save it for me if you can, without a Battle AI for my ships...there's no point in playing anymore.

Thanks for what will be an amazing game, I hope.

Gonzoguns

 

 

Reply #29 Top

"The analytics on strategy games show that most people don’t really appreciate good AI..."

- oh ffs!!  Most people don't REALLY appreciate anything other than 'junk that mess you up'.  Focus on what ought to be made, and do not go by what most people say.  I feel sometimes that civilization has been barely held together through all of history by a few determined people going particularly and exactly AGAINST the current generated by the people at large. [And that's what makes analytics useless to civilization.]

But basically an insightful post, mr Frogboy

Reply #30 Top

The only ai worth having in a game is absolutely hardcore and powerful one. Which will play without mercy to humans. If someone can't stand the heat then they should go and play japanese date sims. Or choose a smaller difficulty in GC3