Victory conditions, score and gameplay

I disagree on both counts.

The proper thing for a tech victory is to change "AI Tech Victory" to a LOSING condition. That is, if the AI gets to Final Frontier first, game over, player loses the game. Right now, if the AI gets there first, all that happens is that player can no longer win by tech, and must win some other way.

The AI in GalCiv cannot win any way other than by conquest, and mostly it can do that only in the early part of the game. The AI does not use starbases worth spit, except to mine resources. It won't build cultural bases, offensively or defensively. It can't use a terror star at all. And it can't "win" by tech. What's left for the AI? Total military domination. That is its only legitimate option.

If player can keep relations high enough to stay out of wars, he can have his pick of victory: conquest the long way (invasions), short way (culture bomb to flip everything) or shortest way (terror stars). Don't even have to PAY FOR the terror stars, just do them all at once and no amount of debt is too much. Player can get to a rapid cultural victory by acquiring most of the influence resources. Then it takes just a few party palaces here or there, and a few settlements on rocks in between AI planets, and game is won.

In a certain sense, the tech victory is the hardest to get. You HAVE TO beat the AI to Final Frontier, and it takes a long time, during which it can be increasingly tough to stay out of wars with AI's who swallow the rest of the map.

Until the AI is taught how to use cultural starbases at least a little bit, and how to use terror stars at least a little bit, nothing else is going to matter.

NONE OF THE VICTORY CONDITIONS ARE ANY HARDER THAN THE OTHERS. That's the bottom line. They are all the same. From a dominating position, player can do whatever he pleases. When in the hole, player just has to avoid war, and though that can be tough, it's doable in most situations, through a variety of strategies.

The idea that it's "harder" to get a conquest win looks nice on paper, but in practice, that's nonsense. Players can turn off cultural victory and play a cultural game, flipping everything, then be credited with a conquest win. I mean, come on. Everybody's got to admit that loophole renders the "conquest bonus" meaningless. Terror stars are no harder, especially since the AI doesn't react to them at all as a threat. And standard warfare is aided by the fact that half the AI population gets blown up in orbitting colony ships and transports, cutting the number of ground soldiers needed in half.


Tinkering with the influence formula would be a bad idea. It works great now, except when player gets hold of all the influence resources. Well, to fix THAT, change the AI to stop trading away its starbases. No starbase trades, ever, and then at least player has to obtain the starbases for himself, somehow. Of course, there's still the gifting of ships to an enemy, to try to get them to blow up a starbase for you. So I don't know. The end game in GalCiv is pretty much all mop up anyway, no matter how it is sliced, so perhaps it is better that it goes as quickly as possible, including leaving some contextual absurdities in the game.

I CERTAINLY don't see any point to forcing "extra party palaces" to be built. How does that make anything harder at all? It only makes it TAKE LONGER. That's not difficulty.

The AI needs to be made more threatening in the later stages of the game, if the game is to be made harder. Let it WIN if it gets to Final Frontier first. Teach it to build its own cultural bases, if it has enough military to defend itself, and teach it to use terror stars. No other substitute will do.


- Sirian

I have to agree with Sirian on this, make it possible for the AI to win and teach it to at least defend against cultural bombing and terrorstars. One way to do this might be to make it possible to attack and destroy a non-ressource starbase and terrorstars without declaring war. Then if the AI is taught to destroy starbases in neighboring sectors, you will have a much harder time to win by culture or alpha-strike terror stars. Constructors within a sector you own should also be fair game to attack without war declaration.

To make the various victory conditions interesting, the scoring should reflect how well you achieved the victory. Cleaning up the map with terrorstars to get a military victory should score no more than the current tech score. Here are a few ideas I had on how to score each victory conditions:




If the scoring was done along those lines, you would have to make tough choices each game depending on the map you get. If the AI could actually win by any of these vitory conditions, Galciv would become the best 4X strategy game so far.


~SDC~

Sirian,

I agree with you that i would be a good idea to


I also agree that the best thing would be to make the AI able to respond better to Human behavior, but my point was to try and come up with a simple solution to what i see limiting my own gameplay (score-wise)

Regarding tech victory - you say you disagree, but i don't quite under stand why you think this change would be bad. I wanted to make it harder to exploit the "free production" exploit Sirian have described in his AAR (single planet - tech victory)

Regarding Cultural Victory - as it is today you are quite right what you have to do is simply grab influence ressources, mine and colonize a planet, then the show begins.
Because of the way influence is calculated your influence universe-wide will rise as a result of each planet that flips.



Exactly my point

I wanted to achive 2 things:
1) Make it harder to "spread" your influence
2) Make it harder to culture flip planet (i agree this point might need more than just changing cultural formula - you might need to beef up the AI's cultural resist)

If the divisor was depend on the diffulty level it would also make maso players work harder for their points, also i would make it harder to culture flip on larger maps (which gives more points)

JaxomCA,



This would be problematic if you defend you SB's with capital ships, when would an attack be an act of war ?



I believe this is already taken into account score-wise, because i think tech-score depends on how far ahead of the AI you have been all the way thru the game (i don't know this for sure)



I like this idea (not commenting on the percentages), because it would be harder to achive a cultural victory with more AI's alive.


I figured there is 2 ways to make the victory conditions more equal, either get the AI to act more like a human (it would have to make more decision and take more things into consideration) or try to make it harder for the human player to exploit the AI's weaknesses.

As i said above the first is the better, but also the more complicated code-wise, if you could achive a "look-alike" be making more simple tweaks to gameplay it would be easier to get the changes made.

What i would like to hear is, if anybody thinks that the suggestions would open other exploits or destroy parts of existing gameplay.
[Message Edited]

If you protect your starbases, the AI to would have to decide to declare war on you or surrender to your overwhelming culture.



I don't think your proposal would make it harder, it would only take longer to achieve the same results. I believe the scoring changes would make it harder to achieve the highest score while providing interesting strategic choices. It would not make it harder to win, that can only be fixed by working on the AI.


~SDC~

I agree with Sirian.
I miss the point of reducing the amount of free research. I don't want to have tech resources or buildings be worth nothing, which is effectively what they'd be without a free bonus or a nerfed bonus.
That's perverting the game for no avail IMHO. To make it harder, have the ai try to compete for tech victory instead of researching excaliburs or warships. If the ai tried to win by science and could win by science, you wouldn't need any nerf.

Cultural victory: You're basically proposing that people spam space with more starbases. It's not harder, it's just longer and more boring. Such a change wouldn't make the game more fun, at least for me. It would also mean that planting colonies with a few settlers on hinge worlds/sectors would become more effective.
Also, if you need 20 constructors to flip a star system, it becomes cheaper to build a terror star instead and blow it.

JaxomCA - i think we pretty much agree. As i said the best would - no doubt - be to make the AI react better to human moves (eventhough i think it is hard to do)

LDiCesare,

My point about lowering the part of the bonus, that is free production, is based on my own playing style. I grab 2-3 tech ressources and mine them, this makes it pretty easy to out-tech the AI. I'm not suggestion lowering the free part to nothing just maybee 50% instead of 66%.

In regards to influence i have found that once you have flipped a few planets your universe wide influence will grow and therevy helping you to flip yet another and so on. This makes it possible to use energy on starting the flipping and then just sid back and wait for the rest to flip (more or less).

Its possible that this could also be solved by beffing up the AI's cultural resist.

Also the point about having to use more starbase to culture flip would mean the games takes longer and sooner or later one of the AI will have grown big and overtaken the others, which leaves you and him - at this point it is pretty likely the AI will declare war.
[Message Edited]

I hate to disagree with my fellow RB player, Jaxom, but I see a fatal flaw in all of his "100%/50%" scoring proposals, and that is this:

Basing scoring on how strongly player dominates the competition is usually MORE about facing weak competition than it is about putting in a strong performance. That is, the mechanism is way too easily "played". People will set up cardboard cutouts of opponents whenever possible, and would be discouraged from taking on stiffer challenges, because they would score less due to the fact that player would less strongly dominate.

I'm afraid I believe that whole line of thinking runs into the ditch, as far as improving the effectiveness of scoring on bona fide measurements of results.


For scoring to become more meaningful, Stardock must accurately identify what is vs what is not a challenge to the players, and find ways to measure those factors.

The current system runs into a dual problem.

Problem One: Players are rewarded for milking the score by building more trade goods and wonders, and researching at least most of the tech tree.

Problem Two: Players previously found ways to set up cardboard cutout opponents and knock out a lot of "quick one hour games" on Beginner or on Tiny maps at higher difficulty, and they were scoring more points per block of time than those playing normally.

Where is the balance between these two problems? They force the scoring in opposite directions. The first problem pushes for a solution where faster finishes are rewarded. The second pushes for devaluing faster finishes because they can be exploited in certain situations for abdurdly rapid scoring.

The only solution I can find is to better measure what is vs what is not a "challenge". Theoretically, if this could be done more accurately, both of these problems could be solved at the same time. Exactly how to do that, though, well, somebody would have to fund my efforts if I were to work on it to devise a system, and that doesn't appear to be in the works.




Weakening the effect of cultural starbases does not achieve these goals. "Takes longer" does not automatically mean "harder". It may only mean "longer, more tedious, and less fun".

My recommendations share your goals, but attempt to reach them by introducing resistance on the part of the AI. Nothing less will work. If the AI cannot respond, in any significant way, to a cultural campaign by the player, then as long as player can stay out of war (so the bases don't get targeted), NOTHING ELSE MATTERS. How long it takes DOES NOT MATTER. The game is ALREADY OVER, and it is only a matter of getting it to recognize this.

Changes have to be targetted to the root causes of problems or they will only make things worse, regardless of intentions.




I'd say you didn't quite understand what was going on in that game. A huge chunk of "free research" comes from the Technological Capital, Omega Research Center, etc -- the unique "you only get one of them" research items. Stacking these on a single planet, with ALL of your income fueling the base research on that one planet, allows a situation that cannot be recreated in a typical empire. All it means is that the one-planet-challenge is viable, because player can make his one planet a superplanet.

The amount of free research available depends chiefly upon player's ability to rack up multiple research resources.

GalCiv is dominated by the resources. Period. Military resources dominate capital ship warfare. Influence resources dominate cultural campaigns. Research resources dominate the tech race. Economic Resources WOULD be dominant, except there's a cap on effective tax income at around 4X the PQ of any given planet. Thus, more than one economic resource is rarely useful. Same with Morale resources. You can't get more than 100%, thus when you are maxed out on 100% at all your planets, there's no benefit to be had. But if there were, the Morale resources, too, would dominate.

Now if you don't like the fact that the resources rule the roost, you can lobby for that. But asking for changes that, in effect, are targetted at nerfing the cumulative and perhaps out-of-control effects of controlling most of the galactic resources -- it's not aimed well.

Perhaps if there were multiple settings to vary the number of galactic resources, we could have different spreads of games. If resources were particularly rare, for example (zero to two of each type on a huge map, say), the whole game would play differently. Player could not amass control of five or six of a certain resource and press that advantage on to rapid, easy victory.

Brad seems especially fond of stacked effects, though. The number of galactic resources is high by design, and so are the importance and domination of their cumulative effects.

That's why I say the most important change to actually achieving the goals we mostly agree on among those participating in this thread, is to change the AI to STOP TRADING AWAY THEIR STARBASES. Would also be a good idea if they were taught to defend them, because the "defensive modules" are a total joke. (Overlord killer Mark II? Give me a break!) They mean nothing after the very earliest part of the game.

Perhaps adding hit points to the base with each module added would help a bit, but maybe not. Starbases are highly vulnerable, and players can be much more effective about targetting them than the AI can, given the same resources on hand with which to work.


- Sirian

No. It's much much much more likely I'll win by alliance at this time.
Again, Sirian is right, the problem with culture flipping is that the ai doesn't try to fight it. Let it build culture starbases, then it will address the problem.

As for tech, I tend not to grab resources until very late in the game, so nerfing them would not affect me at all. Lessenning the free proportion wouldn't change much for me. In fact, it would rather hurt the ai.

Sirian - i hate to say this but....i've come to the point where i almost agree with you on all accounts



I did see why this was so significant in this type of game (one planet), but i still think it's a possible exploit.

Bottom line...i was hoping for an easy way to achive those goals, but i can see your points (i haven't given up on the idea tough - just need more time at the drawing board).

What i have realised is that tech victory is probably still hard to achive if you don't exploit as much as i do (and therefore score shouldn't be nerfed as much as it is)


That why i don't mind alliance victory score nerfed the way it is

This proposal concerns only cultural victory. It takes skills to own 90% of the galaxy without eliminating a race or 2. You would also have to consider how alien wars are going and take a more hands-on approach to maximize score.



The way I understand it, there are already some measures of how you are doing compare to the major AI. For instance, having more population than everybody else from the start will produce a lower score since you begin in a dominating position. The challenge is appropriatly measured by giving more points to higher difficulty level and bigger map size. This may not be perfect but it's a start.





I completely agree with this. The bases may be expensive to acquire but are worth way more than the sking price. You should have to get your hands dirty to grab most of the ressources on a map. With the curent system, I usually own all the non-military ressources by the time star federation is discovered, even on gigantic maps. I used to park constructors near all the ressources I discovered to grab them by opportunity but I don't do that anymore. It's much simpler and cheaper to simply buy them.





~SDC~

I don't see it.

I've played a LOT of alliance games, and my share of just about every configuration. The score consistently falls in the same small ranges, regardless of whether I dominate early or catch up late. This, assuming I play it out through most or all of the wonders, late into the tech tree. Taking an early victory is the only thing that seems to affect score, dropping it by a lot. All middle-range and late victories are pretty much interchangeable.

My OPC game on Medium scores the same as a dominate-from-the-start medium conquest. The only thing that seems to really matter is the wonders. Score will lag a bit if I get locked out of almost all the early ones, and miss a couple of later ones, too.

If there are any mechanisms in the game for boosting score if player comes from behind, they're cancelled out by the scoring mechanisms that reward player inherently for having a strong economy/research/military and building as many wonders as he can afford to pursue.

As far as I can tell, map size and difficulty level are the dominant factors. They set the range, and then everything else affects the score ONLY WITHIN THAT PREDESTINED RANGE, with small or negligible fluctuations.

I'm sticking with the notion that scoring based on the gap between player and AI is a poor mechanism, because it will be easier for most players to find ways to undercut and diminish the AI than to elevate their own performance. Thus the idea would not achieve what you are claiming.


- Sirian

I think you misunderstand me, or rather I explain myself poorly. I am more interested in measuring how well you perform for your given victory type than I am in the gap between the player and the AI. The only place this gap should matter is when you get an alliance victory, which is quite easy to obtain as soon as the alliance tech is known. Of course there should be other factors as well, like how many turns it takes to complete the game, a fast win should score better than a slower one.



As they should be, the longer real life time it takes to play the game, the better the reward should be. Habitability should probably be a factor too.

In any case, I doubt the scoring mechanism will change at this point. I'd rather see them work at improving the AI than at improving the performance evaluation.



~SDC~

All of your suggestions involved the gap between player and AI. For research, it was how much of a tech lead you have and how many techs you researched first. For culture, it's how many majors are left alive. For alliance, how you performed relative to the allies.

These are all struck the same fatal blow in terms of exploitability. Let's take the "how many majors are left alive" notion and pick it apart. Start by setting all the majors to the same alignment. Good would be safer, but any would do. Majors make nice with one another, more or less, and they all live.

See? Example after example can be concocted to show how the competition can be skewed to undermine the performance of the AI's or otherwise exploit the gap, to further the score.

Writing rules, and laws, is a challenging venture. Lawyers and legislatures don't get paid big bucks because the work is unskilled, that anybody can do it well. Designing a modern video game is tough work. It takes a lot of skill to get it right, and even then, thousands of players will play it, a significant chunk of whom dedicate themselves to finding ways to break the game so they can exploit it into oblivion.

In my view, the scoring ought to ignore the gap entirely, and focus on the challenge aspect. If the challenge can be measured even reasonably well, a functional system can emerge. GC has done a better job at this than most games, but there would still be room for improvement.

I agree that the AI should be higher priority than the scoring, though. Solid game first, performance evaluations second.


- Sirian

This is to make a tech victory while turtled up in a corner less attractive than being out there in the middle of the action. There is already a mechanism to grant more points if you are the first to research a tech, this would penalize you if you live off the AI research by playing the trading game. Some mechanism to reduce the value of each techs as time elapse would be nice too. The earlier the tech victory happens, the most impressive it is.





When the major are not fighting amonnst themselves, they are eating up the minors. When there are no minors left, they fight amongst themselves. If one major is more successful at eating up the minors, or was lucky on the terrain layout, he will become too big to handle and start eating up his neighbors. Just for the fun of it, try a a cultural victory on a large/huge map and set all the AI to good if you want. I believe you will find it is not so easy to keep all 5 majors alive.



Yeah, alliance victory is the tough one. The idea here is to reward you well if you take an alliance victory just to avoid doing the mop up. And reward you not so well if you take the alliance victory just to avoid being mopped up.



~SDC~