:LOL:@Sirian!
End of quote
Dismissing me with a laugh, eh? We'll see who laughs last. You won't be able to dominate with this kind of trick. If your view is to win out, it must do so on the merits.
Of course, playing at less then Maso for a player like you would probably be offensive to you... just like not using any of the cheeses.
End of quote
You think you have me pegged, but you're not even coming close.
The majority of humans I've played in various strategy games cannot beat me. So I do not expect a bit of software code to do better then human brains. Why do you?
End of quote
Assumptions inherent to your question are not true.
I do not expect a bit of software code to do better then human brains. Why do you?
End of quote
I don't.
The golf handicap system offers a useful model. Two players of different ability can be matched. In theory, the handicaps applied to each player as a result of their recent performances will gauge their relative ability levels. When each player's handicap is subtracted from their number of strokes, the results become comparable.
Players end up playing against themselves -- their own performance history -- rather than against the other player. No direct comparison will be meaningful, especially if there is wide variance in the skill levels. Yet with the handicap in force, each player is playing against his past range of performance. If the weaker player has a good game, better than his average result, the strong player must do even better relative to his average performance to prevail.
The odds are evened. Each player has a fair chance to win a given round, and in fact, the outcome will not only remain in doubt, but rely on each player's performance. The one who has the better game relative to his ability will win.
That's primarily what I want from a single player computer game: an uncertain outcome, where the result turns on the strength or weakness of my performance. Provide me a handicap dial that I can turn up or down to choose the strength of my opponents, to choose the amount of handicap they are given to prop them up, then give me opponents who are sophisticated enough to provide variance to the strategic challenges.
Predestined results are no fun. Can't-win situations are boring. So are can't-lose situations. So are opponents who are completely incapable of surprising me, who are completely predictable in what they will do.
The handicap system is as essential to GalCiv as it is to Civ3 or MOO. Difficulty levels, map settings, opponents, mods, and more. Problem is, the GalCiv AI only has one strategy, one behavior, one threat. That's war. War is all it even pretends to try to do. Not so for Civ3 or MOO1.
Civ3 has a tech victory: the spaceship. The AI actively pursues a launch victory. It prioritizes launch-related techs. If it is too far behind in the launch race, it will try to use warfare against you. Staying out of wars in Civ3 is possible, in many instances. (Not all, but that too is good.) However, staying out of wars does not guarantee victory. The player must out-tech the AI, one way or another, or else deny it the strategic resources it needs to build all the ship parts. Granted, most games of Civ3 have been determined well before that point, but the AI is CAPABLE of posing a threat (several kinds of threats, actually) in the late game. The GalCiv AI is not.
If you can maintain good relations in GalCiv, peace and stability are certain. If relations grow poor, warfare is certain, unless the AI has bigger fish to fry in the form of an even worse enemy. Either way, once you learn the mechanics of the diplomacy, the game loses its meaning. If you bribe an AI, you can be certain to remain on good terms. Remaining on good terms leaves you certain to win the game. If you refrain from bribing an AI, you are certain to end up at war if your alignments are mismatched, unless you are receiving all of that civ's trade routes because you are their only friend. That may happen naturally, in a few instances, but you can be certain to create the situation if you bribe wars to start. Enough bribing can see all the AI's at war with all the other AI's, and all of them having no choice but to send their trade to you.
Contrary to your repeated assertions, your blind and presumptive assumptions, I am the exact opposite of the loophole seeker, the exploit abuser. I eagerly embrace the handicap concept and will choose to add whatever it takes to the AI's handicap to prop it up.
This approach is the defining quality that sets Realms Beyond apart from the other Metaverse Empires. You've been gone from the GalCiv scene for a while. I didn't even show up until after you had faded away. You're no newbie to the game, but you are unattached to an empire and perhaps unfamiliar with what has transpired during your absence.
I've never bribed the GalCiv AI's into a knot, so that they are all at war with one another. I know it can be done, and that if done, the game loses its meaning. Thus, I have had the restraint not to do that. Yet if bribing to war is taken completely off the table, the game also loses an important feature. If we can't ever use the diplomacy tools because they wreck the game balance, then we don't even HAVE a functional diplomacy rule set, do we? Well, that's the problem. GalCiv does not have a functional diplomacy. Peace is certain with good relations. War is certain with poor relations, unless the player is militarily dominant (in which case it's all moot anyway). Paying on credit gives the player a virtually unlimited bankroll to fund initiatives. Take away the ability to use credit and certain options (like being able to purchase trade goods, military assets, or some techs) come off the table, too. Too expensive. Lose the ability to use bribes, both to improve relations and to pay AI's to attack one another (to start wars and distract a strong opponent), and there's nothing left with which to work. Rivals of mismatched alignment will be destined to end up at war with you. So either way, the outcome is predestined. Use bribes and victory is certain. Don't use them, and war is certain, turning this into a war game.
Right up to the point at which you master the diplomacy in this game, the game is very enjoyable. The moment your diplomatic skills reach the point of "can't lose", and you also realize that there's no use in voluntarily restricting your use of diplomacy because if you do, you end up at "can't-avoid-war", then you're done. The jig is up, the fun is over.
For a handicap to be able to work, there has to be a way to create uncertainty. GalCiv doesn't have it. On the one hand, if you make use of the diplomacy, you're certain to be safe, and certain to win. On the other, if you don't use it, you're certain to end up in war with particular opponents.
I'm highly experienced at designing handicaps, at patching game balance to cover loopholes, to remove or paper over exploits, to craft a playable game. The Realms Beyond has hosted a tourney for Civ3 that has played forty events and is still going strong. There is a list of rules, of tactics and moves that have been restricted or taken off the table, to prop up Civ3 and give it a robust environment for playing meaningful games. The same rules have been used as the basis for dozens upon dozens of succession games. The rules list is as long as it needs to be to get the job done, but no longer. Some items resist rule writing and are handled on the honor system. Each event is tailored to provide a different strategic situation, some special rules or elements that will try to throw even our most experienced players for a curve or two, to let them venture into uncharted regions of gameplay. If you want to learn more, click this link:
http://realmsbeyond.net/civ/epics.html
The sad fact is, I've concluded I CAN'T do this for GalCiv, in its current state. The AI is too weak. The unbalanced elements in the gameplay resist good rule writing. With more work, more revisions, GalCiv might yet get there. We didn't assemble our tourney for Civ3 until eight months after that game's release, and it took two more patches after that for Civ3 to reach a truly stable balance. It too needed a lot of work to plug leaks, prop up the AI, and smoothe imbalances. No game this complex can hope to work perfectly out of the box. A long interactive patching process is necessary, with thousands of players putting the game through the rigors and giving feedback. GalCiv has enjoyed that kind of support and Stardock is still working on it.
Despite your dismissals, MOO1 got more things right than any other 4X game to date. There's still plenty of good lessons Stardock could draw from it.
Man, you are dreaming about MoO1. ... they all declared war as soon as they ran out of worlds to colonize in range.
End of quote
And the problem with that is... what? You're criticizing the AI for trying to win the game. That's ridiculous.
MOO1 had two victory conditions, and both of them required a civ to be one of the largest two civs in the galaxy to qualify for victory. Thus, if a civ not among the top two were to rest on its laurels and not seek to gain additional territory, it would in effect be giving up and settling for a loss. That would not be competitive behavior, nor would it make for much of a game. Larger civs had to fend off smaller civs to win, and in fact, the more territory they could obtain and control, the better.
The nuances of the game lie not in the choice of war and peace, but of who gets targetted and why. Unless somebody showed a grave weakness, the pacifists would tend to wait for more aggressive civs to target them, and then fight back. The pacifist is more like a passive-aggressive. His gameplan is to seek to avoid squandering resources in warfare, let the other civs duke it out, weakening one another, and build up his own technological and production advantage. Even the pacifist would go Imperial and start conquering the galaxy if they gained a strong enough position. Thus, each personality had a valid strategy, a different emphasis, but all were aimed at the same goal. That is far FAR greater intricacy than you are crediting to it. The MOO1 AI would set objectives: territory to try to claim. It would try to choose weak points to attack, to bring enough force to get the job done, and if beaten back, it would gather a larger force and try again. It would even adapt its ship designs, if it survived long enough to make adaptations. And diplomacy would affect its choice of objectives. Then there was the other half of the personality, determining which areas of tech it focused on as priorities. Having different tech balances among the opponents posed different challenges for the player. The principles would remain the same from game to game, but the details would differ, requiring the player to adapt, to think, to look for weaknesses that would shift and change.
The diplomacy mattered. You dismiss the AI as random, but it was much more nuanced than you describe. It is a FAR cry from the "predestined wars" of the early Sid games. Honorable civs would stay on good terms with their friends. They would develop what amounted to hatred for their enemies, though, fastening on a particular "dishonorable" opponent and sticking to them like white on snow. Effective diplomatic play could see you allying with such a civ and developing a reliable friend. Erratics tended to be too random, and thus ineffective at furthering their own cause, but also dangerous, in their degree of unpredictability. All of the personality types were coded to go for the win if they reached a dominant position, if they obtained a decisive advantage. That's what made the game worth playing! Given enough handicap to perform competitively with the player, the AI was capable of winning. There were enough uncertainties and variables to lend meaning to the diplomacy. Players might be able to play civs off one another, with the right moves, but this wasn't always easy and never could be guaranteed. Player had to stay within a certain range of the lead AI, or it could all come crashing down on him. (GalCiv doesn't have this element, to its detriment -- although it does have a solid ruthless streak that is absent from Civ3, to its credit).
Of course the MOO1 AI would play aggressively. You could make friends with certain civs, though. Some took more work than others. Friendships might last all game long, if you opted to play for diplomatic victory. If you opted for conquest, friendships and alliances were always temporary. That the AI might decide its use for you had come to an end before you got around to deciding the same about it... more power to them!
That GalCiv allows for alliance victory is partly responsible for its AI sometimes reaching a place from which it will not declare war on you. That's appropriate to the alliance victory condition. BUT... since the AI is not capable of pursuing the cultural victory, the game is over once the military threats are neutralized. There is no way to "lose by alliance". That's the same as a military defeat. MOO1 did better in that department, hands down. You can lose the game by diplomacy without being wiped out. The absence of this possibility from GalCiv puts the lie to its victory conditions. The GalCiv AI can win only by military means. To lose the game by conquest or by alliance is the same: the player must be eliminated. To lose by culture would be viable, except the AI only poses a threat via culture when it controls all the influence resources -- and it is pathetically easy to take those away. You can buy them. And if you couldn't buy them, if that were taken off the table, starbases are easy enough to destroy. Oh, and the fact that a civ has to be at peace with everybody to win by culture... all you have to do to stop a cultural AI win is to declare war, so even THAT one forces the AI to have to eliminate you to beat you. No matter how you slice it, you can only lose in galciv by being wiped out militarily. This is not enough. The game ought to be able to do better.
If SD spent 10 years on improving the AI so it knew all the best/correct values to plop in a proper Game Theory grid and choose the correct responces and paths to beat 80% of all humans when set to "Hard", there'd still be people able to beat it at it's best/most cheating levels of artifically inflated "toughness".
End of quote
Faulty argument. Again, you ignore the principles of the golf handicap concept. There is no limit to how much handicap can be added to the game. And all that is necessary is that the handicap be sufficient to match the AI to a given player. Let the player set the dial, choose his own challenge level. If the settings for the highest difficulty level are not enough bonuses to prop up the AI, allow the handicap to be dialed up even higher. Every player can be matched with enough AI bonuses to make the game's outcome uncertain, IF the AI is sophisticated enough, variable enough, unpredictable enough to pose a legitimate threat.
If the AI is strong enough, it can be propped up with handicaps. But for that to work, it must not have fatal weaknesses for which there is no workaround. Civ3's AI has fatal weaknesses, but they can be compensated for by taking certain tactics and exploits off the table. If we take the fatal moves off the table for the GalCiv AI, there isn't enough of a game left to be worth replaying. Stardock might yet fix that, but that's where it stands for me at the moment. Unless the GalCiv AI can be given another way to defeat me besides total military elimination, I've gotten all the fun I can from this game and I'm done with it. That is painful to say, but I don't see where Stardock is served by me being unwilling to say it. Brad and Cari need honest, intelligent feedback if there is to be any chance for the issues to be redressed.
- Sirian
Stardock Forums
