21-05-2002 03:16									#31
	Support Apolyton, buy Galactic Civilizations: Deluxe Edition

quote: Originally posted by SpencerH To some extent, I've become an opponent of the great wonders. IMO they've been totally screwed up in CIV3. At the lower levels where you dont need them, you can get em all. At the higher levels they counter the AI advantages (which is how they were envisioned in civ) but the AI tends to get them 1-2 turns before me so I dont always bother.

The problem with Civ3 style wonders, as I see them was that: Wonders were either random (due to great leaders), or Wonders were handed out according to who got the tachs first (because you can't rush-buy them).

I liked the concept of small wonders that actually required things to be available... but I disliked that they were available to all.

I think that all wonders should depend on a variety of things:

Oracle Mount (a super-holy temple thingy on a mountain)

Effects: Can't think of anything, so let's assume it's like the Civ Oracle and doubles the effectiveness of all temples.

Requires:
- A mountain tile somewhere in the province building it.
- At least X influence in any one diety.
- A certain construction tech.
- Either the expenditure of X mana of the diety with which you have the most influence, OR the presence/sacrafice of a clergy leader, etc.
- A temple in the province.

This way, you would actually have to use things to build it besides just production points, and it would be fun IMO to fulfill the requirements to finally make the Oracle Mount. Since buildign the wonder (waiting X turns) is not very fun, we can at least make the seeking of the wonder and the completion of teh wonder entertaining.

	21-05-2002 04:07									#32
	Remove this text

Excellent, Cyclotron! Quite excellent! And in fact, as the ideas for the Commercial release are beginning to gel and mature here, and with the rise of prominant individual figures (what other games in the genre might refer to as "heroes" what these minor wonders will amount to are Candle'Bre's version of "Quests" that the other games' heroes might go on. In fact, we could incorporate that into the game as well. I could see, for example, one requirement of a minor wonder being the kidnapping (new espionage ability that's based off of assassination %) of some rival faction's (insert notable leader here...Priest, Justicar, War Captain, General, etc).

Essentially then, whilst the rest of the game is swirling around the players, there are a number of these "quests" that can be embarked upon (depending on the game situation, etc...as you say, not all will always be available) that may provide a badly needed edge to boost your faction over the top! Bravo!

-=Vel=-

	21-05-2002 05:10									#33
	Help yourself to an AD-FREE life

I like the idea of having wonders dependent on events, but I don't want to repeat the Civ3 mistake of necessary events... Your first leader needed to always be an army, so you could get more leaders (heroic epic) and train more armies (academy). In other words, it was crucial to get a leader and use him for that purpose... even if you didn't need an army.

Here are some ideas:

Rallying Memorial: Requires the death of one Military Leader in combat; maybe kicks up the defense/morale of all units in the province it is built.

Rockface Citadel: Requires the building of at least 5 walls in your provinces, and their upgrade into keeps/castles/high walls/whatever. Additionally, the province which you build it in must contain mountains or hills, and you must have won a major (x units) battle against an opponent in the province. Instead of a regular wonder, the Rockface Citadel is a super-fortress tile improvement, something like triple/quadruple defense and vision of the whole province even if you lose control of the province.

Moon Obelisk: Requires at least 4 field tiles in the province it is built and x Ollux influence. Your astrologers can calculate the position of the moon with this pseudo-magical relic, and all food production in fields in this province is increased by 50%.

Adaaunx Sigil: Requires at least 4 different greater sigils in any of your provinces, and the sacrafice of one of your "Mage" leaders. A special sigil (see magic thread) that acts as a focal point for all High Magic spells, albeit at 25% higher mana cost.

Diabolical Interrogation: Not a building, requires some magical requirement as well as the capture of an enemy leader. All enemy terrain is visible to you for one turn, as well as all enemy units there. The wonder has various compromising effects depending on the leader you capture: A military leader might mean the weakening of that Faction's troops. This may be built multiple times, one for each leader, but lowers honor each time.

Jus' some ideas...

	21-05-2002 17:13									#34
	Browse Apolyton AD-FREE

What is the main reason that the computer players follow a simplier set of rules? Too difficult to program the real thing?

	21-05-2002 17:25									#35
	Support Apolyton, buy Civilization 2

Telling the computer players the rules of the game is easy. Telling the computer how to beat a good human player using those rules is almost impossible without either a) being a programming genius and making the game have very high specs to cope with a complicated AI routine, or b) get the computer to cheat (eg. free money, free units).

Making the computer have a simpler set of rules will make it easier (although not easy) to program a competitive AI that doesn't obviously cheat.

	21-05-2002 17:31									#36
	Got spare money?

FP is right on the money here. Each simple little addition to the rule set complicates AI evaluation a lot. Really.

The other popular road to competetive AI (not that they're mutually exclusive, mind) is to streamline the game around it from scratch. CB grew out of a board game, so they don't really have this option - they'll have to pursue the other one relentlessly.

	21-05-2002 17:47									#37
	Support Apolyton, buy Call to Power 2

quote: Originally posted by cyclotron7 Here are some ideas:

Very nice ideas there cyclo. The Rockface Citadel in particular sounds really cool.

I really like the notion of having the wonders dependent on a series of events and conditions, like quests ... I think that will help create a more realistic and vibrant world. It will add more breadth to the game IMHO.

	21-05-2002 20:44									#38
	Got spare money?

Two thumbs up on the whole suite of ideas there, and yeah...that RockFace Citidel is coolio....

-=Vel=-

	21-05-2002 21:40									#39
	Put an end to popups!

Most of my ideas are adapted for MP... but it seems like you could program the AI to somehow fulfill each requirement, as if it was a shopping list. For example, once the tech for the Rockface Citadel becomes available, the AI would make sure it had at least 5 upgraded walls, and then move a large army just behind one of its vulnerable border provinces. The other player might attack the undefended border province, and the AI would counterattack with the large troop force as to win a large battle an recapture the province... of course, it would only do this if the province had hills or mountains and was strategically important.

That's a tall order, though, so mainly the idea of multiple conditions is for MP.

	22-05-2002 18:50									#40
	Support Apolyton, buy Civilization III: Complete

Yeah, that's a good point...not sure how we can convince the AI to regularly and reliably go after the building of small wonders. OTOH, if we take care to make sure that they're more interesting/cool than unbalancing, perhaps we won't need to worry over it.

One thing I definately wanna do is come up with a longish list (30-50 possibies), which should give us a lot of variance. That's one thing I really hate about most games. They don't have enough stuff in them....you get to see pretty much everything there is to see in 1-2 games.

Not on my watch tho....

-=Vel=-

	22-05-2002 19:45									#41
	Get a bigger avatar today!

quote: Originally posted by Velociryx One thing I definately wanna do is come up with a longish list (30-50 possibies), which should give us a lot of variance. That's one thing I really hate about most games. They don't have enough stuff in them....you get to see pretty much everything there is to see in 1-2 games.

The main reason behind this is probably game balance, and that is something that we must be aware of. Making 30-50 variations that are evenly balanced and not game-breaking is going to be an almost impossible task. It seems inevitable that of the, let's say 40, small wonders in the game perhaps half of them will be a pretty low priority for the human player, maybe to the extent that they don't bother to fulfil the requirements to get them.

I dunno though, I suppose the complex requirements for the small wonders, while it kinda cripples them for the AI, might improve this state of affairs for the player. If we make them hard to get then we can afford to make them all reasonable powerful ... the player should be pretty sure that the benefits will outweigh the costs of course ... and we should try and aim it so that no faction gets more than a couple of small wonders per turn. I think that can be done, I just want to be sure that we don't end up with 10-12 'good' ones and 15-20 fillers, I suppose that's up to us to decide since we're gonna make them up I suppose. Balancing the requirements so that they're challenging, but not too much, will be tricky, but we'll get there in the end.

	22-05-2002 19:49									#42
	Increase Your PM Length

Also ... we should allow multiple factions to build a single small wonder, and even a single faction to build multiple copies of a small wonder (if it fits to allow that ... it would for Rockface Citadel, maybe not for Diabolical Interrogation).

Did that paragraph make any sense? It sounds right in my head, but looks weird on the screen ... too many 'singles' and 'multiples' I think ... but I hope you get what I mean.

	22-05-2002 22:53									#43
	Enter the AD-FREE zone

The quest ideas sound good. I've been thinking about them since someone mentioned warlords III on another thread. If you havent played it, it was fun because of the quests and getting the goodies with success. The problem was that the quests became the raison d'etre of the game in order to win rather than a fun side game.

	23-05-2002 02:56									#44
	Support Apolyton

quote: Originally posted by FrustratedPoet Also ... we should allow multiple factions to build a single small wonder, and even a single faction to build multiple copies of a small wonder (if it fits to allow that ... it would for Rockface Citadel, maybe not for Diabolical Interrogation). Did that paragraph make any sense? It sounds right in my head, but looks weird on the screen ... too many 'singles' and 'multiples' I think ... but I hope you get what I mean.

Okay, I'll try my best:

Most of the wonders above (Memorial, Citadel, Sigil) can be built only once by any faction (like Civ3 small wonders). They can be built only once, so if you lose them they cannot be restored. The Sigil and Memorial become inert once you capture them from another faction, but they are not destroyed unless you tear them down, which causes an honor loss.

The Moon Obelisk is specific to a diety, not to a faction. There will be one or more of these small wonders for each diety... they can only be built once each. The only restrictions on these are that they require X influence in that diety to build, and you must maintain X influence or else they become inert until you have that X influence again. These do not become inert when you capture them, provided you have the requisite influence. You can destroy them, but that causes you to lose ALL influence with that particular diety immediatly.

The Interrogation is in a class by itself. Rather than a wonder, it is really more of a controllable event. So long as you have the magic requirements/ other reqs, this comes up as an option to build whenever you have captured an enemy leader. You can ransom the leader, or sacrafice him to build the Interrogation. The honor loss happens as soon as the leader dies, so you can lose honor without finishing the Interrogation if you interrupt the building. Basically, the Diabolical Interrogation crushes them mentally and emotionally... so they are effectively useless, and are removed (as in the Ollux spell, Lunacy).

EDIT: Since the Interrogation is not a permanent building, it cannot be captured and does not become inert, but if progress is interrupted in the province building it (by invasion, among other things) the interrogation is lost and you may no longer build it for this leader. The project costs more shields for better/more experienced leaders (harder to crack) but you also get more info from better leaders.

	23-05-2002 03:20									#45
	Support Apolyton buy from Amazon

OK, I think I get it now ... thanks Cyclo.

	09-06-2002 05:32									#46
	Support Apolyton

Wonder Idea list:

UNIVERSAL WONDERS (One per game only):

Oracle Mount: Increases the effects of all religious buildings by 50%
Reqs:
- A mountain tile somewhere in the province building it.
- At least X influence in any one diety.
- A certain construction tech.
- Either the expenditure of X mana of the diety with which you have the most influence, OR the presence/sacrafice of a clergy leader, etc.
- A temple in the province.

Pinnacle of the Sky: Fog of War is lifted permanently from all provinces adjacent to where the pinnacle is located.
Reqs:
- A mountain tile in the province
- A high construction tech
- At least 5 watchtowers, forts, etc. in provinces you control

Shimmering Palace: Increases happiness in all your provinces, and doubles commerce in the province it is built. Replaces your capital building.
Reqs:
- Two active gold mines in th province or provinces adjacent to that province
- At least two trade routes through the province

Wrathful Orb: An actual demi-diety, trapped within this great crystal orb. Has a percentage chance (10%) each turn to cast a random offensive spell against an enemy province or unit stack. 20% of the time, the spell will target you instead. Any player may pay 10 mana of any kind to stop the Orb from casting this turn, or you may pay 5 to do so.
Reqs:
- Sacrafice one magic great leader, or equivalent
- No more than X influence in any one diety
- Must be the agressor in at least one war

Brooding Orb: The machinations of this demi-diety are unknown. Provides science points towards a tech at random intervals, but lowers happiness slightly in all provinces. If any espionage action targets you and fails, there is a 50% chance the theives guild carrying out the operation will "mysteriously" vanish. Unfortunately, the Orb has a very small chance per turn of activating espionage actions for you. These cost you no money, but the Orb will even target allies...
- Sacrafice one magic great leader
- Must have built at least 5 theives guilds, and at least one successful espionage mission

FACTION WONDERS (One per faction only):

Rallying Memorial: Requires the death of one Military Leader in combat; maybe kicks up the defense/morale of all units in the province it is built.
Reqs:
- Death of one military leader

War College: You may train Captains and Generals for gold as regular units in this province.
Reqs:
- At least 5 barracks in provinces you control
- At least 20 military units active currently

Rockface Citadel: Requires the building of at least 5 walls in your provinces, and their upgrade into keeps/castles/high walls/whatever. Additionally, the province which you build it in must contain mountains or hills, and you must have won a major (x units) battle against an opponent in the province. Instead of a regular wonder, the Rockface Citadel is a super-fortress tile improvement, something like triple/quadruple defense and vision of the whole province even if you lose control of the province.

Diabolical Interrogation: Not a building, requires some magical requirement as well as the capture of an enemy leader. All enemy terrain is visible to you for one turn, as well as all enemy units there. The wonder has various compromising effects depending on the leader you capture: A military leader might mean the weakening of that Faction's troops. This may be built multiple times, one for each leader, but lowers honor each time.

DIEIFIC WONDERS (One per diety only):

Moon Obelisk: Requires at least 4 field tiles in the province it is built and x Ollux influence. Your astrologers can calculate the position of the moon with this pseudo-magical relic, and all food production in fields in this province is increased by 100%.

Submerged Temple: All of your lakeside provinces gain increased happiness. In addition, you may now use flood spells and their equivalent on any province two provinces adjacent to water (i.e., not just on adjacent provinces to water).
Reqs:
- X Ollux influence
- Must be built in a Lakeside province, with a Harbor

Cathedral of Dominion: As long as this Cathdral stands, the province it is built in never converts, and is twice as powerful in converting neighboring provinces to J'Honsa. All units built in this province gain +1 attack permanently, and never retreat.
- X J'Honsa influence
- Requires barracks and temple
- At least 10 units must have perished against a single foe with J'Honsa influence less than X

Prophecy Stones: Looks similar to stonehenge in some ways, but much bigger. All provinces bordering yours have fog of war permanently lifted from them as if they were your own.
- X Kaylar influence
- Requires temple

	09-06-2002 23:39									#47
	Got spare money?

Hi,

A couple of thoughts about wonders. Amazing how your mind wanders when ironing shirts for work but never mind

1. How about making Great Wonders dependant on building one or more Small Wonders? As a non Candle'Bre rationalising I would give the ancient Egyptian monuments. They started primitive, got more advanced and finally ended up with the Great Pyramids. The experience used in building the earlier monuments enabled the construction of the later. By comparison, say five province walls give the Rockface Citadel, which in turn enables the construction of.... oh I dunno but something pretty good.

Gamewise I think this ties in with an idea I saw on the tech tree thread, in that a given faction will *not* reach the top of all three (four?) tech trees, but has to pick and choose a bit. Similarly with Wonders - the requirements for building the Great Wonders will mean that a single faction will not be able to build them all.
Balancing the effect of the Great Wonders may be tricky though, they'll have to be useful enough that they warrant being built, but are still intended to be kinda 'icing on the cake' - pretty useful and perhaps essential to a *couple* of strategies, but not gamebreakingly powerful. Which ties in with a comment on another thread - from Vel I think, which said more-or-less that a lot of the additional features in the game - espionage, magic, influence and stuff, should be extremely helpful, but probably not strong enough to base an entire winning strategy on.

As for AI wonder building. Would it be feasible for the AI to pre-designate a province as a 'wonder province' and then have a fixed build queue which would provide all the physical prerequisites for wonder construction, provided the necessary techs or what have you were discovered?


just a couple of random thoughts anyway.

Rich.

	11-06-2002 06:29									#48
	Help yourself to an AD-FREE life

Orbs

I originally thought of the first two orbs just as wonders, like the above. They are semi-deific beings which are contained in a physical body. Who knows where they came from, or what they truly are... when they are discovered, all that hints of a consciousness is a vague feeling of great power that surrounds them. It takes the sacrafice of a magic leader to "awaken" them, and even then they manifest only as a voice in the mind, subtle emotions radiated from within.

The orbs sound like a cool idea to me, so I have for your enjoyment some descriptions of orbs. Just for fun...

Brooding Orb- This orb is a dull grey like all orbs, units its activation. At this point, the orb is a glossy dark grey, brewing in the orb like dark storm clouds. The feelings associated with this orb are suspicion and taciturnity. The puropses of this orb are totally alien, and it rarely expresses its wants to anyone.
PRO: Espionage retaliation. Enemy espionage attempts have a greater chance to be detected, and if they are there is a 50% chance the theives' guild perpetrating the act vanishes without a trace. The orb does not take being spied upon kindly.
CON: Forced espionage. The orb may carry out an espionage action in your name every so often... which may be against one of your allies. You must take the honor loss and any diplomatic reprecussions, because there is no way to prove it was the orb's doing. "Did I do that?"

Wrathful Orb- Once activated, this orb is a dark red, churning with streaks of bright red. Feelings of the orb include blind rage and terrible hatred. The orb does not communicate any desire short of utter annihilation of all that is not itself.
PRO: There is a chance each turn to target enemy with offensive spell. Has a percentage chance (25%) each turn to cast a random offensive spell against an enemy province or unit stack. Any player may pay 10 mana of any kind to stop the Orb from casting this turn, or you may pay 5 to do so. The orb prefers spells of wanton destruction, like floods or pestilence. Targeted spells (immolate, lunacy) are not as favored.
CON: You might get burned, too. There is a 20% chance that the spell cast will hit you instead. "Ouch!"

Watchful Orb- This orb is sky blue when activated, with white blotches slowly meandering over it. Occasionally, the orb can look eerily like a great blue eye. Emotions associated with it are intense suspicion and envy. The orb is relatively communacative and probes for information wherever it can. People near the orb always feel very uncomfortable, as if they are being stared at form behind.
PRO: Enemy provinces on your borders have the fog of war partially lifted from them. In addition, you may see through enemy Sleepless Sentinels (see society model thread). Excellent intelligence.
CON: The orb may give enemy a random view into one of your provinces. There is a 50% chance each turn it will reveal one of your provinces to all enemies for one turn. "Hey! How did he know that?"

Ravenous Orb- Once activated the orb is a constantly churning brown. Emotions of the orb are a cold indifference coupled with a terrible hunger of destruction. One of the most vicious orbs...
PRO: Enemy population can... vanish. Population in provinces adjacent to yours may simply dissapear... at least they can't blame it on you.
CON: Your captured units may vanish too. Units you have captured may vanish forever, and you take the honor hit fomr killing them. It's especially bad if you captured a leader. "I didn't mean to kill your leader..."

Null Orb- After activation, this orb is jet black in color. The room it is in feels icy cold, and sound and light seem to be dampened by its presence. There are no feelings associated with the orb, only an icy emptiness.
PRO: Counteracts the effects of other Orbs 50% of the time. When an orb targets you, even if it is your own orb, the effect may be entirely dispelled.
CON: Drains your Mana. The orb is anti-magic, and randomly drains a random amount of your mana every so often. The more man you build up, the more likely it is to drain some. "I thought I had enough to cast my spell!"

Soothing Orb- This orb is a light golden color when activated. It exudes feelings of complacency, even irrelevancy. It is fairly expressive in its purpoted desire of making all beings content, but its true purpose is unknown.
PRO: Greatly increased population happiness. The soothing orb makes your people quite content, although excessively bad treatment will still cause reprecussions.
CON: May decrease population productivity. Certain projects in your provinces may stall because the production of the province has been halved for a turn. "Work faster, dammit!"

Blood Orb- This orb is a solid blood red when activated. Feelings associated with it are seething anger and a delight in the pain of others. It expresses its motives only rarely, although it makes its love for combat and death clear, as well as its desire for eternal war.
PRO: Often gives an offensive boost to your units in battle. Your units may randomly receive a +2 offensive bonus for an entire turn. This can be very valuable.
CON: Random loss of hp. Your units may randomly lose one or more hit points for no reason. This never results in the death of your units, but it certainly can be an annoyance. "How did they die so fast?"

Ancient Orb- This orb remains the grey color of dormancy even after activation. It is reputed to be old in the extreme, and it hints at knowing secrets far older than the cataclysm. Feelings associated with it are a scorn borne of age as well as pessimism.
PRO: Random tech advances. The orb may from time to time give you a large amount of research points that goes directly toward what you are researching.
CON: The orb may summon Nilrogg units. Randomly, Nilrogg units (as per the Ollux spell "terrible manifestation) may appear in one of your territories, causing havoc. This is the rarest of all events, but its power makes it significant. "AAIIIEEEE!"

	11-06-2002 08:59									#49
	Help yourself to an AD-FREE life

To allay some fears about AI inflexibility, we AI people haven't definitely decided on anything, but one thing we are seriously considering is using neural nets.

	11-06-2002 14:54									#50
	Enter the AD-FREE zone

Ramo: would you care to explain why? Not that I doubt you've researched the subject extensively, but as far as I know, people use NNs for tasks like image recognition, and goal matrices for most games. I'd be interested to hear why you think the ordinary approach is "wrong" and how NN can improve on it.

Last time someone tried to hype NN usage in a game context we ended up with BC3K, after all!

	12-06-2002 00:42									#51
	Sure. Support Apolyton

Ok, problems with traditional AI's:

1. Internal flexability. If we change small aspects in the game, certain strategies and tactics that the progammer uses for the AI may be totally invalid. Particularly with the editing capabilities we'll provide y'all, this could pose a problem.

2. External adaptability. A traditional AI can usually implement only rough variations of strategies the programmers come up with. Which brings me to:

3. Elegance and simplicity. Granted, these are rather superficial reasons, but things that cut down on the amount of code (and therefore bugs, as well as time needed to modify the code) are good.

A learning AI could fix these problems to a great degree. The verdict is still out on whether such an AI could competently play/learn a grand strategy game (or even parts of the game) and whether it's too expensive.

Last edited by Ramo on 12-06-2002 at 00:54

	12-06-2002 02:51									#52
	Re: Sure. Support Apolyton

quote: Originally posted by Ramo 3. Elegance and simplicity. Granted, these are rather superficial reasons, but things that cut down on the amount of code (and therefore bugs, as well as time needed to modify the code) are good.

Not sure if the total lines of code would actually be smaller for this neural net AI. Are you sure of this?

quote: The verdict is still out on whether such an AI could competently play/learn a grand strategy game (or even parts of the game) and whether it's too expensive.

This isn't just a dreamy theory discussion. They actually intend to make a game. You need to resolve this issue before pushing ahead with a "radical" coding scheme. Otherwise you are setting the project up for pain and failure. (In other words, nothing wrong with being innovative, but the burden of proof is on you to show that you can make this scheme work...ene so many others have rejected it.)

	12-06-2002 05:21									#53
	Support Apolyton, buy Alpha Centauri

quote: Not sure if the total lines of code would actually be smaller for this neural net AI. Are you sure of this?

It isn't necessarily the number of lines of code that's the problem, but, well, how elegant the code is. As in the number of special cases, the complexity of the triggers for each case, and so forth. Basically, I want code that won't make a programmer go insane looking at it.

The main advantage with respect to the design when using NN's would be the nearly complete lack of special cases.

quote: (In other words, nothing wrong with being innovative, but the burden of proof is on you to show that you can make this scheme work...ene so many others have rejected it.)

How can I have any proof without trying it first?
What we're planning is to make a few different designs concurrently.

If NN's end up failing horribly, don't worry 'cuz we'll have other options open.

	12-06-2002 12:03									#54
	Remove this text

You can do more analysis to understand the difficulties you've already mentioned, Ramo. Leave the cheap HS debate world...

	12-06-2002 13:06									#55
	Support Apolyton, buy Galactic Civilizations

quote: Originally posted by Ramo If NN's end up failing horribly, don't worry 'cuz we'll have other options open.

If you think you can pursue several different tracks, given that this is quite possibly the single most difficult implementational task you face, then I won't agree we you. I'll still be interested to see how it pans out, of course.

It's not like you have a deadline, after all.

Is this NN stuff in for the free version, BTW?

	12-06-2002 18:32									#56
	Support Apolyton or Terrorists Win

quote: Originally posted by cyclotron7 Orbs I originally thought of the first two orbs just as wonders, like the above. They are semi-deific beings which are contained in a physical body. Who knows where they came from, or what they truly are... when they are discovered, all that hints of a consciousness is a vague feeling of great power that surrounds them. It takes the sacrafice of a magic leader to "awaken" them, and even then they manifest only as a voice in the mind, subtle emotions radiated from within. The orbs sound like a cool idea to me, so I have for your enjoyment some descriptions of orbs. Just for fun... Brooding Orb- This orb is a dull grey like all orbs, units its activation. At this point, the orb is a glossy dark grey, brewing in the orb like dark storm clouds. The feelings associated with this orb are suspicion and taciturnity. The puropses of this orb are totally alien, and it rarely expresses its wants to anyone. PRO: Espionage retaliation. Enemy espionage attempts have a greater chance to be detected, and if they are there is a 50% chance the theives' guild perpetrating the act vanishes without a trace. The orb does not take being spied upon kindly. CON: Forced espionage. The orb may carry out an espionage action in your name every so often... which may be against one of your allies. You must take the honor loss and any diplomatic reprecussions, because there is no way to prove it was the orb's doing. "Did I do that?" Wrathful Orb- Once activated, this orb is a dark red, churning with streaks of bright red. Feelings of the orb include blind rage and terrible hatred. The orb does not communicate any desire short of utter annihilation of all that is not itself. PRO: There is a chance each turn to target enemy with offensive spell. Has a percentage chance (25%) each turn to cast a random offensive spell against an enemy province or unit stack. Any player may pay 10 mana of any kind to stop the Orb from casting this turn, or you may pay 5 to do so. The orb prefers spells of wanton destruction, like floods or pestilence. Targeted spells (immolate, lunacy) are not as favored. CON: You might get burned, too. There is a 20% chance that the spell cast will hit you instead. "Ouch!" Watchful Orb- This orb is sky blue when activated, with white blotches slowly meandering over it. Occasionally, the orb can look eerily like a great blue eye. Emotions associated with it are intense suspicion and envy. The orb is relatively communacative and probes for information wherever it can. People near the orb always feel very uncomfortable, as if they are being stared at form behind. PRO: Enemy provinces on your borders have the fog of war partially lifted from them. In addition, you may see through enemy Sleepless Sentinels (see society model thread). Excellent intelligence. CON: The orb may give enemy a random view into one of your provinces. There is a 50% chance each turn it will reveal one of your provinces to all enemies for one turn. "Hey! How did he know that?" Ravenous Orb- Once activated the orb is a constantly churning brown. Emotions of the orb are a cold indifference coupled with a terrible hunger of destruction. One of the most vicious orbs... PRO: Enemy population can... vanish. Population in provinces adjacent to yours may simply dissapear... at least they can't blame it on you. CON: Your captured units may vanish too. Units you have captured may vanish forever, and you take the honor hit fomr killing them. It's especially bad if you captured a leader. "I didn't mean to kill your leader..." Null Orb- After activation, this orb is jet black in color. The room it is in feels icy cold, and sound and light seem to be dampened by its presence. There are no feelings associated with the orb, only an icy emptiness. PRO: Counteracts the effects of other Orbs 50% of the time. When an orb targets you, even if it is your own orb, the effect may be entirely dispelled. CON: Drains your Mana. The orb is anti-magic, and randomly drains a random amount of your mana every so often. The more man you build up, the more likely it is to drain some. "I thought I had enough to cast my spell!" Soothing Orb- This orb is a light golden color when activated. It exudes feelings of complacency, even irrelevancy. It is fairly expressive in its purpoted desire of making all beings content, but its true purpose is unknown. PRO: Greatly increased population happiness. The soothing orb makes your people quite content, although excessively bad treatment will still cause reprecussions. CON: May decrease population productivity. Certain projects in your provinces may stall because the production of the province has been halved for a turn. "Work faster, dammit!" Blood Orb- This orb is a solid blood red when activated. Feelings associated with it are seething anger and a delight in the pain of others. It expresses its motives only rarely, although it makes its love for combat and death clear, as well as its desire for eternal war. PRO: Often gives an offensive boost to your units in battle. Your units may randomly receive a +2 offensive bonus for an entire turn. This can be very valuable. CON: Random loss of hp. Your units may randomly lose one or more hit points for no reason. This never results in the death of your units, but it certainly can be an annoyance. "How did they die so fast?" Ancient Orb- This orb remains the grey color of dormancy even after activation. It is reputed to be old in the extreme, and it hints at knowing secrets far older than the cataclysm. Feelings associated with it are a scorn borne of age as well as pessimism. PRO: Random tech advances. The orb may from time to time give you a large amount of research points that goes directly toward what you are researching. CON: The orb may summon Nilrogg units. Randomly, Nilrogg units (as per the Ollux spell "terrible manifestation) may appear in one of your territories, causing havoc. This is the rarest of all events, but its power makes it significant. "AAIIIEEEE!"

Cool! I like them. A thought - what about assigning an element to each orb and then needing to place the orb on a sigil of that element to activate it.

In fact that could be a victory condition. For a thaumaturgical victory you need:

1. 5 binary sigils (sigils that confer abilities in two of the five Elements - see Magic thread).
- optional : the five sigils should be 'chainable' ie they might be fire/air, air/water, water/earth, earth/spirit and spirit/fire (thus completing a circle.)

2. An appropriate Orb placed on each sigil

3. A quinary (??) sigil, ie. one that powers all five Elements.

4. The Ancient Orb placed on the quinary sigil!

The quinary sigil won't necessarily be at the centre of the other five. However the easiest way to make such a sigil would probably be to divert one of the lay lines (see Magic thread again) from one or more of the binary sigils to a focal sigil. So it might end up being roughly in the centre anyway although this is not essential.

This is probably more of an 'Easter Egg' way to win than anything else - these requirements seem kinda high to me. Havn't really got a jazzy name for the final structure either, but basically its awesomely powerful enough that the other factions surrender to its owner.

Rich.

	13-06-2002 06:27									#57
	Enter the AD-FREE zone

quote: You can do more analysis to understand the difficulties you've already mentioned, Ramo.

Ok, I'm almost certain that NN's could be be used just fine in many circumstances; for example when I'm trying to approximate simpler functions (situations).

I don't know if the NN's could effectively be used in more complex circumstances, or if this approach will be able to be pieced together effectively. And there isn't any theoretical analysis that could answer my questions better than actually *doing it.*

Besides, since I'm the only real proponent of NN's on the team, I'll probably be the only person wasting my time on it.

quote: If you think you can pursue several different tracks, given that this is quite possibly the single most difficult implementational task you face, then I won't agree we you. I'll still be interested to see how it pans out, of course.

Creating multiple implementations of an AI is not that difficult to accomplish. The neural net itself was created relatively quickly (at least for me ), and the implementation I've already planned out more or less, which could be created reasonably quickly, as long as I don't make too many major bugs, once other parts of the project come together.

Same thing with another AI design I have in mind.

quote: Is this NN stuff in for the free version, BTW?

If the NN works out well, I don't see why not.

	13-06-2002 21:05									#58
	Support Apolyton, buy Civilization: The Boardgame

quote: Originally posted by Ramo Creating multiple implementations of an AI is not that difficult to accomplish.

Oh, but it is. While I'm sure doing a basic CS-cookbook NN net isn't especially time-consuming, actually getting it to do AI worth the paper it's printed out on is. This goes for all AI approaches.

Either you're incredibly talented and have solved the task that has eluded so many development teams - in wich case I suggest you get off the CB team and go work for a real developer right now, they'll be happy to pay you a lot of money - or you're seriously underestimating the difficulty of development ahead.

	13-06-2002 21:15									#59
	Suffering from ads?

Ramo,

Is it that hard to think about approaches to tasks without trying them? Why not do a little more research on NN work in practice? At least talk to other people who've tried this approach for a game AI and find out what their expereince was. Is that so hard to do? Am I the only one who thinks this way? I don't know jack about coding, but I've learned some things about life and about problem-solving and accomplishing tasks.

	14-06-2002 00:20									#60
	Lose 30 kilos (of popups)

Guys,

I'm going to start an AI thread here so this thread can get back OT.

R:PM