"That's why I made EBZ a modern action hero when I recounted the great battle with the undead in my GROSS recruitment thread.

He's using a combination of a tech rush with the Technician model of deliberately playing real easy games on small maps to really churn them out. While I would have a problem with playing way below my skill, I cannot fault the results."


I enjoyed your piece, Gerakken. One of these days, I will find time to collate all the classics of role play stories in this board for posterity.

Yeah, it boggles my mind how someone like Technician (who was a beta tester I believe) could play 50 games at Beginner level. But hey, if it floats his boat, I have no problem with it. What I really want to see is if they would continue to do so if and when the scoring is truely adjusted for difficulty.

Hmmm I thought the purpose of the Metaverse was for honorable competition. I fail to see what is honorable about flooding the server with cakewalk victories on tiny maps. Frankly I am surprised that the developers permit such an obvious design flaw. In the strategy game genre it is typical to use victory point multipliers based on skill level. Generally the "normal" setting would give 1.0 * Points while sub-normal would be * .9 .8 .7 etc. for each drop in difficulty below the median. Coversely each step above the median aka intelligent *1.1 bright *1.2 etc. receives and incremental positive multiplier. This of course assumes linear progression in difficulty which as we know is not the case in Gal Civ. Maso is Wayyyy harder than Genius and the score multiplier should reflect this leap.
Even if Stardock has not had the foresight to implement such a method, I think the comunity will still be self balancing in the end. Ultimately the most valuable commodity even us gamers have is time. How much of your time are you willing to waste playing a game at a skill level that is in no way challenging? Since there is no prize for points in the metaverse I have a feeling the only people resorting to such methods in the long haul will be young children, and adults with a serious lack of self esteem.

Theleb

Ok Wanted to post my thoughts on Ray's Walkthrough of my War of the Ring challenge, and how it differed from my two failed attempts at this same challenge. All you beginners listen up!

Race Picks:

Max ship repair
max ship hitpoints
+5 PQ
+15% weapons

No habitable planets in range, starting colony to the nearest large clump of unexplored planets.
Buy Scout Outright (mitrosoft) send off to explore other planet groups in colony range
Que colony ship until new site found then prepare to purchase.
Que Soil enrichment.
Send Hero to autosurvey
Res comm theory
pass turns, new colony, get usual assorted junk from anomolies
Finish colony on earth send to PQ16 planet discovered. que another colony ship.
Research comm theory
Ecounter two of the other races, one of them the Drengin in both cases, with the Drengin adjacent in both cases.
research Universal trans.
attempt to trade tech with other races to stay current. Futile as every race seems to have STARTED the game researched all the way through diplomacy so I have nothing to trade or sell.
No good planets left so 3rd colony is PQ14. Every new colony que soil enrichment and constructor as have no warship tech yet.
Refuse first tribute demand from Drengin which is about 1000 more than my treasury has.
Research diplomacy.
reserach trade
drop my forth colony on another PQ14
Meet last remaining civs
trade two techs for one with a minor civ, all major civs have all of my tech and many more.
Refuse second Drengin tribute request which is for one of my best planets.
Begin researching sensors
opponent civ develops TG translators
Now face tough choice to rush buy a constructor and hose my economy utterly, or have no chance at all of making a starbase at any of the resources.
Drengin now posess defenders on every planet, multiple transports, several starbases, freighters. They also have every tech in the galaxy and then some, as well as the ability to support a fleet which is vastly larger than their economy can permit. It is at this point I realize that the Genius AI plays no smarter than any other, it simply gets techs and ships (with no maintenance) for free on a regular rapid schedule well outside the bounds of the game rules.
Research stellar cart
Drengin declare war, I lose two planets in a heartbeat, still have no ship techs, and they have corvettes deployed en masse and half a dozen transports.
Trade one-two techs with other civs, they are still technologically way ahead of me.
Other civs give me a couple weak ships in my war against Drengin, Drengin not at war with others, just me.
lose more planets to drengin, Beg other civs with good relations for enough technology to build at least some kind or warship. Of course they wont give tech away.
Lose another planet to Drengin, switch research to weapons/industry.
Finally build my first constructor, destroyed instantly by Dregin.
Drengin systematically declares war against every other race in the galaxy, and fights and WINS, against every race in the galaxy simultaneously.
actually get a constructor built and drop a starbase in home sector. All resources long since taken.
Drengin destroy my starbase.
Drengin destroy my freighters.
Reserach a couple more insignificant techs that all the other civs have. Down to 2 planets.
Altarians who have exact opposite alignment as Drengin surrender to them.
Drengin Conquer all minor races
Drengins acquire first glactic wonder and grav accel and tri steel.
Torians surrender to Arc.
Produce my first couple of battle axes for planet defense, killed instantly by drengin.
Drengin invade my last two planets killing me off. They are still at war with every civ, own 3/4 of the galaxy and are unstoppable.

GAME OVER

Mean play time 2 hours (taking turns slowly)
200 or so points
Played in Iron Man mode, no use of Ctrl-N.

Theleb

First lesson, research Weapon tech tree first. You have to go on the offense almost immediately.

Don't think you were looking for gameplay suggestions from me by posting your AAR but I'm going to offer them anyway

1. Leasing is extremely costly. Only lease if you are going to get extreme value (such as Grav acc for Rush), ie don't lease scouts or constructors.

2. It's useful to manually pilot Hero in early game. For rushes, my hero goes beeline for the little metal pieces which contain ship bonuses.

3. After more games, you would be familiar with the tech research path for the AIs. Focus research on other techs so that you will have something to trade.

4. Keep the end in mind. Since you have selected Rush picks, focus your research on techs that will help your rush. Trade, sensors and stellar cart do not. Those can come later.

5. Don't worry when the AI develop TGs, trade for them. They are often cheaper than leasing them yourself.

6. Constructors and SBs are low priority in rushes. You can get them later after establishing youself.

7. Fatal to have the Dregs attack you first. If they had warred against others first, their home planets would usually not be too well defended, leaving them ripe for a sneak attack. Defenders can be taken down by star fighters, esp with military picks. Gift your scout to Dregs for a minor relation boost. Or bribe Dregs to attack others first.

Seems like Eldin has departed from the Council.

All the best to you, Eldin. Have fun and keep on Galciving.

Well boss

I'm afraid that with my new serial number I've put the group score in jeopardy. I'm trying to rebuild my old score... bear with me

Don't worry too much about our current group score, philopoemen. The metaverse is a marathon not a sprint and considering that scores age, the biggest empires wouldn't be up there for long if they can't perserve.

I'm more concerned with buidling a stronger community spirit and personality here in the Council so I really appreciate the posts from all of you. And from our guests who like to hang out at the White Tower.

For those of you who can't find time to post, drop by when you are able to.

#530 by Citizen philopoemen - 7/4/2003 3:20:44 AM

First lesson, research Weapon tech tree first. You have to go on the offense almost immediately.

I am not sure I agree with you on this Philo. If we look at the AAR Ray posted on his smashing sucess in my War of the Ring (WOTR) scenario we notice his first four techs were as follows:

comm theory
univ trans
diplomacy
impulse drive

#530 by Citizen philopoemen - 7/4/2003 3:20:44 AM

First lesson, research Weapon tech tree first. You have to go on the offense almost immediately.

I am not sure I agree with you on this Philo. If we look at the AAR Ray posted on his smashing sucess in my War of the Ring (WOTR) scenario we notice his first four techs were as follows:

comm theory
univ trans
diplomacy
impulse drive

#530 by Citizen philopoemen - 7/4/2003 3:20:44 AM

First lesson, research Weapon tech tree first. You have to go on the offense almost immediately.

I am not sure I agree with you on this Philo. If we look at the AAR Ray posted on his smashing sucess in my War of the Ring (WOTR) scenario we notice his first four techs were as follows:

comm theory
univ trans
diplomacy
impulse drive

Guys my apologies for the triple post my internet conn hung and I could not tell if the submit routine was running or not. RAY can you edit out the redundancies with list master priveledges?

Hey Ray thanks for the strategy suggestions on my posted WOTR failures. I actually was looking for some direction but it was late, I was tired, and not being too eloquent.

point on leasing well taken.

ok seems starbases are not too important on higher difficulty settings.

I have not learned yet exactly what types of bonuses are found in each type of debris so have not been able to optimize this to a single focused strategy.

I noticed that the AIs do not tend to develop sensors, stellar cart so that is why I research them. The problem seems to be that on higher skill levels the ship techs opened via the shields path seem to be absolutely vital. So if I try and research those to not be eliminated early, once again I have nothing to trade.

I was not aware that the race picks I used were rush geared. I am still in the process of learning the balance and pace of the game.

My concern with AIS getting trade goods is that its usually my principal enemy that ends up with grav accel and tri-steel, and there is no way they will lease them to me.

Aye I agree It would be best to get the dregs to attack someone else first, but I see no real way out of this mess. Every other civ at normal or higher difficulty will have more techs and ships than me, which makes me the weakest; and threrefore most desireable target for the Dregs AI. The only thing I can offer to appease seems to be worlds which is a quick ticket to defeat in the early game.

Galactic Civ Concerns:

While I am still relatively inexpereinced, my play history and reading of this thread has lead me to some early concerns about this game.

1) One of the reasons I was attracted to Gal Civ is that while it is a 4X game, it offers paths to victory other than building up the biggest tank you can, as fast as you can, and mobbing your opponent with them. Unfortunately this appears to be a farce, as on high difficulty levels it looks like on most boards, we are back to the old Rush military blitz strategy that dominates both the RTS and turn based strategy market. Now I am a wargamer, so I do not mind some of this, but it seems to me the game loses some richness if there is only one dominant way to victory. Civilizations III helps compensate for this by making culture far more powerful than it is in gal civ. In Civ III you could not roll over an enemy by sending a single transport to a group of your enemies planets and grabbing them. What would happen in Civ III, is that within a few turns unless they were HEAVILY garrisoned, (and sometimes even if they were) they would all flip back to the orginal owner. This encourages other means of acquisition besides military conquest, or requires heavy garrisons which prevent the rush alpha strike mentality. Sadly in Gal Civ culture planet flips take WAY too long to achieve, and never occur in response to a group of barbarians grabbing one of your worlds.

2) I am finding the AI disappointing. The AI that provides the greatest enjoyment for me, is not the kind that gets to break game rules to be competitive, but the kind that is actually SMARTER. This allows you to watch the AI and actually LEARN something. the higher end gal civ AIS tend to follow the model launched by Sid Meyer in Civ I. The higher the AI, the exponentially greater it cheats, and the more agressive it is. Thats it. Nothing one can learn from, no way the human can replicate. This of course leads to very regressive strategies. Taking the example of CIV I, the way the human would have to compete with the high end AI (They get units and research for free, quickly, and pay no maintenance) is to stop all research yourself and then take over an AI controlled city to steal a tech. Then you would sally your defnding units out of the newly captured city leaving it defenseless for the AI to immediately retake. Then you attack and capture again, stealing another tech, REPEAT PROCESS AD NAUSIUM. This IMHO ruined what little "believeability" the game had, and led to a less than satisfying play experience. I can see this game is heading in the same direction...
By way of camparison I will use the WarCraft III AI as an example of how to play well, but within the rules. The only major game rule the AI breaks is that it knows the entire map ahead of time, and never has to explore. (This by the way is typical of almost all strategy games RTS or Turn based, for some reason developers cannot create decent exploration/scouting Algorithms for AIS that allow them to use only the knowledge they have legitimately gained and stay competitive) Of course all these other games do not tell you that the AI sees the whole board for free at game start, at least Gal Civ informs you that the AI is using this standard cheat. Frankly with the metaverse available, and established technologies such as neural net, I think Stardock wrote a cop out AI for this game. If they can not come up with good AI then they can allow the players on the metaverse to do it for them and simply incorporate that strategy via neural net. No cheating would be necessary.

What do the rest of you think?

As Staffa had noted in another thread "placating hostile civs", the Crippling game feels a lot different from a Maso one. Once a player gets a good feel of diplomatic mechanics of this game, it is not difficult up to Crippling to maintain good relations and to play a peaceful game.

However, on Maso, due to the massive AI bonuses, it is near impossible to maintain any form of military parity. The Galciv AI is designed to punish military weakness and Maso games rapidly deteroriate into war.

Why should Maso be made like this? The answer is that Brad wants a level that would thwart even the most experienced Galciver. It is evident from the fact that the Maso difficulty was increased time and again during this 3-4 mths after release once it was clear that there are players who beat it consistently. I still remember that at the very beginning, the Maso AI just had money bonuses, but now they have massive ability bonuses like +30PQ (contrast this to the fact that most of us thought that the +15PQ racial pick previously available in the 1.01 bonus pack was way overpowered for humans!).

Culture is usually less of an factor in the early game than Civ 3 because it takes a difference of a couple hundreds of influence points before they start to bite. But players have reported effective flipping early on through selecting influence racial picks.

In the late game, they can be very effective with the high influence bonuses(from trade goods, wonders, influence resources and high pop planets). I could frequently flip all remaining planets in an enemy sector quickly through conquering a single enemy planet with the lowest pop.

They can work defensively like Civ 3 too if the target Civ has very high influence. But I usually follow-thru with my invasions so I don't get into that situation. Note that planet influence is 5x the strength in the home sector as compared to adjacent sectors.

You will find the AI smarter when you play at Intelligent level, which is the level where the AI and human is on par. They have deliberately dumb the AI down at the lower levels to make it accessible to new 4X players. Beyond Intelligent, there are no additional AI routines but straight AI bonuses.

I can tell that you are probably an experienced TBS player. My usual recommendation is for vets to play at Normal immediately, then gradually progressing up. You will not find any level lower than that challenging at all.

The Civ 1 AI had blatant cheats like the 1 turn wonders but in Galciv, there aren't any blatant ones up to Intelligent other than the ones already disclosed like the map knowledge.

I don't think that the AI is a cop out. I've played my fair share of TBS and RTS and I think it is one of the better ones that I've seen (at Intelligent). The key reason I believe for the better AI is that they have kept the game mechanics and effects simple. There were a lot of fancy stuff in MOO2 but most of the time, the AI don't know how to use them, leaving the human with a tremendous advantage. There are problems like this too in Galciv (AI doesn't seem to know how to use AMMs, terror stars, etc properly) so there is definitely room for improvement.

Yes, personally I'm progressed beyond finding Intelligent challenging but I seriously doubt if there is an AI that would be a challenge to humans in the near future. Brad mentioned that he took out the "Transport to scan for danger each turn" routine because it would simply eat up the CPU cycles and cause the game to be unplayable for poorer machines. I like the fact that the AI has bonuses at the highest levels so that it would still be challenging for someone like me who has played a while.

The Metaverse was designed with the intent to also improve the AI. Though far from neural net (it is more Brad net at the moment), we have seen improvements and game balances made to the game through the analysis of Metaverse data. My only issue with the Metaverse is the seeming disinterest to fix the few more scoring inequalities (I've looked at 1.09 scores and they seem no different from 1.05) and this may result in the departure of top class players like Staffa who was instrumental in bringing about some AI improvements.

The only rules this one breaks are the economy at levels above painful, and pq at maso. Otherwise, it plays a stock game. It knows where the planets and stuff are, but that's it. and I have found that that knowledge is no real benefit. After all, if it were, you'd see the best human players rush to the tech that gives the same to us. We don't. It's been called the totally useless tech by some. Civ3's ai doesn't even come close to this one (production cheat, science cheat, base city resources cheat, military experience cheat, and to top it all off the military attack and defense ratios are changed (WITHOUT TELLING YOU WHAT THOSE CHANGES ARE!!! and a no fog of war to computer players (yes that's right) cheat. Your point that this is a cheating ai relative to other ai's is plain not so.

And lets not talk about wonder poofing and the invinceable phalanx cheats in civ1.

With good understanding of the game, its not that hard
to beat crippling(2nd highest level) without any
military action.

Maso is a bitch in this regard(though not impossible on
larger maps), but having one absurd level isnt
unreasonable

The scoring differation between difficulties is as you
noted badly broken and I hold out little hope it will
be fixed

Also, the difference in score due to the victory path
is unneccesarly punative for non military wins.

Ray: Very insightful comments about MOO2, did you
play it much? I was absolutely fanatical about it
multiplayer for years, still play it multiplayer on
occasion, there are some very good players out there
who still play online.

Moo2 was awesome for the depth it had but the Ai did
not handle it well at all Though if you play
with tactical combat off, it did prove to be quite
abit more difficult, as this removed almost all
the things the Ai was clueless about(tactics and ship
design)

I still greatly perfered playing humans though, tactics
and ship design is what made moo2 fun.

Double post, maybe ill invent something to put here later

[Message Edited]

I'm more a single player person, Staffa. Though I played Moo2 quite a bit, got rather bored after a while. I didn't play it near as much as Galciv.

Heh, I'm going to spam Maso/Tiny/Abundants till 40 games to see if the 40/40 has additional requirements. Solitair did mention that T-Man was considering difficulty and map sizes. I should have met all these by now.

The last 2 I played were great!! Heart-stopping action continuously for 5-8 years fighting against much more militarily superior opponenets. Then it was boring SimGC to get it above 3000pts. My adrenaline levels were so pumped that I had difficulties sleeping last night.

Ray--

Thanks for your thoughts. Your position seems to be that since GalCiv does what most TBS games do, that it is ok. I on the otjer hand see the potential for this game and want it to live up to it like a few rare others have. Ultimately for me the true measure of success for a title is replay value. From what I can see I think Galciv may have a long way to go at the higher levels to keep one stimulated.

I will take your word for it that you can placate hostile civs up to intelligent even if they are much more powerful than you, you have the wrong race picks, and you have nothing to trade. I'll chalk that up to inexpeience, poor lock, not using ctrl-n or reloading.

I however fundamentally disagree that MASO level MUST have massive bonus to be competitive for a high end player like yourself. Letting an AI cheat rather than making it smarter is a cop out, period. How many games of electronic chess have you played Ray? As you up the skill level on a chess machine/program it does not cheat in any way, but I guarantee you will eventually hit a skill level that you will consistantly lose to the AI on. So I will not buy that it is not possible to make super smart AI's when it has already been done.
Now let me ask you this. Would you find it challeneging to play against another human in a multi-player environement? What if that human was just as good at the game as you are Ray. Would he still need massive bonues on top of that skill to give you a good run for your money? Do you think Staffa could beat you head to head at least part of the time WITHOUT any cheats? I think you will submit that the answer is yes. So then the question is not how much cheating you have to allow your high end AI to do to compete with your high end players-- but rather how do you get your AI to be as SKILLED as the high end players using the same rules set. The answer is Neural net. This has already been done in video games, I am not proposing anything revolutionary.
Even the developers of GalCiv realize this, I seem to recall in my registartion to metaverse that one of the eventual features they had in construction was to use the game data players upload to be able to profile a player and let one play in effect a "clone" of that player.
Yes Culture is far less important in this game than CIV III period. I stand by my assertion that that is a mistake. How many of your high end games against high-end AIS do you win a cultural victory on? Why is that? Perhaps because the military route is safer, quicker, and more effective? If you answer yes to this then it means that the game is not well balanced. A well balanced game would be one in which every winnig strategy had an equal chance of sucess on the median board. How far are we from that?
I agree with you that culture can be effective in the late game. Of course if you are in a late game, that means you either are not skilled, or not playing to win. If you are skilled and playing to win your game will read like your AAR to my WOTR challenge, and it will never even get to mid game let alone late game.
Now I am an experienced strategy gamer, but I can tell you I found my games below normal challenging in GalCiv for my first few. In your first couple of games you are trying to figure out the interface, have no idea whats on the tech tree or how to achieve any of the victory conditions. The only game that I did not find challenging is my cakewalk game where I set all the AIS to foolish just to see what they would do. They did not collonize much, built no warships, never built bases on resources, and never challenged me.
Yes Civ I had many blatant cheats including a pleathora neither of us have mentioned. Hence my reasons for using it as the quintesstial example of a poor TBS AI.
MOO2 had an AI that was very similiar to the higher end Galciv. The computer cheats progressively more and becomes more agressive. Yes it did not due well with some of the unique techs. MOO2 is not a good example of how to make good AI. It is however a good example of how to weight games scoring based on difficulty level-- something Galciv does not seem to really do.
As far as the scanning routines and computer speed, it is not that uncommon in games to set up a text file .config that the user can toggle features and pathing nodes etc off/on/up/down whatever so that the faster machines can take advantage of this while the slower ones can still play.
I will have to disagree with your beef that the biggest problem with GalCiv is the broken scoring system. I agree that that is a problem, but not nearly as much of a problem as the redundant GAME PLAY experience. Let me put it this way, when we boil down galciv to its lowest common demoninator, at the higher skill levels does it not come down to who can build the biggest tank the fastest and send it to the enemy en masse? While galciv may have different graphics, sounds and effects, is this not ultimately the same exact game play experience as CIV I, Command & Conquer, Warcraft I, and countless others? It seems to me that the label STRATEGY game implies more than "Tank Rush Game", I wish this title would share that vision. :notsure:

Interesting discussion we are having here, a bit heavy for me on a lazy Sat morning but v interesting.

Will need to run some errands now, but will be back to resume our conversation.

Pontius--

Thanks for your thoughts, but I am not sure I agree with you. By altering the economy you effectively are cheating at the three items that build off it-- military, social, research; aka everything that matters. By altering PQ at MASO level you are in effect applying a blanket multiplier to the cheating the AI is already doing. Now it occures to me that the +PQ Maso cheat may be a double edged sword. Correct me if I am wrong, but once a planet gets a PQ upgrade it is permanent unless reduced by warfare etc. right? So in effect then, it would be possible to abandon low PQ worlds the player holds, thus inviting the MASO AI to conquer them. Then you immediately retake them, and viola you have a much higher PQ world in your empire no?
Also you took my comment about Civ III out of context. No where did I state that CIV III has good AI. What I said was that the cultural model of CIV III had a much more prounounced game effect than the one in Galvic, and for that reason the cultural model, and only the cultural model, of Civ III was superior IMHO.
As to your comment about CIV ignoring fog of war issues for the AI this is standard. AIS always get to see the entire board, every turn, in every game that I can think of.
The invincible phalanx cheat (otherwise known as the phalanx that sinks the battleship) in Civ I does not work exactly as you think. When I talked to Sid Meyer about it at CGDC he explained that he used a sliding adjustment based on the power of the two nations involved in the conflict. Thus if you were winning the game and fighting a CIV that was doing poorly, you would have a big negative to your att/def #'s and the computer would have a big positive bonus. It had nothing to do with phalanx's; spearmen got the same bonus and could fight off tanks etc. Sid did this to try and midigate a common problem in all strategy games which we call "the mid-game sag". Basically at the mid point of most strategy games when you have the edge, the game gets boring as the rest of your time is spent on mop-up. Sid was trying to re-inject excitement into the game by making it progressively tougher to get ahead the more points you had. This was far from the ideal solution to this common strategy game problem, but you have to credit Sid for being the first to implement at least some solution.
Finally I must reiterate that there ARE AIS that do not win by cheating-- unlike this game. I put forth Warcraft III previously as an example. Even at the highest skill level the AI is not allowed to break the rules, and it is very challenging.

I agree. For someone who tires of games rather quickly, I'm amazed that I have played Galciv 3 months non-stop. I kinda suspect that the Metaverse and the nice pixels under my post had something to do with it though. Seriously, I do enjoy the part of the game before critical mass is obtained, the rest of the game I play for the Metaverse which is a totally different topic.

Interesting that you use Chess as an analogy. I'm an average chess player and I've played a bit against programs and over Yahoo. Though I did say that Galciv is simpler in mechanics, it's complexity is most certainly much higher than that of Chess. With such a simpler game, we still require a supercomputer just to pose a challenge to the Grandmasters. Can we get Deep Blue to compute best counters to the next 100 potential moves that a human may make in Galciv? May be possible in the early game but with 100 colonies and dozens of ships in the late game, the different iterations would be tremdendous. And unfortunately most of us don't get to use Deep Blue for gaming.



As I mentioned, I believe the human mind is currently far superior to any AI in gaming. You seem to have a very high expectation of AIs in games. Can you share with us examples of complex strategy games (with or without neural net learning) which has superior AI capable of challenging top human players?

WC3? I've played a bit of the SP campaign and I found it a breeze for those scenarios where you are not faced with overwhelming odds. I certainly don't think I've learnt any moves from the AI. Credit should be given to the developers though for trying to think of unique scenarios which are not the usual build and bash (the frozen throne xpack got a good review from Gamespot bec of this).



Yes, that was a major drawing point considering that this will only be a SP game. Let's hope they refocus on this after the Xpack is done. But I get the feeling that the AI will only be learning our top players' tech research path which is nothing to scream about. Good players would make adjustments to cater to different game conditions.



I don't win on culture any longer because I've switched it off. Else I predict that 90% of my games will be won on culture as the winning threshold is too low and too easy to achieve. Agree that the different victory paths are not balanced and Brad has acknowledged as much by setting different score weightages. However, once a player reaches critical mass, the type of victory is easily theirs for the choosing. This is a long topic altogether.

Can we reach a game state where each victory path is equal in challenge and effort? I doubt so but I do agree that such a dream state would certainly increase the replay value of this game.



Theleb, you have captured quite a number of pertinent issues in your post which have already been discussed in some detail (esp by Staffa). Yes, I agree that the most skilfully played game is the fastest one too. The complication lies in the Metaverse, which does not reward players based on speed of completion but rather on the basis of us following a certain process (in this case empire building). So I admit that many of my games are not won based on the fastest possible moves because I care about scoring and the Metaverse.



Difference though is that Galciv does not cheat (other than those mentioned) till Intelligent. It cheats on Genius and Incredible.



It does on Maso for smaller maps but this is intentional. This level is not meant to be winnable (Brad said so). As mentioned, you can still have very different games up to Crippling/Genius which is the second highest level in this game. Maso is not meant to be played and conquered like the other levels, it is supposed to be an impossible level to frustrate gamers with big egos.

Theleb, I would suggest you put in a couple of more games at the higher levels and see if you would think differently after that. It is only fair to the game before we pass judgement on it.

No, all AI Maso bonuses are racial bonuses so only they get it. If humans capture AI planets, human racial bonuses would apply.



Hmm, but you have certainly pip my curiousity about WC3 again. Maybe I should go compete the SP campaign before I myself pass judgement on its AI.

I however fundamentally disagree that MASO level MUST have massive bonus to be competitive for a high end player like yourself. Letting an AI cheat rather than making it smarter is a cop out, period. How many games of electronic chess have you played Ray? As you up the skill level on a chess machine/program it does not cheat in any way, but I guarantee you will eventually hit a skill level that you will consistantly lose to the AI on. So I will not buy that it is not possible to make super smart AI's when it has already been done.


Chess is a very far cry from any tbs game on the market. It is a closed system. There is no diplomacy. There is no building. There is no spending ratios, no tech, no nothing other than moving a finite number of units (all with the same fire power and no defensive power. Ie: a pawn can completely destroy a queen if it moves correctly without any combat rolls. (this won't happen very often and next to never, but in theory, it can) There are no firepower and hitpoint differences in units to worry about. The only difference in pieces is how they can move. And there is only one ai player, not 5. So the analagy is totally invalid.
A tbs strategy game is further above chess on the complexity scale than chess is above hungry hungry hippos. If checkers rates a 1 on complexity, chess would rate a 10, and a tbs game would rate about 10,000. That's not to say chess is an idiots game. But A closed finite system is far different from the infinite possibilities a human can come up with in a tbs game. In a chess game, the human is strictly limited in possibilities. Not so in a tbs. Very very big difference.


The fact is, you keep talking about the ai as if it is intelligent. That is a misnomer. It has no intelligence at all. It is a machine. Despite the sophistication and miniturization these days, a computer works essentially the same as it did in the vacuum tube era. It is essentially still nothing more than a series of switches that are either on or off. It has the equivalent of rooms and rooms full of these vacuum tubes switched either on or off, and it's all electronic now but a computer still knows only two things. On and off. An "ai" is nothing more than a clever series of if-thens trigger-responses written by a clever programmer. It cannot outthink a human (in an infinite or nearly infinite move game simply because it cannot think. And no program can take into account these infinite possibilities. And the more complicated you make a game, the harder it is to make the ai competitive.
In fact, what you say has been done hasn't as your analogy is extremely weak.