OT: Starcraft II and Elemental

[Doo]

I kinda suspect that Stardock would like to have put more things in at release but didn't have the time and it is good enough as a release version. I suspect a lot of crinkles will be ironed out over time.

You can see the Stardock team sitting around a table saying:

"Only some seeds give good starting resources, a bad seed effectively screws the game."

"Oh poo...."

"Lets release it on schedule and run with it, more maps can be added over time. It will fly."

And thats ok by me.

*edit* Immediately after writing this I went to the Stardock Elemental forum and guess what, Version 1.01 released with more random seed maps

[Folket]

I would buy Elemental if it had hotseat. As it stands now I will pass on that game.

[HoneyBadger]

Gandalf: Stardock gives good AI, but consider that it's AI in GalCiv is operating in a really simple combat environment.

That's always seemed a little suspect to me...

It feels like, once you've got the AI performing well in terms of overall strategy, that it shouldn't be to huge of a step to code in some kind of tactical framework that can go beyond "paper-rock-scissors".

People have been designing competitive chess games for well over a decade, that are able to outperform genius-level human players (1997: Deep Blue vs Kasparov, 1998: Rebel 10 vs Viswanathan Anand etc.), so why successfully create an AI that can perform on a macro-scale, but then not even make the attempt at one that can compete in a deep microgame? It's a letdown, in retrospect.

Another letdown is that the Dreadlords (the big bad Balrog-types) are neither capable of "learning" (they never adapt new weapons or defenses to the tactics of others), nor of fully operating on the grand strategy level (they don't take new planets), so that, to me, is a major failure of the AI towards giving the single player (which this style of AI seems to have been adopted for the benefit of--obviously, human players wouldn't require these kinds of handicaps), the ultimate challenge of trying to overcome an enemy that is not only strong, but in some fashion, smart.

It's a perpetuation of a tired old cliche: that of the supervillain who exists only to wait around for the plucky heroes to knock him off his dark throne. Once again, Tolkien in space. And in this case, there's not even the excuse of a Barad-Dur for the villain to be tied to, or a Cracks of Doom/Ring of Power weakness.

In GalCiv, Sauron has a spaceship.

[LDiCesare]

Stardock's handling of maps is very bad. They've been told to open map scripting to modders, who proved with Civ IV that they can do it.
No, they preferred to do it themselves. Now they have a big map generator, that is slow, that has to pick bits of here and there because they can't think of a non-fractal random map or can't create good fractal random maps and make them fast...

As for the ai, it will be decent enough given time, but right now, the Gold version is just worthy of a beta.
And I concur that Stardock games had good ai's because of simple rules.
Now, it's much harder with Elemental, which has way more varied rules, and in which many tricks will come, including unforeseen ones I believe, like self-hasting fire giants.

[Zeldor]

Is there anything interesting in Elemental that would make you want to stick longer to it?

And some questions:

1. How to deal with Golems? They pretty much insta-kill any armies.
2. Or maybe auto-resolve comabts sucks so bad that with 10x higher attack you lose anyway?
3. Is there any hard limit for tiles for my cities? OR any pattern? Some of my cities have same size, but 2x less tiles...
4. Can I make more adventurers or do I have to run like crazy with my sovereign to check all huts taht pop up everywhere?
5. Do games take hundreds of turns, or is it just me?
6. What's the strongest sovereign? So far it looks like the one starting with 2 units just owns everything, as he can amass insane amount of gold etc from monsters and treasures. what's +1 gold per turn if you can find 200 gold chests?

[sector24]

Originally Posted by Zeldor

Is there anything interesting in Elemental that would make you want to stick longer to it?

And some questions:

3. Is there any hard limit for tiles for my cities? OR any pattern? Some of my cities have same size, but 2x less tiles...
6. What's the strongest sovereign? So far it looks like the one starting with 2 units just owns everything, as he can amass insane amount of gold etc from monsters and treasures. what's +1 gold per turn if you can find 200 gold chests?

3. You are limited to 50 tiles per city. If you put a resource inside your city (gold mine, etc.) that takes up 4 tiles. But it is put behind the city walls, so there's a tactical decision between freeing up 4 tiles and preventing bandits from burning down your gold mine.

6. Statistically speaking, some of the pre-made sovereigns are unbelievably powerful compared to the custom sovereigns. A few have more than 50 points placed in stats alone, and then on top of that they have abilities and equipment. The only advantage of a custom sovereign is the stacking of ridiculous bonuses, like getting all the +movement bonuses and the ability that allows all units to move at the same speed as the soverign (or something like that).

[Doo]

Originally Posted by HoneyBadger

Gandalf: Stardock gives good AI, but consider that it's AI in GalCiv is operating in a really simple combat environment.

That's always seemed a little suspect to me...

It feels like, once you've got the AI performing well in terms of overall strategy, that it shouldn't be to huge of a step to code in some kind of tactical framework that can go beyond "paper-rock-scissors".

People have been designing competitive chess games for well over a decade, that are able to outperform genius-level human players (1997: Deep Blue vs Kasparov, 1998: Rebel 10 vs Viswanathan Anand etc.), so why successfully create an AI that can perform on a macro-scale, but then not even make the attempt at one that can compete in a deep microgame? It's a letdown, in retrospect.

Another letdown is that the Dreadlords (the big bad Balrog-types) are neither capable of "learning" (they never adapt new weapons or defenses to the tactics of others), nor of fully operating on the grand strategy level (they don't take new planets), so that, to me, is a major failure of the AI towards giving the single player (which this style of AI seems to have been adopted for the benefit of--obviously, human players wouldn't require these kinds of handicaps), the ultimate challenge of trying to overcome an enemy that is not only strong, but in some fashion, smart.

It's a perpetuation of a tired old cliche: that of the supervillain who exists only to wait around for the plucky heroes to knock him off his dark throne. Once again, Tolkien in space. And in this case, there's not even the excuse of a Barad-Dur for the villain to be tied to, or a Cracks of Doom/Ring of Power weakness.

In GalCiv, Sauron has a spaceship.

I thought these supercomputers used a brute force AI, that they simply run play out millions of optional moves and where they lead and then pick the best.

[LDiCesare]

Originally Posted by Zeldor

Is there anything interesting in Elemental that would make you want to stick longer to it?

Good potential when there's an ai, a better ui in places, and some more balance. hopefully good modding capability too.

1. How to deal with Golems? They pretty much insta-kill any armies.

You mean fire giants, stone giants, etc? Yes, they kill everything. Best counter: Summon one of your own. Next: You need lots of attack to get past their armor, so need good weapons and lots of units grouped together. So buy a company of soldiers equipped with longsword or heavy warhammers or the like. Just make sure to have many individuals in the unit, each of them dishing out lots of damage.
Alternately, accumulate insane intelligence on your sovereign and cast some spell at it, particularly if you own the shards that go with the spell.

2. Or maybe auto-resolve comabts sucks so bad that with 10x higher attack you lose anyway?

I wouldn't trust auto resolve except maybe for 1 on 1 fights. The tactical ai is so bad that playing by hand means you'll win evenly matched fights easily so...

4. Can I make more adventurers or do I have to run like crazy with my sovereign to check all huts taht pop up everywhere?

You can hire adventurers and the like. Some of them seem to keep roaming the map even late game.

5. Do games take hundreds of turns, or is it just me?

They do for me too. But then I use a magician sovereign, she casts spells, and if it takes 20 turns for her mana to refill, the game drags just because of that.

[Dhaeron]

Can anyone who already played it give a review of the tactical battles? For me, boring tactical combat is a complete dealbreaker. GalCiv 2 was promising until the first fight showed that braindead rock-paper-scissors scheme that had going. There's a reason i always come back to playing MOO3 when i'm in the mood for space 4X.
Well, SEIV is nice too.

[Doo]

I can write a good review in around 48 hours time, so far I have not committed much time to playing it because after work I'm pooped. Friday night I intend to have an epic 6 hour marathon, Saturday morning I'll give you a fair and proper rundown of the game.