OT: Stardocks Fantasy TBS

[MaxWilson]

Originally Posted by Warhammer

Regarding the weapons, I think there are several reasons for this:

1) It is much easier to make an effective AI when you are dealing with aggregate defensive and offensive values.
2) As a ruler, you really don't get involved with determining what arcs the weapons on your ships have.
3) It makes the game much easier as far as micromanagement in the late game.

My understanding is that from Frogboy's (i.e. Brad, the designer's) perspective, #1 is the reason. GalCiv was designed around the AI, and making tactical combat simple and symmetrical was the only way to make strategic choices computable for the AI. Imagine trying to create an AI that can predict combat outcomes for ships in MOO2 or armies in MOM or Dom3!

Unfortunately, it also makes the games flat and uninteresting for those who are more interested in playing with neat toys than in abstract strategy. I happen to be one of those who would rather have rich tactical options and an AI opponent which "cheats" around strategic choices by getting 8x my resources[1] than a world-class AI who can beat me every time at tic-tac-toe. Obviously I'd really like rich tactical options and an AI which doesn't NEED to cheat, but for me, the interestingness of my own armies is paramount.

-Max

[1] Being strategically clever is only necessary when you're dealing with scarcity. If you've GOT a bigger hammer you can just use it.

[Baalz]

Yeah, my biggest beef with Galactic Civ was "ok I've now got +4 lasers instead of +3 lasers. Guess it's time to start researching +5 lasers so I can keep ahead of my opponents +4 shields..." Contrasted with Calahan's description of MOO (which is a pretty accurate depiction of how the game felt to me), it just felt very flat to me. Deciding to go with rapid fire shield piercing death rays or MIRV'ing ECM'ed missiles or tractor beams and boarding crews makes a pretty big difference in how your ships fight as opposed to generic weapon A, B, or C. Energy dampeners or shield capacitors? Point defense lasers, a lightning shield or good old ECM? Long range torpedoes or a bigger engine and point blank maulers? Punch through a section of the shield or go for enveloping weapons.

Damn, I think I might have just talked myself into digging up a copy of MOO2...

[P3D]

And in MOO2 you could even enjoy playing against Impossible AI (or whatever it was called) opponents with 400% resource bonus. I loved plasma cannons (they got really small as tech increased).

SEV(IV) is supposed to be similar, but the micromanagement... ugh...

[NTJedi]

Some good news about the multiplayer... looks like they've had some change of heart. I'm thinking the individuals who did send their concern made a difference. Looks like Hotseat and PBEM are not guaranteed yet... so anyone who enjoys those multiplayer aspects may wish to contact Stardock.

EDIT: June 2nd:

{Developer Frogboy reply="4" id="2237359"}
PigeonPigeon:And not to be a pain, but could you shed a little light on whether LAN/Hotseat/PBEM multiplayer will be available? You made a lot of people concerned in your getting back to Elemental post.

Frogboy:
LAN almost certainly. Hot seat maybe. PBEM, not sure yet.

[thejeff]

The trouble with most of the "gets 8x resources" approaches to AI is that they usually wind up with an AI who's by far toughest if encountered early. 8x advantage is very hard to overcome when you haven't had time to use your strategic and tactical advantages to build production faster to offset the AI's mechanical advantage. Later, the AI will still get 8x the resources from each planet/city/province/whatever, but you'll have more of them to counter with. Starting out you've each got 1.

I haven't played GalCiv, but this seems to apply to most 4X type games. On the hard levels, if I get attacked by an AI early I get swamped, if not I win easily. I wonder if a progressively increasing multiplier for the AI would work better.

[MaxWilson]

Originally Posted by thejeff

The trouble with most of the "gets 8x resources" approaches to AI is that they usually wind up with an AI who's by far toughest if encountered early. 8x advantage is very hard to overcome when you haven't had time to use your strategic and tactical advantages to build production faster to offset the AI's mechanical advantage. Later, the AI will still get 8x the resources from each planet/city/province/whatever, but you'll have more of them to counter with. Starting out you've each got 1.

I haven't played GalCiv, but this seems to apply to most 4X type games. On the hard levels, if I get attacked by an AI early I get swamped, if not I win easily. I wonder if a progressively increasing multiplier for the AI would work better.

I don't think the 8x approach is a substitute for a good AI in a 4X game, it's just a substitute for the ability to do tactical analysis. Instead of being able to exhaustively analyze the outcome of a particular battle to decide whether you can engage his 34 missile frigates with your 4 dreadnoughts, just make some unrealistically-optimistic assumptions about the capabilities of 4 dreadnoughts vs. 34 missile frigates, but what you actually send into battle are 32 dreadnoughts.

That doesn't mean you don't have to know how to expand efficiently, but an AI which can't expand isn't going to do well even if you do dumb down combat to rock-paper-scissors.

Edit: to address your point about 8x being too much when encountered early, all I can say is "good point." Progressive does sound like a better idea in that light, unless the game starts at a point where the human already has a lot of tactical options (Advanced start in MOO2).

-Max

[BesucherXia]

I have played the second Galciv2 expansion as a beta tester (preordered it as soon as they started advertising), and was disappointed to find they had broken too many words like the version schedule. IMO they were delaying the game just because they focused on another title(Sin of a Solar Empire) while relying on the fans to improve the game for free. Almost every week people were complaining about the unplayable AI and the infamous bug that made their ships disappear, but its only near the end of the beta phase we saw they were finally fixing them.

I know that was reasonable for business, but when the game was finally released I had lost my interests, specially when I saw the minor bug I had noticed them months ago were still there after so many updates.(the in-game picture of an unique building is missing because they have misspelled the file name). Maybe they have blocked my gmail like spams, who knows.

Though I loved Galciv2, I have never checked their games since then. Prefer saving time for another MP.

[Dragar]

A focus on singleplayer gaming requires AI to be a top priority. I love Dom3 but playing SP bores me to tears. I just can't make muyself play past the early game

The devs here have clearly developed a game for mp only, which is fine, but if stardock want to sell a lot of copies they need to have a great AI, which isn't possible with Dom3 type complexity

[Gandalf Parker]

Ive known some complicated games that did decent with AI by making the AI moddable by the players. I think Dom3 (ok maybe Dom4) should have gone that route.

I think that any game developer should recognize the areas they are leaving light, and make those areas moddable.

[Kaffa]

dom3....the best TBS fantasy game.
what do you think about mod fantasy Fall from heaven 2 ? there's some likeness with elemental?