next »

[Coelacanth]

A lot of older members here have probably heard of or played SSI's epic "Fantasy Empires" at some point or another.

Any abandonwarepirate/old game aficionado will tell you that it's incredibly frustrating to get the game working on newer computers. CPU killers don't do the trick, the battles are all way too fast to play.

And the battles were the only real innovation that the game had. They were the attraction. Played in autobattle mode the game was essentially just risk.

I want to see games with a battle system similar to fantasy empires. I want to see another real time, hotseat strategy game.

I was thinking that with my simple understanding of GML and gamemaker, I might be able to undertake a remake project. But before I start, I want to know what people think of the idea. Would people play it if it was recreated? Are there any new features that any of you would like to see? Has it already been done, and I just don't know it?

[Derek]

I've never played the original, but sure, I'd play a good remake!

According to Wikipedia, the game's AI used a neural network.  Wow, that's pretty ballsy!

[Coelacanth]

I've never heard the phrase "neural network," but the best part of the game was most definitely the player's ability to control the battles, not the artificial intelligence system.

The player had the ability to cycle through the lead units of battalions and control their movement and attacks, or the player had control to go it alone as a hero. It was a top-down, vaguely robotron-like experience. The players didn't just click with a mouse telling armies where to go and where to attack, they became the armies, moving around and attacking manually.

It was real time, split screen, turn based strategy game and it worked flawlessly. I've never seen anything like it in the era of contrived big commercial games, which is why I want to bring it back.

Now that I think about it, the whole thing seems a bit beyond me. But if things get too complex, I'll still have the collective mind of the tigersauce forums to help me along, right?

This is too good an idea to let it go to waste by being scared off by the complex-sounding phrase "neural network!"

Here are some additions that I figure the original game could use, that I think are within my humble realm of tech skillz:

1.) MAPS: the original started to feel boring after playing the same map over and over. I could make multiple maps, as well as a map editor.

2.) Network play: The addition of network play could allow for more than two players in a battle at a time, I think. Allies could merge armies, allowing for more than one player controlling things at each side.

The idea strikes me as nifty, though I don't know how it would play out as to who would have control over the movement of a merged army in the world map during the actual turn based sections...

Still, allies could merge armies temporarily to garrison each other's castles/countries.

That's all I got so far. Feel free to add onto the list. I played around with the idea of adding new units/heroes, but that ultimately seems unnecessary to me, plus I don't want to spend the rest of my natural life on this.

[Derek]

I don't want to throw a big bucket of ice water over your head or anything, but when you said you had a "simple understanding of GML and gamemaker," it kind of raised a red flag for me.  Reading more about this game, it seems like a remake would be a difficult undertaking for even an experienced team.

Even without the neural network (which is essentially an AI that learns from its mistakes in a fashion that is similar to how we think the human brain works), it seems like there would be some big design challenges... not to mention a whole boatload of graphics and sound assets involved.

And then you're talking about network play!

Before you get overboard with the design, it'd probably be a good idea to just throw together a simple battle system that has everything you liked about the original game.  My guess is that that itself will be more than an ample challenge for a single developer.

Or you can tell me to kindly fuck off and I wouldn't hold it against you. :D

[Coelacanth]

You're probably right.

I'll try to recreate the battle system, and report on how that goes. From memory the AI didn't seem all that complex. I'll scrap the idea of a "neural network" for something far simpler and predictable, since I have no idea what goes into that anyway.

Gotta start somewhere, right?

[Derek]

Neural networks can be absolute crap because they depend on a lot of ambiguous factors and by the time the network is seeded you really have no idea how it works.  You just kind of pray that things turned out alright (much like a human brain, actually).

I just thought it was interesting because as far as I know only a few games have ever tried using one for their AI.

[Zaphos]

Hmm, thought this was interesting -- a quote from Dyack (of Silicon Knights) about the neural net in Fantasy Empires:

Fantasy Empires had a learning neural network and an agent that would try to learn from what you were doing. It was part of my master's thesis. We did a lot of tests on it. It actually adapted, it helped people, and it actually was a great test case. We put it in Fantasy Empires, and it was a little bullet point on the back for marketing and no one noticed. No one cared.

( - Gamasutra Interview)

[Seth]

Hm, I wonder, even if most people didn't consciously notice it, if it really did enhance the game

[PoV]

Well, I think the game still relied on superior numbers to win battles.  So smart or not, if I had double your units, I'd win.  This game came in a pack of SSI games when I played it.  For me though, Stronghold and Dark Sun were the games of note from that pack.

[Chris Whitman]

Hmm, thought this was interesting -- a quote from Dyack (of Silicon Knights) about the neural net in Fantasy Empires:

When you think about it, how many people are going to take notice of an AI that simply makes better decisions, or learns or adapts?

I think it comes down to the fact that, by and large, bad artificial intelligence is extremely noticeable, while good artificial intelligence is largely unnoticeable because it makes human-like choices. We notice the incongruity of Dwarf Fortress dwarves picking up things that are on fire and carrying them home, but it won't necessarily stand out when a group of units in a war game behave in an organized fashion instead of milling about aimlessly.

I don't think that means good AI isn't a worthwhile feature, it's just often not the kind of thing people rant and rave about.

[Derek]

I don't think that means good AI isn't a worthwhile feature, it's just often not the kind of thing people rant and rave about.

In this particular case, I think bad marketing has to take most of the blame.  If you didn't know what a neural net was and saw it in a list of bullet points, you probably wouldn't give it a second thought.  It sounds like some made-up pseudo-science hogwash (and maybe it is, kinda).

But if the marketing team had decided to press the fact that the AI learned from its mistakes and could adapt the way a human brain does, I'm sure that would have been of interest to a lot of folks back in '93.

I do agree AI is usually more noticeable when it's bad than when it's good, though.

[Lurk]

The AI was'nt that impressive, maybe because it was one of those hybrid strategy games. You moved your armies in a turn-base phase, but the battles were played in a very flawed arcade sequence. You could very easily win with a ranged-attack hero, by killing the enemy while they were stuck in the castle, shooting through walls. Tactical map graphics and events were nice though, I don't get why they put in the top-down fights. Maybe they were planning Dark Legion, which has very similar gameplay, but much better combat graphics. I think those games tried very much to recapture the spirit of Archon. Those were the days

[nunix]

I really enjoyed FE and I'd be pretty happy with some kind of port, even; I'm pretty gosh-freaking sick of Tolkienesque (and thus D&D -- that's another argument, save it for another thread or PM) fantasy (I MEAN IT) but every now and then I play a game or two of it and enjoy it.

The fun stuff:

* An interesting world terrain, with region gold bonuses and unit recruiting (nabbing the Broken Lands got you both an easy three-region bonus + Night Elves and Orcs)
* Levelling up your player-god-hero (the avatar the player is ostensibly ruling via, though never seen, but who influences magic and other things)
* Keep and castle upgrades
* Taking your hero and blasting holes in the enemy's  keep walls, or sometimes just chopping them down with a freaking sword
* Greyhawk > Forgotten Realms. Suck it, fanboys. (See: first parentheses)

I'd honestly call the game something like Dynasty Warriors +.. I don't know, maybe the Total War series would be best. If it was made tech-generations ago in SVGA graphics, and a CD voice-over was a big deal.

To answer your question, would people play it? Sure. That's a ridiculous question, man. People will play anything. Will it have a huge audience? Hire a marketting team if you want to know that (and be prepared to have wasted the money). If it's a thing you like, and a thing you want to do, then just do it. People who like the game will play it and continue to play it. That's the only way anything good ever happens.