Welcome to the Home of Game Strategies and Stories
14 May 04
Friday again. It's highlight time. Since I went on about AI, I figured I needed a game that took some chances with AI for the highlight. I've also been playing it pretty safe on the highlights. For games with little downside, it's easy to get people to agree on highlights. So, I'm going to take a chance. We're going to look at the highlights of Black and White (BW). Yes, we all know there was downside to BW. It was discussed at length across the internet. Still, there are good reasons why the rush of initial positive reviews flowed in.
Obviously, the biggest highlight in BW is your creature. The creature is a reward for finishing one of the very early quests. Since BW is a 'god' game, it's nice to have an active avatar in the game. Of course, in BW you don't directly control your creature. It's an 'exceptional' animal. It's huge for a start. It can grow bigger and more powerful. It can fight other creatures. But the best thing it does is learn. And boy does it. Pretty much anything you do around the creature, it will try to copy you. You can then reward it (petting it) or punish it (slapping it) or just ignore it.
Early on, you have to keep a close eye on your creature. It can and will try everything. You have to be around to make sure it understands the things you want it to do and those you don't. Your creature can be a huge help in the game if trained well. Your creature doesn't want to be good or bad. It wants to please you. Its appearance is a constant reminder of whether you've been a good or dark 'god.'
I could go on and on about the AI of the creature. I'll limit myself to say that if more games used AI of that caliber, I wouldn't have written a series of articles complaining about AI. I can also add that if you play a good 'deity' that you can grow quite fond of your creature (and maybe a little proud). It's impressive when unscripted events can evoke an emotional response.
The next obvious highlight is the graphical and game engine of BW. While strategy gamers are often proud that they don't need the latest eye candy, this engine is still impressive after a couple of years. It's also useful. You can zoom way in to see the day to day lives of your villagers, or you can zoom way out to look at the whole empire and territory. The engine is powerful enough to create the sense of a living breathing world for you and your creature to play in. BW also sports realistic physics so you can throw boulders or even villagers and see the expected results.
Moving on, there's the mouse control system. While some would argue the design turned out to be a weakness, I've got to admire the risk taking involved in making such a complex game that could be controlled with only a mouse. From movement, to camera, to spell casting, to creature training you could do it all with just the mouse. There's also key binding if you don't like the mouse only method.
Finally, I would point out the way your actions impact the world as a highlight. On a basic level, if you want more people, set up more couples to have children. If you need a boulder out of the way, pick it up and move it. If you want more crops for your people, create rain for them. If you want more magical power, set more people to worship you. Beyond that even, if you are good and kind, the world becomes brighter with more plants and animals. If you are cruel and evil, the world becomes darker and more barren. The same changes can be seen in your creature. A kind, caring creature is bright and happy and intelligent looking. A cold, cruel creature is dark and menacing. This kind of subtle feedback adds greatly to the game. Even better the changes are slow and gradual so you might not be thinking about how evil you've become until you look around the stormy, blighted world. That really adds to immersion.
As we go to today's game, I struck by the fact that development must be cheaper in Russia. Otherwise, how could 1C have so many games in its publishing pipe. Today's installment is XIII Century: Sword & Honor (official site) from OSW Games. It's a 3D RTS with a strategic board ala the Total War series. Playable factions include the British, French, Spanish, Germans, and (of course) the Russians. Hopefully more details will be available as it nears release in the second half of 2005.
Jason
Comments?
13 May 04
There are massive lightning storms in the Houston area today, so I'm going to keep this short. E3 has started. The early impression is that the handhelds are going to dominate the show. While generally weak on the strategy and tactical front, there is some hope.
KOEI has announced that they will support both the Sony PSP and the Nintendo DS. Right now the only officially announced title is going to be a version of Dynasty Warriors for both systems. However, KOEI has confirmed a strategy title is in the works for the PSP.
There are other hopes as well. Konami has confirmed a new Metal Gear game for the PSP probably for the launch. That should please the tactical stealth crowd.
Pictures of both systems have shown up on the net. I have to say that both look better from a design point of view than anything on the market now. Still, the PSP looks killer. If Sony can keep the price point down, it looks like a killer. With high end games, a proprietary movie format (that includes digital rights protection), and a sleek modern look surrounding the 16x9 screen, it could turn into the next ipod like must have device.
Today's game is the next in our Russian series. It's Brigade E5: New Jagged Union (official site) from Apeiron. Gee, that title sound familiar. Wonder what the gameplay is about? Well, you lead a team of mercenaries dealing with coup on a tropical island. It's real time until combat then goes turn based (or really some sort of phased timed actions). Hmm, well, if you're going to take ideas, you might as well take from the best. It's due out in the last quarter of the year.
Jason
Comments?
12 May 04
I'm going to try to wrap up my thoughts on Artificial Intelligence today. I'm not a huge sports game fan. I'd probably save a lot of money if I did. Have you ever watched the Madden Challenge? Those people play one game constantly all season. Wow. Of course, sports games focus on a real life activity. That means that little of the design has to go into the game world or the rules that apply. The designers can focus on making a deep, rich and realistic game. Then each year, they release, get feedback and start the process again. While that constant pressure can lead to some stumbles, overall the games get better and better each year.
Well, those games are modeling the thinking of real people (coaches and players). Sports fans are sticklers for realistic outcomes so how do sports games do it? When they get the overall AI in good condition, the answer is simple. They use sliders. Sliders are exactly what they sound like. They are adjustable values that the user can set to get more realistic play (or more action oriented if that's what they desire). Generally, the more sliders the game offers, the happier the sports fans are. Some people have even built an online reputation for coming up with good, realistic slider settings for games.
So why don't we see more of this in strategy games? I think part of the problem is how AI has traditionally been handled and part of the problem is that designers don't often think about it. Any good strategy game should be like a sports game in that there are many paths to victory. Some may be hard or easier, but there's no one pure way to win. Sounds like there's room for sliders to me. While the initial underlying AI may be harder to program, the designers would be giving more gamers the game they want.
That's really the point isn't it. Many cool games get shot down in reviews or your friends because although the unit design was good and the balance was good, the game didn't play the way they wanted it to. So what kind of sliders should we be looking for? Aggressiveness would be a good start. The slider wouldn't affect the quality of play at all. Put it toward defensive and the AI will focus on defensive units and structures. It would focus on expanding to defensible areas. Put the slider in the middle and the AI would balance protecting its position with creating an offensive army. Slam the slider to aggressive and the AI will focus on creating skirmishing forces to harass you while building a massive army to crush you. If you manage to get behind his lines however, he'll be in trouble.
Similar things could be done with other sliders. You could even have linked sliders. For instance, say for a fantasy RTS, you could have linked archery, infantry, cavalry and siege sliders. If you crank up one, the others go down. Some others could be specialized units vs. mass numbers of units. Unit sacrifice tolerance could be another. Maybe army size vs. tech level. Maybe you could even have linked sliders for every unit on a side. As many different ways as people can come up with to play, you could have sliders.
So what sliders would you have?
Today we have yet another Russian game for you. It's Private Wars (official site) from TS Group. It's a tactical shooter with squad commands. It's promising visuals to rival Doom 3. We'll have to see if the tactical aspect will stand up to Ghost Recon 2 or Operation Flashpoint 2. Still we can't complain about another opportunity to virtually fight the war on terror. No release date has been set yet, so we'll guess you shouldn't look for it until next year.
Jason
Comments?
11 May 04
Continuing with our AI theme, today we're going over the dangers of poor AI in the RTS genre. While turn based strategy games are often impressive with their AI, RTS's tend to throw the old more, bigger, faster and cheaper at you. It's pretty rare to see the AI in a RTS game impress you.
Now we're at a point where many have called the death of the genre immanent. If nothing changes, I might agree. I tend to believe, however, that there are many areas of improvement that can not only save the RTS genre, but also revitalize it. Improving the AI is the best area.
In a RTS game, there are generally two types of AI. There's the individual unit AI and the overall enemy strategy AI. Unit AI has seen some progress in recent games. Now it's not shocking to see infantry march ahead of archers. It is still surprising to see archers move to cover and high ground, but we'll get there. After the early years of the genre, most RTSers are happy that their units can move from A to B without getting stuck on a tree.
Let's focus on the overall AI for now. Most games still just throw units at you. Admittedly some now actually build a group before throwing them at you, but some still string lone units at your defensive structures. If you lose to the AI it's not because of some brilliant strategy it used. It just out built you or out teched you. While efficiency is an easy task for a computer to do, it doesn't make an interesting enemy. Ideally, you'd like to face your Napoleon, an opponent who can win with less.
There are two problems with a merely efficient AI. The first is the tipping point. You build and skirmish to a point. You win your first major battle and take resources away from the enemy. Too often the game is over. It's just a matter of mopping up the left overs. If the AI even tries to take something back, it often just goes after what it lost right where your army is sitting. We need AI that will send probing forces to find your weak points and seize those resources.
The second problem is the exploit. The problem with using an efficient AI is that it still has to work. It wouldn't be any use to have an efficient AI that always lost handily. It wouldn't be a fun game. If efficiency does work, it opens the door for human players to use similar shortcuts. It's not very strategic for players to mindlessly follow hyperefficient build orders and produce the same army every time and just throw it at the opponent. First the gamespace needs to be configured to reward sound strategy. Then we need AI that uses strategy and not efficiency. Hopefully the new development teams are moving this direction and will help save the genre.
Sticking with our Russian theme, today's game is The Cuban Missile Crisis (official site) from G5 Software. Cuban Missile crisis combines RTS gameplay with a turn based global tactical mode. It answers the question: What if diplomacy and the blockade had failed? Look for the doomsday clock to strike at the end of the year.
Jason
Comments?
10 May 04
Today I'd like to discuss Artificial Intelligence again. In case you can't tell, it's one of my favorite subjects. Like I mentioned before, I like to think about since the industry seems to avoid the subject like the plague. Ok, so that's not true of everyone. It's not hard to understand the problem though. Programming involves a lot of binary thinking, zero and one. Realistic AI tends to involve fuzzy thinking, all those unpleasant values between zero and one.
For instance, some current AI involves "state switching." A guard is mildly alert in normal mode. Something unusual enters his "sense" zone (sound, sight, or even just movement). The AI changes to "alert." He might start moving toward the disturbance or reporting in an alert or even going to look for some help. Once he detects something wrong (say a body), he goes to "alarm" mode. His weapon is up, he's sounding the alarm. He might attack or go for cover or run away. Programmers can use random numbers to select certain scripts to run and weight them to give a more "realistic" performance.
The overall effect is good, but the state switch itself is as jarring as the sudden appearance of an exclamation point over a guard's head in Metal Gear Solid. It takes the player's sense of immersion away. You are playing a game, but it's something you want to get lost in.
My theory is that the reason you have the sudden hits of Battlefield 1942 and City of Heroes is that these are worlds you can lose yourself in. But you also have allies (and/or opponents) that don't act in predictable ways. When a game can surprise you not once but over and over again, that's when you get real fun.
Tomorrow, I want to go over the particular dangers of scripting and limited AI in strategy games especially real time strategy. Perhaps, that will point us on a direction away from the "death of the RTS."
Today's game is another Russian import. You've got to love the energy and creativity being released over there even if you don't love all the games. It's Starwolves (official site) from X-bow Software. It's a turn based 3D tactical spaceship combat game with some RPG elements. You play a bounty hunter in the wilds of space. It's trying to be a strategy version of Privateer. We'll see how it succeeds later this year.
Jason
Comments?
© Talk Strategy 2004
