gratz on the engine and what exactly do you mean by the zombies will have less AI the farther away they are, does that mean they will be dumber if they are farther away?

Yes.
When they are far away, when you can't even see them, they will just be randomly moving around. No need to animate them, or even draw them. Just compute their new position. Or just make them idle.

When they are less far away you might want to have them walk a bit, pause, turnaround, walk some more, and react to the environment.

When they are close to you they have to detect you (line of sight) and move towards you to attack. Also detect more collisions. Maybe try to break through your barriers etc...


My point is, there is no need to actively run AI scripts on a zombie which is on the other side of town.

First off to the original poster. That is a FANTASTIC idea (the general concept anywaays), and a great and yes, fairly untapped segment I do believe. I think part of the reason for this may be that (I remember reading this in a gamestura report somewhere back in 2000-2002) that zombie horror games outside of REvil were performing badly in the marketplace. Of course that was then, and I dont believe anyone has tried it with a sandbox effect. And LOGY just really made it shine:

Quote: Original post by LOOY
I'd prefer it if the players didn't spawn in groups, but started in random houses, and would have to band together using some sort of radar. The zombie numbers should increasce as the game goes on so eventually players must hide in a larger building as you said.

Except I would remove the radar :). One of the fun things about zombie/horror games is you dont see things coming. Your afraid of opening that door because there could very well be some undead brain sucking mutant on the other side of it. Honestly this could be one of the best co-op style online FPS games out there if it were done up right. I mean take 10 people, stick them in a fairly decent sized city in their bedrooms armed with only a pistol, and set them loose. And respawns can be very painfull aswell, you know that if you lose a guy to a zombie, said person becomes a zombie, and his actual charactor responds in another bedroom or something.

I dont really like the "last man standing" gameplay. It could easily be exploited so it lasts forever, and it seems a tad boring. Another idea would be gathering living NPC's. Then you have groups of people hunting for NPC's still alive and evacuating them to particular points throughout the city. Elements of the game include finding people to band together with (the more people you find the more protected you are obviously), finding ammo and resources, and evacuating civilians out of the area.

Wow,

I've been kicking around an idea similar to this for a few years now. I've written a rough outline for the concept and continue to look for easiest way to bring it to life... I keep taking small steps along the path... which led me tonight to search different game development forums with thoughts of collaboration. I am not a programmer, more of a creative artist who games a lot and dabbles in mods.

I was excited about Left4dead, having done some mapping for the HL mod "Day of Defeat" and considering HL2's engine for a mod incorporating the concepts I was working on. I think left4dead has a lot of potential for being an exciting game, but it doesn't compete with what I was working on... I am not talking about a level of superiority, just a different genre and experience.

My concept isn't an MMO or even a multiplayer game though. As others have referenced in this thread, I didn't want to deal with the problem of player controlled zombies or the conversion of living>undead>dead>living conundrum.

I want to see a zombie survival RTS. It can have a multiplayer option, but players only control the living and once the last survivor dies, game over.

To my knowledge, there hasn't been a strong zombie RTS yet.

Imagine my surprise, finding this thread practically at the beginning of my search!

This is actually pretty cool. It reminds me of the zombie mod on counter strike source. That was a lot of hours of my life being a zombie...haha. In that one the zombies were much faster than you but when you shot them they'd bounce back a bit. You should look up videos for it to see how they handled any situations you can think of. I am working on getting kDevelop and SVN Workbench and MySQL set up on my debian and after I get used to all of that stuff I am going to dabble in SDL. I may use this idea a bit to learn game development better but I will try it out in a top down 2d style to simplify things for me.
Won't get very far very fast though since I work 10-12 hours a day 6-7 days a week ;)

Well ive got several people to respond to now, lets get started:

Response to Viperr:

Alright so i did understand you correctly, although how far would be to far to make the zomb's unresponsive to you?

Response to PaulCeser:

Well im glad you see how there is a need for a game like this just like i do :), and yeah we probably should get rid of a radar, but then that raises a problem, if the people dont have a radar or a means of communication how will they find each other? People will spend more time looking for each other than actually killing zombies is what i think will happen. My answer to this is if its on X-box 360 their should be a map, and then players can go into their friends list on 360 and tell each other where to meet up. Tell me what you think about that idea.

Response to ak_satchel:

Alright now then i know you said your not really a programmer but how far have you come on taking your "small steps"? Do you do concept art? Is that what you mean by "small steps"?
Also im not sure whether or not i am really going to be excited about L4D, the general concept, you and some buddies kill zomb's, is good, but the downside to it is only 4 buddies?, people playing as special "boss" zombies with special abilities? to me this just doesnt sound like a fantastic game, but dont get me wrong it does sound good.
Next up is How do you make a zombie game an RTS? Are you going to be looking at a map with little dots that are zombies and you tell them where to go? How will barricades play into this? wouldnt it be more fun to take direct control of a human? If you made a zombie game im almost 100% positive people will be dissapointed if its not a MMO or a FPS.

Response to Cryovenom:

Yeah i playey zombie mods on counter strike source still today, and i dont know about having a bounce back in this game, i just think we will do one head shot and they drop, all though you will have so little ammo and so many zombies that there will still be a major problem.

Well i replied to everyones post's, keep posting guys :)

Here is a general idea of the zombie rts I would like to create:

The player chooses a class to start with, just like choosing a race or side in traditional rts, except you don't start with a colony/build unit and a few grunts- you just control one character. Your class selection determines where you start on the map. Each class of course will have different strengths and weaknesses including the strengths and weakness of the corresponding starting area.

The player will have a few minutes to become familiar with surroundings - Zombie concentration starts slow and builds over time.

The player doesn't "make" new units at all, all resources the player will use to win the map/game will be scavenged or rescued. For example, the player will have to rescue living npc's throughout the map in order to control more units(numbers of living survivors will diminish as the game progresses). Rescuing more survivors can be accomplished by proximity - I have a few different ideas about this; either the proximity can be static, proximity can vary depending on unit class (a leadership stat) or the unit's class will need to have a minimum value (leadership stat) to "unlock" the npc and then be able to control them. This resource model will require the player to be aggressive in exploring and rescuing as many survivors as he/she can in order to have a higher chance of success beating the game. Survivors are placed randomly (for a demo they can be fixed) throughout the map with different, hidden countdowns associated with their lifespan.

Each survivor will be a different class, also prompting the player to explore and rescue... for example: the player will be motivated to explore medical facilities to find a medical class if they don't have one already.

As stated, the amount of zombies will start off slow at the beginning, but gradually increase. There can be a fixed number of zombies on the map with the total potential number dependent on the amount of survivors, or they can be constant (spawning along the perimeter or outside the cordoned area)-It depends on the game engine and other variables.

I would like to see "fresh" zombies move faster than old ones, creating more tension for the player if one of the player controlled survivors gets eaten.

Each class can have an ability tree that gets unlocked as more items are looted from buildings and corpses.

Barricades can be built (class dependent).
Depending on the engine, it would be nice to have player controlled vehicles.

Multiplayer option could include players competing for survivors throughout the map - creating different factions.

Maps can be linked in a linear fashion with a goal in mind (reaching the airport , army base, or harbor) so the player will start in the suburbs on map one, map two can be liberating a mall, map three can be a vehicle map where the player drives the survivors to an army base, etc... this can lead to plot twists... the player might even need to escape a zone before a nuclear strike for example.

Get the idea? I agree that there is a lot of potential for a zombie horror game that's untapped. I think an RTS with sandbox features gives the player the most options for fun gameplay associated with the concept, (FPS are limited on scope and linear maps, mmo's are too player dependent and have balance challenges)

That was just a general summary of different concepts referenced in my development outline, off the top of my head. I can talk about the concept all day long - pretty excited about seeing something like it come to life :)

I am still looking for a decent game engine to get a basic demo going... even in 2D. If anyone has any suggestions for an easily mod-able rts engine, that would be great! I've considered Company of Heroes, Spring, and currently looking at the Torque engine with the RTS starter pack. I don't care about physics or features like volumetric fog... an engine that supports real time lighting would be nice (for day and night sequences)




Sorry Blarkage, I realized this post is piggy-backing your's and I should probably start a new thread since I am looking for collaboration. However, I think it's great that you were thinking that the genre was underrepresented too...

Quote: Original post by Blarkage
Well ive got several people to respond to now, lets get started:

Response to Viperr:

Alright so i did understand you correctly, although how far would be to far to make the zomb's unresponsive to you?

Response to PaulCeser:

Well im glad you see how there is a need for a game like this just like i do :), and yeah we probably should get rid of a radar, but then that raises a problem, if the people dont have a radar or a means of communication how will they find each other? People will spend more time looking for each other than actually killing zombies is what i think will happen. My answer to this is if its on X-box 360 their should be a map, and then players can go into their friends list on 360 and tell each other where to meet up. Tell me what you think about that idea.

Maybe the distant zombies would follow active zombies. Active zombies would be the ones that are within close proximity to a person, OR they would be ones that have decided to follow active zombies. So in this case, the close ones would find you and the far ones would follow each other in chains which could simulate them finding you like in the movies where they all just mass in to one area that the victims have decided to use as a fort.

As for communication, why not have positional sounds based on the person's location. If they speak in one spot and someone is close enough to hear, they can hear you. So if you're in desperate need of help and ammo, run around town quickly and scream! It'll add to the atmosphere and if you are being annoyed by some 8 year old yelling in to the mic, just run out in to the zombies gun-blazing hoping you can find another place you can call home :)

Lol cryovenom that whole thing about how loud something is determines how far you can hear them from, but i love the way you put it with the 8 year old :)

and also thats a good idea with the zombies in a chain, i think thats what we should do for the game.

Quote: Original post by Blarkage
Lol cryovenom that whole thing about how loud something is determines how far you can hear them from...

Well, in games like counterstrike, when someone speaks it's global. I thought it would be cool to just run by someone laughing like peter from family guy and then fading off in to the distance.

Right now I am thinking about what it would be like to implement a voice recognition system on to a separate server so when everyone is playing, it would take what they say and display it in a chat dialog. Of course it's never accurate but the idea sounds fun.