The one thing that has always annoyed me about current RTS game models is the absolute lack customizability. C@C, Warcraft, Cossacks, etc etc. They all have one thing in common and that would be a static tree of units that almost never changes. Sure all time passes in some games you can get new units and upgrades but it doesn't measure up to what I am looking for. Imagine this- Instead of having a game where units are based on the age or technology, the units should instead be created by the player. Impossible Creatures did this on a very fundamental level but it's lack of variation resulted in a finite number of useful combinations for combat. Here is a fundamental example of my concept. Suppose you had a RTS game that ranged from early time up until say the Age of Reason. Players would be given an entire research tree each with a different class of weapon or armor and then subtrees of various kinds of weapon and armor. The player would get a certain number of research points per age developement that they would use to create thier own tech tree. Like this Type Weapon Armor Ranged Sword Shield Body Bow-Longbow-X Bow Katana-Longsword-Scimitar Mech X Bow-etc etc Claymore-Rapier-etc etc Each weapon type gives a smaller tree of specific weapons or armor or special equipment to choose from. Then you have a unit creation screen which displays a normal human in clothing with base stats. Then you equip specific weapons to each hand, armor, and special items. Each peice of equipment affects stats such as speed, armor, defence, damage dealt, attack skill, HP. They player names the unit and it's added to the player's unit tree. Give them special training techniques to the tech tree to further modify the base statistics of a particular unit. This is just a basic example of the concept. Why stop there? What is to prevent you from allowing players to run from the stone age to the modern day. Provide vehicle, air, and ship design schemes with additional weapon and equipment trees to research each age. Research mining techniques or lumber cutting to find new types of minerals or wood to construct items with. Research forging or construction techniques combined with new alloys and minerals to result in thousands of combinations for weapons and equipment for players to play with. The problem with RTS games in particular is that developers tend to focus more on graphics and less on variation and replayability. A player is more likely to put money into a game and it's expansions if he can replay the same game with a new experiance each time. The possiblities for a game with this much player control is limitless in it's potential.

While this opens the possibilities to new types of games for your users the majority of them will be unbalanced and not fun. The hardest part in any RTS game is getting it balanced in such a way that no one unit is more valuable than the rest. As soon as you open up the door to changing unit values then you open the door to unbalancing everything.

While I like the idea of opening things up to the user to allow for replayability you really have to come up with a way to calculate the balance of units and let the users realize what they are doing.

Have you ever heard of WarZone 2100? It does not go though the ages like your proposed, but it does pretty much everything else you have said. Freaking awesome game, but the way they did resource gathering killed it. I'd like to see a remake of it.

I've Thought about that sort of idea before.

The main issue becomes a balancing act. Anything that is worth researching gets
researched, and anything that isn't is avoided.

If you make a hard-counter set of research topics, then you run into some of the
problems of games like homeworld, where in your research tree can
become a futile waste of resources when the other player builds specifically to counter you. Since it becomes to hard to backtrack and start a new branch on the tree. Games can become frustraiting if you don't know how to rush a topic faster
than they can counter it, but there is plenty of variety.

If you make a soft-counter set of topics, then you run into the Diablo2/WarCraft3
problem of players heading for the most damaging attack that suits their style of
gameplay. It works out well for letting players play the way they want, but there
end up being a limited number of stratigies out there.

So balancing for the type of game you want becomes very very hard if you start to let the tree branch.
--Anyway, just some ramblings for you to think about.

The balancing becomes up to the player. It's akin to a good RPG. For example, a friend of mine is obsessed with making characters that are completely opposite what they are supposed to be. He likes to make speedy magic dwarves in Arcanum and hulking technological Elves.

Obviously if you give the player enough tools he will constantly be creating new tech trees to suit his or her opponent. This is where the fun and challenge is. If you give the player enough resources the challenge becomes outwitting your opponent by making a better army than he does. Having certain combinations of units will always be better than others at certain tasks. We all know the average player exists to rush in massed formations of the strongest unit possible. So the challenge is not in balancing the units, the challenge is making other combinations of units viable for other tasks. We need to make it so that tasks such as raiding trade routes, scouting, attacking resource points, guarding, patrolling, lobbing artillery, are all viable options. The game would have to be designed so that Mass formations of the strongest unit possible is not really the best way to win.
The current mindset with game developers is balancing all the units. Units should not be balanced. The play styles need to be balanced not the units. If you balance the possible viable options and ways of waging a war, then all the units begin to have a purpose other than rushing in and killing everything.

Quote:

Original post by Cian OConnor The balancing becomes up to the player.

I don't think you understand what he means by balance. The problem is that you - the designer - have to make sure that there are no combinations of components that result in an undefeatable super unit. Otherwise, players will eventually discover that combination and use it exclusively, because it's the most powerful unit they can make. And as a result, your game will have even less variation than more conventional RTS games where you don't get design your own units.

This type of balancing is hard enough in a conventional RTS with fixed unit types. Starcraft is considered one of the best balanced RTS games out there, but that balance didn't exactly happen overnight. Rigorous testing and tweaking throughout the development, and continued tweaking for years after the release, in various patches. And that's for a game with ~40 different unit types. With say, five armour options, five weapon options and five base unit options, you're already looking at 125 possible units.

Another problem with the 'design your own' approach to units, is that designing units takes time away from the game. Isn't there enough micromanagement in RTS games already, without the need for players to spend time designing his units?

Quote:

The problem with RTS games in particular is that developers tend to focus more on graphics and less on variation and replayability. A player is more likely to put money into a game and it's expansions if he can replay the same game with a new experiance each time. The possiblities for a game with this much player control is limitless in it's potential.

I disagree. Again, look at Starcraft for an example - despite being ancient and having simple graphics by today's standards (2D, 8bit colour!) it is still one of the most widely played RTS games around. Log on to battle.net and you'll see that there are still thousands of players on any given server. There are only three different races (but they are very different) and they only have around twelve different unit types each. (rough estimate, haven't counted them up properly)

RTS games are most fun when played against human opponents. This is where the majority of the variety and replayability comes from - trying to outthink your opponent.

I have to disagree yet again. There may be certain combinations that result in a super unit for combat sure. Thats why you must make other many styles of warfare. Suppose trade is vital for a nation. Slow pounding units can niether escort nor attack quick trade units. Therefore you must have a raiding unit. Make it so that fixed fortifications can only be beated by seige units. This defeats the massive rush tactics. Perhaps having generals in the game who when assassinated reduce the power of an army by 1/4. In Cossacks, raiding was very viable as one unit could take over or destroy any mine that wasn't protected. The number of mines and size of the maps made it had to raid or protect mines without quick cavalry units. Such units wouldn't stand up against any infantry.

Yes I understand the weapons must be balanced so that no one super unit can be created. But there should be more focus towards balancing the ways of waging a war to make weaken unit combinations just as viable or effective.

As for micromanagement. If you don't want to micromanage in game, then why not out of game? Who is to say you can't presave the techtree path you want to follow. That way when you start the game, the tech tree you would normally follow is highlighted for easy research progression until you decide to deviate from it. Then it becomes much quicker to simply select what you want and go. Same thing with unit creation. Allow precreated units that, much like a normal RTS, will be greyed out until the proper research is conducted. By pre created I mean allow a unit creation tool that can be used out of game and saved as an army profile before the game starts. One that can be changed during the game if needed.

Quote:

Original post by Cian OConnor I have to disagree yet again. There may be certain combinations that result in a super unit for combat sure. Thats why you must make other many styles of warfare. Suppose trade is vital for a nation. Slow pounding units can niether escort nor attack quick trade units. Therefore you must have a raiding unit. Make it so that fixed fortifications can only be beated by seige units. This defeats the massive rush tactics.

Sounds simple enough in principle, but be wary - actually putting it into practice is another thing entirely. What if one of the early raiding units is sufficiently powerful to crush the enemy before he even gets to fortifications? What if it's possible to build a generalist unit that can do everything? What if I can build an indestructible seige unit - that might not be able to destroy trade routes, but can destroy everything else, thus making the trade routes irrelevent anyway?

It's a simple rule of thumb, and it applies to all software: the more options you give to the user, the more ways your program has to go wrong. This is particularly true of RTS games, because it applies not just to the stability of the actual program, but to the complex interplay and balance of all the units.

I'm not saying it's a bad idea, but I am saying it's a damned hard one to get right. It's a hard enough job to balance 40 units well, 400 possible combinations is asking for trouble. 40 well balanced units makes for a better game than 400 poorly balanced ones.

Quote:

As for micromanagement. If you don't want to micromanage in game, then why not out of game? Who is to say you can't presave the techtree path you want to follow. That way when you start the game, the tech tree you would normally follow is highlighted for easy research progression until you decide to deviate from it. Then it becomes much quicker to simply select what you want and go. Same thing with unit creation. Allow precreated units that, much like a normal RTS, will be greyed out until the proper research is conducted. By pre created I mean allow a unit creation tool that can be used out of game and saved as an army profile before the game starts. One that can be changed during the game if needed.

I definitely think this would be the way to go, a bit more like the conventional RTS but with a 'design your own race' option. Could definitely be interesting, if you can solve the astronomical balance problems already mentioned.

Thats where I believe the strategy comes in play. Modifying your army to counter whatever the opponent has to offer. Balancing the weapons will be an important factor in the game to be sure. But once you balance and tweak the weapons, the tool then becomes to players to decide which combinations work best to counter what. Once it's all figured out, it's easy to see just how much replayability a game will have. Especially if you come out with expansions every year with a new age, new weapons and research trees, new architectual styles to choose from. The possibilities are limitless because the add-on's would be less coding intensive than most expansions are.

Quote:

Original post by Cian OConnor The one thing that has always annoyed me about current RTS game models is the absolute lack customizability. C@C, Warcraft, Cossacks, etc etc. They all have one thing in common and that would be a static tree of units that almost never changes. Sure all time passes in some games you can get new units and upgrades but it doesn't measure up to what I am looking for.

The RTS games have exactly the kind of ability to customize you're talking about, but it's done on one level higher that you think of. While you want to build units out of pre-made parts, the RTS games allow you to build varied entities (armies) out of pre-made units.

This might seem rigid and limiting at first glance, but when you think about it more, the ability to customize your base units instead of them being pre-made doesn't add really that much to the game, if anything (aside maybe from visual aspect) What i mean is... if the game gives ability to customize say, speed of your unit on 100-point scale, together with armour and firepower, and have these three attributes linked together to not allow players make units with everything maxed ... i'd dare a guess that due to players' tendency to min-max, majority of designs made by them will have the attributes set to various combinations of min, 25%, 33%, 50%, 66%, 75% and max ... simply because anything less quantized is going to make too tiny difference on the battlefield to matter, and having units with clear-cut combination of attributes makes it easier to decide what roles they should fill in battle ... as opposed to large number of units with attributes different by just few points.

And the RTS games provide you with exactly that, pre-made units with attributes quantized enough to perform with noticeable differences and easy to pigeonhole into specific roles (tank, scout, escort, transport etc) The challenge they give is to put these units together into well-oiled coherent war machine that can smash the other guy's attempt at doing the same. Being able to micro-manage a tiny part of that army so you can replace a sword on one unit and make it do 75 damage instead of 74, at the expense of armour going from 33 to 32... it seems to obfuscate the game rather than add to it.

edit: btw, check out Mega Lo Mania ( http://www.the-underdogs.org/game.php?gameid=695 ) for a very simplified version of what you want -- the game allows continuous research of weapon and armour technology (as well as buildings and vehicles) over the period of whole human history ... then you can use these researched weapons and gear to put together units that can be sent on the battlefield.