My current game development project is as a volunteer programmers for the Glest Advanced Engine, which is one of the 2 currently active Glest development forks. Glest is an open source cross platform RTS game.
GAE is both a game, using the original Glest tech tree, and an engine for creating RTS games. It has a decent modding community and several campaigns/new factions that modders have developed.
The original Glest website, from which you can find both Glest and GAE, and also Mega-Glest, is here:
http://www.glest.org

The feature I am currently adding to GAE is a RPG like Hero system. Heroes can either reset after every fight, or have their current stats stored in a save file. The second option is to allow modders and game creators to create campaigns with various units that exist beyond a single fight.

Due to the fact that GAE doesn't really have support for RPG style stats and skills, as an RTS engine the closest it has is unit wide upgrades and a rudimentary unit specific "level" system based on how many other units a unit has killed, I am mostly going to be writing this system without dealing with existing game code limitations. I will be referencing the "stats" of a unit and that is pretty much it. This is quite different from my previous work where I was mostly expanding on existing systems, like allowing layered upgrades or adding on hit attack effects.

Given this freedom I am currently trying to decide the specifics of my implementation.

For instance I can structure the overall system in various ways:
Classic level gain based on exp which provides "points" to raise stats or gain new skills or increase the power of old skills
Gaining skill points or stat points directly from experience
Spending experience itself to increase stats

Spending experience seems to allow the most freedom. You can specify the cost of 1 point in a stat, rather than having to say a stat or skill point is worth 5 movement speed or 1 attack strength or 2 armor or w/e.

Further I could use multiple types of experience, for instance fighting enemy units gives combat experience, which can be used to increase combat stats, whereas units that do economic tasks such as gathering resources could get production experience towards carrying larger loads per trip, or gathering faster. You could even have science experience whereby units that work with research/upgrades could become more proficient at those tasks.

However some players may prefer simpler mechanics, and generally those people constitute a larger number, so modders who use GAE may prefer to use that. I have been considering implementing multiple systems but that is a lot more work for me.

I was wondering if anyone here had any thoughts on these issues.
What is your preferred growth system in RTS games, if any?
Do you support leveling in separate spheres?
How do you prefer skill systems?
Diablo II style?
Warcraft style?
General thoughts on heroes/growth systems?

Do you mod games?
If so:
As a modder do you prefer an engine/game to have a lot of choices?
Do you like to pull aspects for several different genres?
Are you interested in Scenarios or random maps?
Do you like a lot of economic game play in RTS games or do you save that for TBS games?

What sort of problems have you encountered in trying to mod games?

GAE allows access to the source code, do you like to be able to mod aspects of source that constrain your options?

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]What is your preferred growth system in RTS games, if any?[/background]

[/font]
Heroes should grow via neutral objectives or enemy deaths (such as in wc3). This design encourages player interaction. The growth can be in the form of experience points items, or straight up stat increases. It is also important to restrict snowballing of heroes, so make sure that there is a punishment for letting a hero die, scaling with the strength of the hero.

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Do you support leveling in separate spheres?[/background]

[/font]
I'm not quite sure what this means?

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]How do you prefer skill systems?[/background]

[/font]
I prefer skills to be either predetermined or easily selected.

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]General thoughts on heroes/growth systems?[/background]

[/font]
[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Overall, hero complexity in an RTS should be kept to a minimum, because an RTS is largely about macromanagement, multitasking, and strategy. Forcing a player to spend lots of time on setting up a hero will make the game less fun, I feel. Part of the reason I personally did not like WC3 was because of the hero system. It was always optimal to get an early hero, because everything that you kill without a hero present is wasted EXP. It was impossible to scout most hero abilities, so you don't know what ability their hero has selected until you are fighting against it. For example, it is nearly impossible to know that an orc player is going for BM harrass. You can only make an educated guess. [/background]

[/font][color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Generally, heroes are very difficult to deal with without a hero of your own, which makes a hero-less build not viable. [/background]

[/font][color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Heroes in SC2 campaign are implemented in a much more desirable manner (except for Tosh, but he's a special case). They are largely low-micromanagement, but they feel good to use. They are strong, but mortal. I feel like if there was some way to produce heroes in game, as long as they were balanced between races, they would fare fairly well.[/background]

[/font]

[background=rgb(250, 251, 252)]What is your preferred growth system in RTS games, if any?[/background]

Heroes should grow via neutral objectives or enemy deaths (such as in wc3). This design encourages player interaction. The growth can be in the form of experience points items, or straight up stat increases. It is also important to restrict snowballing of heroes, so make sure that there is a punishment for letting a hero die, scaling with the strength of the hero.

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Do you support leveling in separate spheres?[/background][/font]

I'm not quite sure what this means?

[color=#ff0000]Economic, Military, Scientific

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]How do you prefer skill systems?[/background][/font]

I prefer skills to be either predetermined or easily selected.

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]General thoughts on heroes/growth systems?[/background][/font]

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Overall, hero complexity in an RTS should be kept to a minimum, because an RTS is largely about macromanagement, multitasking, and strategy. Forcing a player to spend lots of time on setting up a hero will make the game less fun, I feel. Part of the reason I personally did not like WC3 was because of the hero system. It was always optimal to get an early hero, because everything that you kill without a hero present is wasted EXP. It was impossible to scout most hero abilities, so you don't know what ability their hero has selected until you are fighting against it. For example, it is nearly impossible to know that an orc player is going for BM harrass. You can only make an educated guess. [/background][/font]

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Generally, heroes are very difficult to deal with without a hero of your own, which makes a hero-less build not viable. [/background][/font]

[color=#282828][font=helvetica, arial, verdana, tahoma, sans-serif]

[background=rgb(250, 251, 252)]Heroes in SC2 campaign are implemented in a much more desirable manner (except for Tosh, but he's a special case). They are largely low-micromanagement, but they feel good to use. They are strong, but mortal. I feel like if there was some way to produce heroes in game, as long as they were balanced between races, they would fare fairly well.[/background][/font]

I totally disagree that RTS games are about macro management, although I would prefer it if they were. Most popular RTS games are micro APM fests.

Its turn based strategy games that are about macro, like Dominions.

Majesty actually was mostly about macro but its not a traditional RTS.

If this RTS is using large armies then managing individual units would be a lot ot handle so using a automatic leveling system based on experience would be more appropriate but a smaller scale squad-based game would allow micromanagement easier. In that case, I would personally prefer to be able to spend experience points on upgrading my units. I am also a fan of eliminating a level/tier based system in favor of something more similar to a sports game's player rating where every individual point used has an affect, however miniscule it may be. With an RTS, it would be best to keep the unit specific attributes at a minimum (speed,power,accuracy,etc.) maybe 5-6 at the most.

I think that the spending of stat points should be streamlined.

I totally disagree that RTS games are about macro management, although I would prefer it if they were. Most popular RTS games are micro APM fests.

Its turn based strategy games that are about macro, like Dominions.

Majesty actually was mostly about macro but its not a traditional RTS.

SC2, unarguably the most popular RTS, is very macro intensive. A large portion of APM goes to macro. Micro is only important and APM heavy in specific situations such as marines vs banelings, zerglings vs hellions, etc.

However, if you mean to say that popular RTS nowadays are not about macroscopic decision making and strategy, then I can agree with you. However, non-micro RTS games I have played suffer from the issue of poor AI, making it extremely frustrating for players.

Also, no matter how non-micro you TRY to make the game, the element of unit pulling will always exist in any RTS. (Pulling a low hp unit out of range of the enemy, then sending it back in).

EDIT: Also, leveling in separate spheres is too complex. A better implementation of the same concept would be to just create a hero that specializes in each sphere, so you can make the decision when you recruit/train the hero. The total war series (although not truly RTS) does this, and it works brilliantly.

[quote name='AltarofScience' timestamp='1337891348' post='4943004']
I totally disagree that RTS games are about macro management, although I would prefer it if they were. Most popular RTS games are micro APM fests.

Its turn based strategy games that are about macro, like Dominions.

Majesty actually was mostly about macro but its not a traditional RTS.

SC2, unarguably the most popular RTS, is very macro intensive. A large portion of APM goes to macro. Micro is only important and APM heavy in specific situations such as marines vs banelings, zerglings vs hellions, etc.

[color=#ff0000]This was actually an argument I have seen on some RTS specific forums. Many of us don't agree with the starcraft definition of macro. Apparently among starcraft players keeping 15 barracks producing but never queueing more than one unit is considered macro because its not fighting specifically. And people who don't play starcraft were like, wtf are you starcraft players talking about? That is micro management. It clearly fits the definition. I and others have argued that pretty much any aspect of an RTS where APM is important is micro.

[color=#FF0000]I suppose you could argue that the micro/macro distinction has moved beyond management.

However, if you mean to say that popular RTS nowadays are not about macroscopic decision making and strategy, then I can agree with you. However, non-micro RTS games I have played suffer from the issue of poor AI, making it extremely frustrating for players.

[color=#ff0000]Dominions doesn't count because its a TBS series, but Majesty has almost no micro. You can't even directly control units. To be fair, it makes many classic RTS fans incredibly angry. If you expect your super micro/APM intensive skills to carry over you would be in for a shock.

Also, no matter how non-micro you TRY to make the game, the element of unit pulling will always exist in any RTS. (Pulling a low hp unit out of range of the enemy, then sending it back in).

[color=#ff0000]This is also not possible in Majesty. Units do their own thing, you can only influence them with reward flags.

EDIT: Also, leveling in separate spheres is too complex. A better implementation of the same concept would be to just create a hero that specializes in each sphere, so you can make the decision when you recruit/train the hero. The total war series (although not truly RTS) does this, and it works brilliantly.
[/quote]

[color=#ff0000]Leveling in separate spheres could be very complex if you had to micro every unit starcraft style.

[color=#FF0000]Imagine that you were playing a game where you made an army and sent it off to guard an iron mine. The map was like 2048x2048 and you had many many areas to deal with. You might expect to have those units defend that are by themselves with minimal interference.

[color=#FF0000]I was actually thinking about a sort of compromise between Majesty style influence and Dominions style order setting.

[color=#FF0000]You would have direct control over your Hero unit but other units would be controlled by a set of orders. You might set the group order on some units, and then set the garrison order. So that group would stay together and defend the target of the garrison order from attack.

[color=#FF0000]To get units to attack something you would assign them to a group, or not, but grouping may be more efficient and make them cooperate compared to many single units, and assign an attack order. Then they would look for nearby enemy bases. Maybe you also assign priorities like kill towers first or kill other units or w/e.

[color=#FF0000]In this case checking back on units to spend experience isn't as much of a problem as you are largely relieved of micro, aside from choosing what units to produce.

[color=#FF0000]Macro would involve selecting unit specializations, like you have options on how to evolve your mages guild, not the single mages, although you have that too, but like, this guild is pyromancers and this guild is geomancers.

[color=#FF0000]I had been thinking about some pretty complex magic, like summoning rituals, so mages may be able to switch between direct control and orders, but standard units like archers and such would only be able to be controlled with the order system.

[color=#FF0000]You might just have a separate leveling system for non hero/mage units. I have a lot of freedom on how to handle this, so I kinda wanna try out some new methods of dealing with things. You might just set specializations for regular units and then power up automagically instead of you specifically dealing with them.

[color=#FF0000]I really want to get rid of the idea of massing units anyways also. I don't like it.

Also, no matter how non-micro you TRY to make the game, the element of unit pulling will always exist in any RTS. (Pulling a low hp unit out of range of the enemy, then sending it back in).

Not true. You can have sufficiently non-specific commands, and/or delay in command execution, and/or mechanics which make that kind of disengaging not worth it. Or make it automatic, as it pretty much is in real combat.

Having non specific commands or minimal direct control really works on cleaning up the micro.

I guess the general sentiment in this thread favors the older RTS-style games. I apologise for my generalization, then. Since I'm a fan of more standard RTS games, I speak from experiences such as C&C, AoE, WC, SC, etc.

However, I do feel like it's fair to assume that your target audience will include people like me. In my opinion (and with no offense intended, I understand conflict of opinions can be touchy) you're imposing mechanics from very specific RTS games onto the genre. For example, Majesty is a different take on the RTS genre compared to most RTS games. You might enjoy the gameplay, but I have never even played it.

I will use your definitions for micro and macro from now on, though.

To reference micro-light "standard" RTS games, AoE would be a prime candidate. Unit pulling is still effective in AoE, but less so because of the large mix of melee and ranged units melee units will be taking most of the hits and it is obviously silly to pull melee units taking damage from ranged units in AoE. Positioning micro is also lessened because of the formation options. Queueing units is relatively not punishing. It also emphasizes gameplay based on your definition of macro, where a large portion of thought goes into unit selection and technology. Certainly this is not an APM-fest by your standards?

But the points I made about heroes also applies to AoE. I wouldn't want to spend all this time on maintaining a hero, because it means my influence is more localized to the hero than multiple map positions. I want to worry about what buildings I'm going to build and where they are positioned. I want to worry about scouting my opponent. I want to worry about finding resources. It is definitely viable in AoE to send a couple of units to defend a position, and not have to maintain them. This is generally achieved through static defenses, walls, and ranged units. On the other hand, micro is definitely important in AoE when you're attacking, because you need to prioritize targets.

From what I gather from your posts, your order-only gameplay is no different from standard RTS games and can have APM applied regardless. You can still order retreats (I would hope), so in the case of a large map, superior APM in order to watch and evaluate multiple battles at once would still beat out those who have low APM and are slow to react.

Really, how one would minimize the amount of micro required in an RTS is by automating it. For example, in SC2, if units didn't cost resources to queue, had infinitely long queues, and had order-based controls to the point where stutterstepping is not effective, then it would pretty much match your expectations of an RTS?