RTS game features and when to tie a bow

I haven't blogged in a while but I am still working on my Torque RTS, Orcs vs. Martians. It's coming along quite well, and is very playable at this point. It seems like there have been countless challenges and technical hurdles to overcome in the development of this game, but so far I've gotten through them all, or at least the important ones. Currently, I'm working on the game's special effects, like making buildings shake when hit, particle effects, etc. Fun stuff.

A useful development tip: do not click this link if you ever want to get work done.

I had lunch with my former boss a couple of weeks ago. We still keep in touch to discuss the software biz and Longhorn football (go Horns!). He said something like "Sometimes it's just time to tie a bow on a product, and ship it." I think I'm agreeing with that idea at this point.

What's funny is, I still have a list I made at the start of this project, "What's fun about this game". I think I've only implemented 50%-60% of the features on that list. Maybe that's a good ratio for a game? I have no idea. Development seems to take forever, even when I feel like I'm working productively. I still haven't finished implementing the good ideas I got from Beta 1 feedback, eleven months ago. But I think it's time to tie a bow on this game, meaning dropping game features I had hoped to get in.

Even with only 50%-60% of the features I wanted, I think OVM now has enough to differentiate itself, and make a successful product.

OVM's got:

Unlimited upgrades. Players can upgrade their firepower, range, view distance, armor, etc. -- indefinitely. Goodbye, stalemate games. I love winning via superior technology, so of course OVM had to have this feature. Unlimited upgrades also make games less predictable, as you have no idea how upgraded your opponents are, until you encounter them behind the fog of war. You may be in for a shock after thinking you had the game won. Upgrades gradually get more expensive, to keep them from spiraling too much out of control.

Random maps. OVM's random map generator is powerful and can produce a wide variety of maps. I love exploring an unknown RTS map; there's just something about discovering a big ol' pile of resources, or a really strategic location, and then rushing to take control it. The map generator has a high success rate at producing maps that both look good and are chok full of interesting tactical features. Hopefully, that will give OVM high replay value. One downside though is sometimes the generated maps are pretty unfair.

Strategic terrain. Terrain in OVM plays a strategic role on multiple levels: waterfront property tends to re-grow trees fastest (yielding more resources); high-elevation terrain gives vision and range bonuses; and choke points of course are always candidates for fortification. I think it's fun to have terrain whose strategic value is defined by multiple criteria.

A.I. that doesn't cheat. The A.I. does not cheat by seeing the whole game map - it has to explore, just like you do. Every A.I. player has his own, limited view of the world, and reasons with that view only. Fog of war is fog of war -- if the A.I. can't find you, he can't attack you. I think it's more fun that way. Besides, I hate it when A.I. cheats (I've seen a survey somewhere that said most gamers feel the same). And it's a hoot to see OVM's A.I. make the same blunders a human player does - like trying to find a shortcut through the fog of war, only to be blocked, or by sending an attack force against an enemy base, only to discover the base no longer exists. The A.I.s' view-of-the-world continually grows stale and out-of-date after it explores, so it also continually has to re-explore. Me likes.

A.I. that fortifies choke points. The A.I. analyzes the passages & intersections in the terrain and, if it can find a few key choke points that protect a lot of land, it fortifies those choke points. It correctly recognizes dead-ends as not needing fortifying. I've noticed that even many AAA RTSs don't do that correctly(!) This feature makes the A.I. hugely more capable and more human-like, and IMO, more fun. (and it was darn hard to implement.) It also works off the limited-view-of-the-world feature, so the A.I. only discovers choke points as it explores.

Convenience. The game is user-friendly and easy to get into (I think). There are no building dependencies and no tech dependencies to memorize. You can jump right in, and start building / upgrading everything. Workers automatically start harvesting if idle for a couple of seconds. Workers avoid harvesting where enemy buildings can shoot them, reducing micromanagement. Being attacked off-screen causes a helpful pop-up with an arrow pointing in the attack direction, and info about who's attacking you. I intended a lot of these convenience features as a way to lower the bar to playing an RTS, so that semi-casual players could get into the game whereas otherwise they might not (hopefully enlarging my customer base).

I'm proud of those.

One thing I haven't done, and am considering not doing, is having a campaign mode. My main focus has been random maps. Can an indie RTS sell well without a campaign mode? or would that be a fatal shortcoming?

Obviously, Orcs vs. Martians is not the be-all, end-all RTS -- but it's not meant to be. I hope its got enough going for it to be a hit...Lord knows I've invested the time in it.

blog


- - - - -