Friday, November 4, 2011

What's Missing from RTS Games?

I managed to work through a couple of animation bugs this evening.  The odd thing about a bug-squashing streak is that it makes the next frustrating bug much, much more frustrating.

Instead of banging my head against the only QWERTY keyboard in all of France and running the risk of having to type on the abomination that is the AZERTY keyboard, I decided to do some brainstorming on the GameDev subreddit.  The result is this post where I ask the following question:  what are your biggest gripes in RTS gaming?

I won't lie, I was sort of hoping to be praised as a visionary, but the tough love I got from a few redditors was both useful and insightful.  The following points, in no particular order, were made:
  • Build orders are not the enemy.  Don't hate them.
  • Unit responsiveness is of capital importance
  • It sucks when RTS games go on and on forever
  • On the other hand, it sucks when you win by running the clock
  • Don't use units as cannon fodder
  • UIs need some serious innovation
  • Cover systems suck
The last one was quite an eye opener.  I had actually enjoyed the cover system in Company of Heros, but the point was made that it restricted a player's use of the terrain.  Since Project Orbit will rely heavily on terrain influence, cover systems may be more of a bug than a feature.

Food for thought, as they say...


  1. (This is evizaer from reddit.)

    What the redditors say is mostly invalid or entirely based on taste. It's more a symptom of the current (terrible) state of RTSes instead of what CAN and SHOULD be done. Existing RTSes cater to a twitchy crowd by being mostly micro-based and tactical or action-oriented. You shouldn't trust people on reddit without thoroughly checking their statements to see if they are at all valid. You should trust your own judgment and the results of your own experiments.

    Play RUSE. That game had very little in the way of build order memorization because you had to remain flexible and build in reaction (or in prediction) to your opponents moves. Build orders are not an awful thing--you just don't want them to dominate the way that people think about strategy in your game. To minimize build order prevalence, force players to stay fluid and play based on how their opponent plays (or try to encourage them predicting their opponent's behavior).

    The validity of cover systems depends on the scope of your game. You should decide if your game is on the operational, campaign, or tactical level (or some combination, though you'll benefit from picking a scope and sticking to it). At any but a tactical level, direction cover is meaningless, though defense-altering terrain still makes sense.

    "Unit responsiveness" complaints are an attempt to encourage you to introduce micromanagement burdens in order to artificially increase the skill cap. Increase the skill cap by designing a game that emphasizes planning instead. Good AI handling the more tactical kind of maneuvering can more than make up for "responsiveness" being slightly compromised.

    The "games too long" complaint can be countered through some kind of continuous-countdown victory point system. Please do not allow resource strangulation to be a viable long-term strategy, because then you end up with long games.

  2. The previous poster has it mostly right. I would trust reddit for very little. I do disagree with his stance on unit responsiveness, but I suspect we may be thinking about different things. Units that respond slowly are totally fine. Units that appear not to respond, or otherwise just stand there and dick around (Company of Heroes, Dawn of War 2) frustrate me to no end. It's less about the speed at which units respond, and more sending a clear indication that the unit is responding.