Procedurally Generated Combat Scenarios


OK, I didn’t intend to make a 3rd “Future of FPS” article, because God knows I write enough about FPS here as it is, and a lot of people don’t want to read about “crappy shooters” all the time. I actually meant to write the next “future” entry on RPG’s. But, I realized I left out one major innovative feature which I have often discussed and suggested on tactical shooter forums.


In short, no matter how realistic the weapons, aiming mechanics, movement mechanics, and environment “set pieces” are in tactical shooters, they all have one huge gaping hole in realism; you play on the same confined “battle arenas” wave after wave, round after round, match after match, month after month, year after year.

..and why should it be any different? Virtually every other game out there, barring dungeons in a few old RPG’s, feature static levels that never change. That’s just the way it is. Furthermore, to a degree, soldiers are briefed on combat zones before going in, so even in real life they’re not going in totally blind.

However, this simple factor causes probably all of the absolute most unrealistic tactics and behavior in tactical shooters. It’s not just about knowing the lay of the land, it’s the fact that everyone starts in the same places at the same time every round. Thus, all combat in the game turns into a routine where people take the same paths and run through the same chokepoints and danger zones at the same time every time you play.

The most immediate thing that comes to mind is the drawing of knives during the beginning of every round of Counter-strike in order to run to the chokepoints faster. Obviously, in real life, when you enter a combat zone you would never just sling your gun, pull out a knife and run around like that; someone could pop up around the corner at any moment and cap your ass. But when everyone is running the map on routines, you know exactly when and where to expect enemy encounters and could practically run around the map with your pants around your ankles and still pull them up in time to enter the firefight in fully-clothed dignity.


Another unrealistic phenomenon, the recent discussion of which sparked my memory and inspired this article, is the technique of throwing hand grenades over buildings into totally unseen other areas of the map at completely arbitrary times to score kills that would appear to the uninitiated to be solely the product of psychic x-ray vision coupled with camera-guided rocket hand grenades. Once again, this is enabled merely by knowledge of routes and routines of the map. Once he knows the target’s location at a specific time, the player can simply practice offline by himself till he’s determined exactly when and where to stand, and at what angle is needed to throw the grenade and “magically” obliterate unseen enemies on the other end of the map.

These are but two examples. I could probably name many more, but if you’ve played a tactical shooter before, you can probably think of countless more on your own. For all intents and purposes, once a tactical shooter has been played enough, the combat essentially runs on a “script”. Run to this place and throw a grenade over the wall, run to this chokepoint and wait for the guy to run through, mow him down and run to the next chokepoint, etc.

Sure, it varies a little bit, and different players can choose between various routes and do things slightly differently each time, but once you’ve been around the block you pretty much know all the different ways the round can play out, then it’s just a matter of figuring out which ones are being used and acting accordingly. Even when a veteran is playing a fresh novice that doesn’t know the routines and simply runs about “randomly”, the experienced player will know all the common n00b mistakes and use those to track him down.

This phenomenon is not all bad. It can be fun developing various strategies based on all the different routes and routines on the map. You see an enemy use one tactic in one round, and predict based on that what he’s going to do in the next, then you determine a path that will get around his trick and give you the advantage. When he changes his approach, you have to devise a new one of your own, and so on.

It’s a fun little game, and a lot of people have gotten a lot of enjoyment out of many tactical shooters this way, but it has its limitations and eventually becomes repetitive and boring (to say nothing of the huge departure from realism it entails).


One way to get around this is to randomize the spawn locations and objectives each round. Even though you know the layout of the map, you don’t know where the enemy is and they don’t have a routine down, so you suddenly have to approach things with a lot more caution. One classic example of this is the custom gametype “Dynamic Take and Secure” or DTAS for Infiltration. It would spawn each team as a group randomly in some part of the map, and assign some random location as the objective. The ‘attacking’ team would have to find the objective, get to it with at least one team mate, and hold the position fending off all enemies for a set amount of time till it was “captured” and the round was won.

The gametype was very popular in the Infiltration community, and become the standard way to play the game for a long time even though it was invented by people in the playerbase and not the actual developers of the mod. Unfortunately few other tactical shooters have adapted anything like this. Maps are always developed with specific spawn points and objectives in mind, and once the maps are built, the developers don’t want to have to go back in and add a ton of spawn locations or do a ton of extra coding to support this “odd” unconventional gametype some fans have suggested which may not even catch on.

However, it’s my belief that this is really the only way to execute a proper tactical shooter these days. Games that structure things linearly with defined spawns and objective locations are just blindly repeating what has been done already, rather than bothering to innovate and push the boundaries of the genre. A game built around dynamic starting positions and objectives will entail a lot more realistic combat strategies and allow for a much greater number of possible ways for the match to play out; essentially causing a “big bang” on the replay value of the map, and consequently, the game.


One thing which is just waiting to be done in tactical shooters, but hasn’t yet (AFAIK) is procedurally-generated maps. Procedurally-generated content has actually been around since the early days of gaming, when there wasn’t enough memory to support large quantities of pre-made content. Based on certain algorithms and parameters, the game would essentially create graphics, or characters, or levels on the fly.

The result of this is that this content could conceivably be different each time you play, and theoretically the game would have practically infinite replay value. While the technique is still used today to some extent in MMORPG’s, the idea is being promoted in a big way by the upcoming game Spore. I’m not going to fully explain everything in Spore, because that would be an entire blog entry unto itself, but suffice it to say that procedurally-generated content is the centerpiece of Spore’s gameplay, and the game is looking to place PGC (I’m not going to spell that long-winded phrase out again) as the next big buzzword and trend in game design.

It’s pretty easy to imagine how procedural generation would be used to drive tactical shooter environments. The developers would create a ton of “set pieces” and “building blocks” which the game would randomly assemble based on various parameters to create a unique, functional, and realistic combat environment each time you played.

This way, players would never know the map “too well”. They couldn’t whip knives out to rush to the objectives faster (doesn’t work in real life anyway), they couldn’t run “routines” around the map, throw grenades into unseen places being sure to score kills, or camp certain entryways knowing for certain that enemies would come through. A lot more caution and teamwork would be necessary to survive, and the most successful players would be those who could adapt to each new environment the fastest, developing strategies that suit any situation, rather than amassing a ridiculous amount of knowledge on a map’s combat “script” and merely developing strategies based on that.

This doesn’t just have an application in tactical shooters. Procedural generation and randomization could go a long way to insert replay value into practically any genre of game. Tactical shooters are simply the best example of how static, unchanging content diminishes gameplay… any aspect of a game repeated tot much will get old after a while. Developers simply haven’t recognized the potential of this mechanic or been forced to rethink how they do things. Hopefully, at some point in the future, games that provide a linear, scripted, unchanging experience each time you play will eventually be seen for the primitive, obsolete, and stale design formula that they are. Until then, all we can do is hope.

__________________
[Only Registered and Activated Users Can See Links. Click Here To Register...]