informa
6 min read
article

L4D: FPS or RPG Part 2

A long delayed follow up to my first post on L4D and the mechanics of RPGs it uses and whether the idea of an AI director is applicable to other genres

After much delay, I get around to a follow up to my first blog post, on how the design and gameplay of Left 4 Dead mirrors classic pen-and-paper game design. After posting up my first blog post, I had a comment that basically said "So what?" What, after all, was the use of recognizing L4D and it's RPG like mechanics and it's story telling mechanics? And that's a good point. Once we recognize what L4D is doing, what can we do with it? More specifically, can we take the mechanic of an AI director and apply it to other genres or other styles of play.

As far as recognizing the story mechanics behind the weapons and items of L4D, I think it's simply boils down into recognizing that it's not necessarily broad choices of 'stuff' that create such story mechanics but a careful restriction of choice and conscious intent. L4D uses only one mechanic for that - weapon choice. Each weapon in the game does not merely serve as flavor and 'power' choice but also a role choice. As mentioned in a recent Gamasutra article, each weapon is functionally a broad class. But especially in the second game, each weapon is also further increased/reduced into a resource - if,  as I did, one studied the placement of the various weapons, one will note that L4D2 is both more generous and more stingy with it's weapons. The increased variety of weapons and the longer levels means more opportunities for weapons.

However, the increased variety of weapons and greater ability of the AI director to vary weapon drops also means that it's not a given that one will recieve Tier 3 weapons - it is as common to see Tier 2 or a mix of T 2 and T3 weapons. More so, it is possible to see several variants of a certain type of weapon - several shotguns or an SMG and some melee weapons. Equally significantly, the abundance of ammo piles is reduced to perhaps one or two per level - generally one at the start and one near the mid-point. Most often, players are reliant on weapon caches to either replenish a weapon's ammo (if it carries the same weapon as they are using) or forced to switch to a different weapon (if not). What this does is encourage weapon usage but also forces upon the players an interesting choice as well as increasing the atmosphere of the game.

 That choice is that of role and class. The unpredictable availablity of role/class forces players to either conserve ammo on a favored (or suitable) weapon or adapt with a less-optimal weapon. The atmospheric effect of this means that players are working with less resources; players seem much more consciously aware of their ammo and other resources in the second game. This paradoxical reduction of resources also further encourages the close knit gameplay of the game despite the daylight heavy visual aspect.

On to the second major story system, the AI director. As a refresher, let's identify the AI director again. It is a system that runs parallel to the difficulty of the game; as a game mechnic, it does not affect the intelligence of the NPCs, their health, their damage, or any sort of 'hard' numbers. Instead, it is a story system - it affects resources, pacing, and atmosphere. In Left 4 Dead 2, it gains the ability to affect path and world layout. It is then, an intermediary between the player story and the designer story - the world story, such as it is. The effect the game can generate the story of the world before the player arrived (or after) and allow the designer to create the important details without having to work on unimportant details.

In open world games like Fallout or MMOs, it is generally difficult it generate some sort of intermediate story - there is the overall plot and the moment-to-moment story/experience but little in way a medium term story. An AI director like system would be useful, I think, in filling in this story. An AI director would be able to take control of locales and such to generate individual elements. Rather then a static room with a mini-boss, such a boss would occur (or not) based on the AI director's understanding of a climax to that sub story and the player's progess. Objects and set pieces would change - not necessarily because of anything the player or opponent did but simply because. By also carefully limiting the number of goals a player has, the AI director could have a little more certainity with which to work within.

One game that has done something like this is, perhaps surprizingly, The Sims 3. In this latest interation of the series, there are three levels of goals (and thus stories for the player to pursue). Wishs are generall short term goals one can do in 10 minutes. Lifetime wishes are long term goals; generally, something one will accomplish over several weeks. Opportunities/adventures are something to be done in several hours. Most notable though opportunties aren't always on the whim of the player - some can only be done in certain hours, others certain skills. Still others -require- a sim to have certain skills before appearing. At least in the first playthrough of the Sims, the illusion of a world reacting to the player/their sims is fairly potent. Designing systems like these, I think, could help the progession of games in terms of story telling. By giving the game the ability to fill-in the game world between designer set pieces and bridging the gap between designer and player.

- - -

Short Random Bits: I'd love to see in L4D3 the ability given to the AI director to place such flavor objects. Such as random bodies/covered sheets near ammo caches, or various locations have the last stands of other survivors.

When L4D first came out, I had a mis-understanding as how the AI worked in regards to stat tracking. I initially believed that the game tracked how well a player was doing and using that data to generate a variance in difficulty within a given difficulty. In effect, a 'producer' relying information to the AI director with player-specific information it could use. Are they good against hunters? Are they fast on the trigger? Etc. More importantly though, I believed it used this information in Multiplayer. If a highly skilled player was playing with a player of low skill, the AI director would 'favor' the highly skilled player - influencing the individual AI so that hunters would be more likely to target the highly skilled player for instance. Thus creating a subtle social pressure - the less skilled player would feel useful but also having some less game difficulty placed on them while the more skilled player would have greater pressure on them and greater difficulty.  I'd love to a game take advantage of all the stat tracking and do something like this; intelligently using stats instead of simply making them fun but useless numbers to look at and compare.

Latest Jobs

Treyarch

Playa Vista, California
6.20.22
Audio Engineer

Digital Extremes

London, Ontario, Canada
6.20.22
Communications Director

High Moon Studios

Carlsbad, California
6.20.22
Senior Producer

Build a Rocket Boy Games

Edinburgh, Scotland
6.20.22
Lead UI Programmer
More Jobs   

CONNECT WITH US

Register for a
Subscribe to
Follow us

Game Developer Account

Game Developer Newsletter

@gamedevdotcom

Register for a

Game Developer Account

Gain full access to resources (events, white paper, webinars, reports, etc)
Single sign-on to all Informa products

Register
Subscribe to

Game Developer Newsletter

Get daily Game Developer top stories every morning straight into your inbox

Subscribe
Follow us

@gamedevdotcom

Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more