Creating a Single-Player Combat Space


PeteEllis

01-KZM_Panorama.jpg

This article is the first installment in a three-part article that looks at the considerations for creating a single-player combat space, using a walkthrough of the first battle in ‘Killzone Mercenary’ as a working example.

Creating a Single-Player Combat Space Part 1

This article will explain how to create a combat space for a single-player campaign, using my work on ‘Killzone Mercenary’ (hereon referred to as KZM) as an example.  There is already a fair amount of literature on the different methods you can use for creating a combat encounter, but I felt that none of it really discussed how to arrange the layout in closer detail, nor did they discuss where the different elements were appropriate.  In my early work I tried to jam in all the concepts for encounter design without fully understanding how they affected the player’s experience.  As strange as it sounds, I discovered there were times when it was better to restrict the number of elements being used to provide a much more focused and coherent experience; sometimes less is more. I will take you through an example to explain what I mean and how this can be the case.

I will use the very first combat arena in KZM as it’s a small encounter where I can explain in depth what goes into even the most basic combat space. This encounter is a fight against assault troopers who are trying to stop the player from escaping the building and reaching the objective building; the ‘Halls of Justice’. I designed bigger encounters that featured many flanking opportunities and complex circular navigability but focusing on a simple encounter allows me to explain certain techniques in detail and where I purposely removed some elements to balance the difficulty and give the player different experiences.

First of all I will explain two important aspects that must be considered for combat creation; AI metrics and weapon choice. I will then take you through a step by step walkthrough of this first encounter explaining in detail the reasons how it was designed and constructed for optimum player experience.

Metrics

The design mantra ‘form follows function’ should be the basis when creating an arena layout; that is that the arrangement of geometry should derive from its purpose. The arrangement should support the function not only of the style of experience you want to create (is it a tight corridor section with close quarters combat or an open space with multiple routes and options, for example) but it should also support the main element that makes up the combat encounter; the enemy AI.

When considering the layout for the AI or non-player characters (NPCs) that will populate your environment you have to consider their metrics. These are the numerical values for how the NPCs move around and use the environment and the differences between various NPC enemy classes. This isn’t something people tend to talk about and so it can be easily forgotten or missed, yet it directly affects how your enemies will move and react.

For example, in KZM the standard enemy NPC class were the Assault Troopers. These soldiers could be given patrols and animations to perform whilst they were in an ‘unalerted’ state, just like every other enemy class.  However, when they were in an alerted state their behavior changed so that they used cover points to move around the combat space. The maximum distance between cover points that an assault trooper would move was 10 meters. This meant that any cover point that was further away would not be considered, so we needed to make sure when creating combat spaces which used assault troopers that there were enough cover islands so they could move around. If there weren’t, the assault troopers would just stay in the same spot and could risk looking less intelligent.

The assault troopers also tried to maintain a distance of 15m whilst they were trading shots with the player. The behavior was that if the player got closer than this range, but not so close that they were in melee combat distance (5m), the assault troopers would retreat to this mid-range distance of 15m. They would also never choose a cover position that was closer than 15m to the player, so when we created combat spaces we had to make sure that there was enough variety of cover positions in the >15m range.

For the production of KZM we used the ‘Killzone 3’ engine and modified it for the PS Vita. In ‘Killzone 3’ the assault troopers picked their cover within a range that was further than 25m from the player, but we discovered that this was too great a distance for the enemy to still be clear and readable on the PS Vita screen. In our modified version of the KZ engine we had to reduce the combat distance to 15m, which meant that the original combat spaces we had created using the ‘Killzone 3’ metrics also needed adjusting in order for the NPCs to still work. It is an unfortunate truth that the game metrics, be it for the AI or otherwise, can change within a game’s development, which means that your combat arenas will also need to be adjusted.

Weapon Consideration

The metrics for both the player and enemy weapons were also considered. As this is the start of the game we can be more certain that the player is using the default starting weapons, at least on their first playthrough, before they have earned enough credits to buy a new arsenal. Therefore, the combat distances of enemy placement were considered to be comfortably within range for the player’s assault rifle.

The enemy assault trooper archetype used assault rifles that were balanced to have a short range of <10m, and a long range of >20m. This meant that their behavior was to try and keep the player within these ranges and would thus move around the environment to try and maintain this. This was important to consider when building the environment so we could determine the amount of movement the troopers were likely to perform. This is important for balancing difficulty as a moving target is harder to hit.

Foreshadowing

As this was the opening of the game, we wanted to make it compelling in order to grasp and hold the player’s attention; we wanted to start with a bang. If the first lot of encounters in the game only included assault troopers with nothing else to differentiate them it may not have been so compelling. Therefore, we decided to include a significant Killzone enemy vehicle; the Helghast Dropship.  Of course it would have been far too difficult to fight a Dropship at this point in the game, so instead it was used as an impressive introduction of enemies into the arena using the rappel ropes from the ship itself.

Using the Dropship at the end of the encounter, it was important to foreshadow its existence prior to its introduction. The level’s opening cut scene introduces the buddy character, Ivanov, and the narrative that he and the player are infiltrating the building whilst trying to avoid the searching eye of the Dropship.

02-Dropship.jpg

The foreshadowing of the Helghast Dropship

Once the player has control they make their way up a flight of stairs learning how the movement works and feels whilst being in a safe environment. Once at the top of the stairs they enter through a door where they are introduced to the new melee attack which utilizes the touch screen on the PS Vita.

First Wave

After a successful melee attack the player enters through the door to the first combat area. The composition shows the exit of the arena in the top left third of the frame.  Central to the player’s view is where the first pair of enemies enter from, ensuring that their arrival is not missed.

03-OpeningComposition_Grid.jpg

The exit to the arena is in the top left section of the opening composition

Starting on the level above, the two assault troopers vault down into the gameplay space, to give their presence a more dramatic opening than merely walking in through a door. Their animation and movement also ensures that they catch the player’s eye if they aren’t looking in the desired direction. These vault down animations were 4m high, the standard height for a room in KZM, which meant this was a metric we had for the balcony and floor above.

04-OpeningEnemyArrival_Arrows.jpg

Two assault troopers drop into the environment from the level above

Once the assault troopers had landed in the arena they became a lot less mobile than their standard behavior so that they were easier to shoot because, as previously mentioned, a moving target is harder to hit.  As this is the very first section of combat the player encounters in the game it was important to ensure that it was easy to get to grips with.

None of the enemies were waypoint/navmesh restricted to certain areas in order to limit their movement as this could potentially lead to NPCs not behaving correctly under differing circumstances. In fact, there were only a very select few instances where we waypoint/navmesh restricted any characters in the whole of KZM; we instead crafted the environments to support the behavior we wanted from the NPCs. This was important for consistency; if you restrict areas and zones for the AI then they won’t behave consistently with what the player has learnt. This would lead to the player not being able to predict their behavior and therefore won’t be able to plan how to attack effectively.

Here, in this first section, the two assault troopers took cover at two upright pillars of high cover and an overturned sofa of low cover. They didn’t tend to venture further into the environment unless the player had for some reason retreated to the edges of the level. The reason they wouldn’t move and advance on, or flank the player was because the other cover options in front of them were within 15m of where the player was likely to be stood. This caused them to be more static and thus easier targets to allow the player to get to grips with the shooting mechanics.

I also chose to mainly use higher cover here so that when the enemies lean out of cover their shooting positions allowed the player to shoot their full body, which was a bigger target than when they poked their heads over the top of the low cover positions.

Low cover positions are great for seeing the enemies move around and change their positions, as the tops of their helmets are visible over the top of the cover. Enemies are much harder to track when they use high cover as it breaks line of sight to them, so this is usually the harder option. However, as they have restricted cover positions and weren’t moving around in this specific situation, it was the best option to use for less difficulty.

First Combat Front

A ‘front’ is the perceived line or boundary that faces the enemy and is the nearest position which combat should be engaged from. The ‘fronts’ used here create boundaries between the two sides; a front for the player and the buddy character and an opposing front for the two assault troopers. This was the simplest setup to start the player off with and it only required two sets of cover points as I didn’t want to encourage the enemy to flank the player at this stage. This section of the encounter only needed these few pieces of cover (in the image below) in order to work and the other pieces of cover were actually for further waves of combat.

05-Front_1.jpg

The two fronts and the cover setups providing it

It’s also worth noting that the cover object which the buddy character crouches behind is positioned further forward than the arrangement of cover that the player is drawn to. This is so that the buddy character is in the player’s view so they always see the buddy’s actions and involvement. It wouldn’t be optimal to have a buddy NPC that the player rarely saw. The buddy is also kept near to the player in order to maintain a close relationship and the feeling of being a team. Empathy is directly related to proximity between characters, so if the buddy was further away from the player they would experience a much more detached feeling towards them.

Continue to part 2 or go back to the homepage

Index:

Creating a Single-Player Combat Space Part 1
Creating a Single-Player Combat Space Part 2
Creating a Single-Player Combat Space Part 3

Copyright ©Peter Ellis 2016. Killzone™ Mercenary is the property of Sony Computer Entertainment © 2013. Killzone is a trademark of Sony Entertainment Europe. Killzone: Mercenary is a trademark of Sony Computer Entertainment America LLC.


spence, Beck, kos8bit and 20 others like this


User Feedback


Cool. I feel I learnt a lot about encounters on Crysis, but working on Killzone 3 really solidified my understanding of when you talk about fronts and arranging the cover to create a no man's land, etc. It's cool to read this, as I think people who've worked on a few titles internalize this stuff but it's great to see it all written down and illustrated as a reminder and also for people new to the process!

You should probably also link to a YouTube video of the encounter playing out too, btw!

PeteEllis, Squad and Sprony like this

Share this comment


Link to comment
Share on other sites

I think this was really excellent. My favorite parts were the pictures, I think having an enemy character leap over a wall and down into the battlefield is a cool little touch that will probably do as you designed, provoke a player reaction especially in the early-game where stuff like that is novel for them.

PeteEllis and Sprony like this

Share this comment


Link to comment
Share on other sites

The way information is divided into layers (TV screen to PSVita screen in this example) and the affect it has on pacing combat, presentation of cover, objectives, aesthetics within each level is a very valuable lesson on perspective.

PeteEllis and Sprony like this

Share this comment


Link to comment
Share on other sites

Thanks very much, I'm glad the article was helpful :)  And I'm glad the diagrams/images were appreciated as I spent a while making them!  But I felt they were worth it :)

@Alf-Life Here is a YouTube link of someone playing this encounter (I just picked the top one in the search), and its good to see they play it pretty much as I intended!

https://youtu.be/iYj-nkEs1xg?t=4m35s

 

Sprony likes this

Share this comment


Link to comment
Share on other sites

Nice article Pete, really shows how much thought has to go into each encounter design.

Not sure I am a fan of the readability of the climb up exit (atleast as shown on the screenshot), always tricky to get good forward gates.

Edited by Vilham
PeteEllis likes this

Share this comment


Link to comment
Share on other sites
On 5/19/2016 at 0:15 PM, Vilham said:

Nice article Pete, really shows how much thought has to go into each encounter design.

Not sure I am a fan of the readability of the climb up exit (atleast as shown on the screenshot), always tricky to get good forward gates.

Yeah, although in this case having an AI Buddy lead you there, dialog, the pipe shimmering and a tutorial button prompt (and probably an Objective Marker too I'm guessing?) utilizing the strengths/"crutches" of the game design is enough.

PeteEllis likes this

Share this comment


Link to comment
Share on other sites

My question to these kinds of thing is always: What do you do in encounter design to pace out combat fatigue? Usually encounters get stale fast, and I find that the introduction of new enemy types (and their quantity) accelerates fatigue rather than creating counter pacing. I see this problem in pretty much every shooter out there, so I assume it's largely unsolved in these kinds of games?

Share this comment


Link to comment
Share on other sites

@Vilham @Alf-Life Yeah you're totally right about the screenshot I used in the article, it wasn't the best angle to show the buddy-boost/climb up.  The next video down in youtube (linked below) shows what happens a lot better (at 8:09).

The readability primarily comes from the buddy leading in this section.  There are a lot of subtleties that support it, like the negative space above the wall is only on that side of the level, where as all the other walls are blocked out, which also supports the 'lead-lighting' here as its a brighter area.  We didn't want the top of the wall to be any lower (regardless of the 4m consideration for the animation) as we didn't want it to seem like the player could jump up there themselves as this would have been a worse situation - if it looks like the player can get somewhere by themselves they should be able to.

 

Share this comment


Link to comment
Share on other sites


Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now