Simplicity vs Complexity

Last week, I meet the animator Candy Guard. She was giving a lecture about her work, and something that struck out at me was how much she based it on reality. She specialised in observational comedy based around everyday, relatable experiences. Part of the charm of her work is seeing scenarios that remind you of your own life.

I asked her how essential it is for her work to include these naturalistic scenarios, and whether or not she could imagine herself doing anything more abstract. The answer she gave didn’t really answer the question. Instead she went into detail about her art style. She uses simple line art for characters and environments, as she considers it to be ‘more universal’ so that more people would be able to identify with the characters. The simplicity of the visuals allow for the characters to become a near blank canvas for the audience to project themselves into; there are less variables that need to line up between the audience and the character for them to be relatable, it makes the character more malleable in order to fit the mind of the viewer.

I was inspired by this, and asked the question ‘would the same thing apply to 3d environments?’ Would the malleability of the player’s behaviour change if the level was visually simple?

I created a block out of a level, which included a selection of the different cues I had been testing up to this point, namely lights and visual threat.

I also tried something new with some lines as paths for the players to potentially follow.

I’ve only had 8 people play the levels so far, but there were some noticeable differences in how players behaved.

At the beginning of the level I had placed this spike trap. To get over it players need to jump onto the raised beam to give them the necessary height to make it over. If they just jump over the beam without looking, they fall in. This makes sure that players fear the other spikes immediately afterwards. I placed a red light by the spikes so that there is at least some clue that something dangerous is down there.

All simple level players didn't look at the first spike jump and dived straight into them. But 3 of the 4 complex level players noticed the spikes, and made it across safely. 

Although there's not enough people to form a solid conclusion, it would seem that the semantics of the red held more weight in the decorated/complex level. Perhaps the stone temple-like visuals held additional semantics that prepared the players for danger.

This area was a continuation of the previous fire level, where I was testing to see how player react to different levels of environmental threat. With that test, I came to the conclusion that players viewed different sized threats as the same as each other; there is no such thing as something that looks a little bit threatening. Watching people play through these new levels gave more evidence for that. Everyone chose to go through the large spiked path crouching under them to fit through the thin gap underneath. All the players also stopped and looked at their options before choosing a path, so it was clear that this was a thought out decision rather then something they did on auto-pilot.

A choice in the levels between whether to go down a light or dark path still brings up the same outcome as it has the past couple of tests; it’s about 50:50. What could be interesting is how players choose which path to go down. Those playing the simple level where generally more cautious about committing to a path, they would walk down one, turn around to see where the other went, and slowly edged forwards into making a decision. Those playing the complex looking level where more likely to down one path and stick with it, although some would backtrack afterwards to see where the other path led. This is something that I would like to look into more, maybe with an additional level focusing in it.

Finally there was the new test with lines.  This didn’t go as expected, as only one player, playing the complex level, followed the line laid out on the floor. The majority of the players took the first corner on both levels. One player on each level took the corner furthest away. It would seem that the manipulative power of lines on the floor doesn’t change with the complexity of the visuals.

Some interesting result came from this experiment although, as stated before, I’m going to need to get more data before I say that it’s a definite conclusion.

Fire - and lots of it

This week, I wanted to look into how people are affected by threatening objects in the environment. I made a level set in a burning building, where player needs to navigate to the exit. Throughout, there are different sized fires that damage the player if they move into them, with larger fires dealing more DPS then smaller fires. There are some fires that are so small that they don’t deal any damage. At a couple of points, the player will need to move through fires in order to progress. If they are killed, they respawn at the start of the level.

I wanted to see how players reacted to the different sizes of fires; whether or not the amount of threat they feel increases with the size of the fire. I’m also interested in what they would do if they are confronted with two routes with different sized fires down them. What routes would they choose, or would they try to find a third option?

During testing, I saw that all of the players avoided the fires as much as possible, no matter what the size was. Even very small fires, the kind of which you could realistically stamp out, where seen as a threat. When confronted with the ‘2 routes with different fires’ scenario, most players took a third option and tried to go in-between the fires in one of the paths.

Something else that was interesting was how players interacted with the smallest, no damage dealt fires. Only a few players discovered that these particular fires were safe, but each of them believed that each example was an isolated incident, like it was a bug. When they went up to another fire of the same size, they still considered it a threat.

I also had half of the playtesters play the level in an 'over the shoulder' 3^rd person view (meaning they had one to one control over what they looked at). The default character in the UDK is a metal robot, and I wondered if players would notice this and start the level thinking that they wouldn't be hurt by fires. However, they didn’t. They played in the same way that the 1st person players did. This reinforces the notion made in the 'Material Semantics' test levels that players don’t seem to think about the properties of certain materials when playing, unless they are taught it.

Another observation is how the majority of players tackled the final section:

For this, the player needs to jump through the fires to get to the exit. There is an additional route that the player can take, but it leads to a dead end.

All of the players went to the dead end, which is to be expected. The majority of the players saw the two fires as an impassable barrier, and so back tracked to find a new solution. What was interesting was that after they came back to the exit, whether or not they had been killed by another fire and respawned, they went back to the dead end, even though they knew that there was nothing there to help them. They likely went there to find if there was anything that they had missed. One player however said, ‘I don’t know why I keep coming here’, which could indicate some sort habit or a subliminal, hard wired human behaviour involving retracing step. In the Vine level there were some players, after falling to their deaths, would retrace the same route that they took previously, even if it was a difficult one. This is something I'm going to read into more.

3rd person cameras part 2

After playtesting the 3^rd person level with a controllable camera, the overall by-the-numbers outcome was similar, with about half and half going towards light and dark. There were a few key differences in how players acted.

Half of the 6 players missed this cake, running straight past it at the beginning. This could be for 2 reasons, either they didn’t think that they could pick up the cakes (something that they do realise when looking at the second) or they didn’t notice it due to the brightness of the light and the white texture on the cake, making it blend in.

Players were also more likely to go down this dark alleyway. 5 of the 6 players went down here, as opposed to just 1 player when the camera was fixed.

When placing the cameras I made a conscience decision to compose the shot so that the most emphasis was placed on the lit corner to see if it made any difference when the player could control the camera, and it has. The fixed camera position looks to have made the dark path less noticeable and thus a lot less likely for the player to go down. From this it would seem that visual composition, and how the information is presented to the player has an extra layer of influence on the players actions then if they were allowed to control how it was presented.

Something I think I could look into whether or not the presentation of information has a greater influence then the information itself.

Also, if a designer controlled camera was used to guide the player, say in a fixed third person view, what would happen if the player was suddenly given control, maybe from a first person perspective. Would they feel uncomfortable with this control? Would it be like having the training wheels on a bike suddenly disappear? Would there be a greater feeling of freedom and choice?

3rd person cameras

This week I created and tested a 3^rd person camera level.

The goal of the level is to collect all the cakes and head to the exit. The players however don’t know this when they start, and work out the goal at the end when they see a number counter increase when they collect a last piece.

The level uses the same light vs dark idea from one of the previous levels, and I wanted to know if I got different results with a new camera. My hypothesis is that players will feel more detached from the character on screen, so would fell less danger, and thus will venture into the darkness more often. There are two different variants; one that uses a fixed camera that the player can’t control, and another that uses a controllable camera attached to the character.

The fixed camera lets me look into controlling the image on screen as a way to influence the players decision. At a few points, I have the image framed so that the prominent part of the image is a dark or a light path. I wanted to see if this has an effect.

I only managed to test the fixed camera version this week, so next week I’ll be testing the controllable camera, along with any other ideas I have in the meantime.

The level is made up of a few different sections, each with a dark and a light path they can go down, like in the original. I took note of each players general action qualitatively, as well as which path they chose the first time they entered a section as raw numbers.

What I found was similar to the original test, that the amount of players choosing dark and light paths first is about half and half. Out of the 6 players who tested it 3 mostly went down dark routes first, and the other 3 mostly went down the light routes. However, within the players own choices there was a near 50:50 split as well. All of the players who mostly went down the dark paths had a 3 – 2 split out of the 5 choices they made. Within the players who mostly went down light paths was an outlier who only took routes that where illuminated.

There were a few notable behaviours. Players didn’t seem hesitant in going down the dark paths as much as they did with a first person camera. They would usually go past the entrance to the dark path, and then back up to see what was down there. Also, there was only one player who went into the first dark area in the level.

Other players came back to this area either after reaching the end of the level, as they need to in order to complete it, or after going into the dark area of this 2^nd section.

It would appear that either players didn’t want to go into the woods, or they didn’t know that they could. It is known that players eyes are drawn towards areas of contrast, so it could be that in this 3^rd person game, light is more about catching the players eye more than creating an area of safety that the player feels emotionally.

More Vine level testing

After having more people play the vine level, it seems that more people share the same though process as my original tester. Generally speaking, players recognised that there was a variety of density in the vine bridges, but when players played the game they ended up being put into 2 mental boxes, bridges that broke and bridges that didn’t. They didn’t recognise that there was an analogue. Player still avoided the bridges that looked the thinnest, but used all the others.

My next step is to take some of my previous maps, and see if I get the same result if it uses a 3^rd person camera.

A possible new direction

My lecturer Josh, after looking at the ‘wood vs vinebridges’ level, had a brilliant idea. Why not make a level where some of the bridges broke depending on how dense the vines were. Some vines would be thin, and snap easily, and others would be thick and not snap at all.  This was great because in many games' mechanics, interaction is very digital; one thing looks like this, so it does this. The idea of having an analogue interaction is something that feels new, and the level could be constructed with the aim of making it 'feel' more realistic than usual.

The main aim of the level, however, is to see if players understand it and like it. Will they be cautious about the vines? Will they be frustrated at the level? Will it feel like they are having to make judgements?

Currently I have only had one playtester play the unfinished level. They partly understood the idea saying that they thought the number of leaves on the vines indicated whether or not it was stable, but they didn’t understand the analogue nature of the bridges.

To the player, the thought process was ‘Will this bridge hold me?’ rather than ‘How well will this bridge hold me?’ To them it was a binary decision to make, because their action too is binary – to cross or not cross the bridge.

On the same day as the testing, I meet Arthur Parsons, game director at TT Games, as he was giving a lecture at the university. The lecture was based around whether or not creating a game for an existing IP stifled creativity. One of his points that really interested me was how an IP game is grounded within the IP in the same way that any other game can be grounded in reality. They are 'grounded in the reality of fiction’. As my research is currently looking at how much of a player’s reality is brought with them into a game, I asked him if the same applies to IP games, if they have ever used the players knowledge of an IP as a way of guiding them through the game. 

He said that it was a case of yes and no. In the Lego Harry Potter games they used Nearly-Headless-Nick as a form of player initiated guidance, where if they were stuck player could walk up to him and he would float off in the direction they needed to go, an analogue to how he helps Harry and the gang in the books. However, unlike the books, as he was floating off he dropped a trail of lego studs, a collectable item. More traditional guidence methods were also used, such as entrances to new areas being lit up with different colours to make it more visible to the player.

I told him about what I was doing and he gave some great suggestions. He recommended that I did some tests with a 3^rd person camera, to see if the detachment between the player and the character that they were controlling affected the results of previous experiments. He also recommended that I look into how the camera is positioned; how framing the information differently can be used to guide.

With the levels I have been making thus far all using a first person camera, I have become too focused on how the player interacts with the world as if they themselves were part of it. Instead of just looking at how different stimuli effect players, I’m now going to look into is how these stimuli are presented to the player, and how that changes there reaction to them.

In the level with the wooden bridges vs the vine bridges, players only went over the bridges if after they noticed that they were being laid out across gap that they otherwise couldn’t get past. It was the context and position that the vines were in that clued players into their functions. Also, in the ‘dark vs light’ level, the fact that the choice between going down a dark path or a light path was presented as a T-junction, and therefore presented as an explicit choice, might be the reason why so many players decided to go down the dark path, against popular wisdom. Players’ had mentally modeled the environment as a binary choice.

In an average game, choosing between dark or light areas isn’t presented as a choice, but rather as part of a wider more flexible environment. Light is used to literally light the way, and very rarely will a level be designed so that a dark area is a choice with consequence, it’s usually use to indicate dead end space. In this sort of environment Players’ don’t model it as a binary choice.

I think I made a mistake to say that players are drawn to dark as much as they are to light. It’s much more complicated.

Bridges and Materials part 2

I decided to take the idea of material semantics a bit further, and incorporate some new things into it.

I want to see if the general mood and feeling of a level will influence the actions of the player. I took the previous ‘metal and woodenbridges’ level and changed the bridges to be made out of wood and vines. Vines aren’t normally associated with the ability to walk across them in reality, but are in fantasy and adventure stories.

I made two versions of the level, one in a realistic environment and one in a more abstract unrealistic environment.

In the realistic environment, players were generally cautious of the vine bridges at first, but by the time they got to the 2^nd or 3^rd bridge, and noticed that they could be used as bridges, they used them more often, I imagine due to the novelty of it.

However, when in the abstract environment, most of the players ended up being more reluctant to go across the vines. This was unexpected, as I thought that players would see the odd world with the floating rocks and platforms and think that the world didn’t adhere to realistic rules. The level itself was floating in an empty sky, with no visible ground underneath. Maybe it was because of this the players felt unsettled didn’t want to risk falling through the vines.