Tag: GAM710 Week 01

This week (well last week now, but I didn’t have time to type it up!) we were tasked with coming up with a concept for a creative app. The theme or constraint was:

“As long as we have each other, we will never run out of problems”

I am here to make games, so as to further my prospects as a game developer and try to land that Hollywood lifestyle where you’re at a wedding and someone says “Hey, what do you do?” and I get to say “I make games for a living…”. So, a game it is. Conflict is one of the most basic of ‘problems’ that humans have and its because we ‘have each other’. Now, I just need to get to a wedding.

Mindmap

After stabbing, swiping and pinching on that paper I remembered that you have to use a pen. So, I got going. I started out sensible enough, thinking about time and task management. I am a GTD’er and I like routine and systems so I thought maybe I should do something in this area. Then the thought of playing as a 1930’s mobster popped into my head. What would he be doing? Collecting of course. Then led to ‘who is he collecting from?’ and the first thing the came to mind was a barber shop. But I think that the barbers in my imaginary town are tough, I don’t know why but they just are. They would fight back and so I started thinking about how that might happen. I came up with quite a lot really, mostly pretty silly, but I smiled a lot while I was working through it all!

So the story goes that Don L’Oreal is a successful criminal boss. He’s a leader of the L’Oreal crime family and respected (sort of) in the communities where he operates, at least by the people that enjoy romanticising the brutal nature of such organisations. Although there are many problems that face a criminal boss, the one that really gets to L’Oreal is the partial baldness he has. His Wife, Alopecia, reminds him that there are more important things to worry about but he just can’t shake the feeling that he would be a more successful boss with a full mane of hair. Because of this perceived deficiency, he wears a hat and insists that all the mobsters that work for him also wear hats. Well, as time goes on and the mobsters popularity grows in certain parts of the city, the trend for wearing hats catches on.

The Barbers and the Hats

Barbers don’t like hats. Not one bit. They know that the hat covers up the beautiful work that they do but worse than that, the barbers know that the average man will go more than six weeks without a haircut if he knows that he can hide the floppy mess under a hat! When the mobster induced trend of hat donning really starts to bite, they know that they have to do something. They create a plan. Well, they start out with a messy plan and then the shave a little off the sides and … You get the idea.

With the help of Guy’s uncle, Tony, they approach a very well groomed Doctor Dread . L . Lock, who agrees that something drastic is needed. Dr Lock has been trying to perfect a way to transplant an entire buttocks worth of skin to the head in order to cure baldness (providing that the patient has the necessary physical properties on the buttock to start with). They arrange to present the solution to Don L’Oreal so that they might assassinate (get it?) him while the procedure is ongoing. The short version is that they succeed and the Don is dead. But looking pretty good…

Don L’Oriel is now just a handsome framed head-shot in black and white, looking off into the distance. Pouting a touch.

Mobsters are ruthless. They are violent, short-tempered and often cruel. What they are not is stupid. They know immediately the Don has been murdered and they want revenge on Guy, Tony and Dr. Lock. The barbers, knowing that there is now a war coming, arrange the support of the other fibre slicing agents of the city and form The Band of Barbers to combat the ‘growing’ threat. Sorry, I’m in full swing now.

Don’t Games have Players?

Yes, yes they do. It’s about time to talk about some actual gameplay. I must be honest and talk first about what I learned during this exercise. As we did not have very long to perform this task and well, the clue is in the title of task (rapid) I think that I could have used the time a little better to clarify the actions that the player would take a little more. That’s not to say that I dont have some of that, I just know that it’s not as well developed as I would have liked. On the plus side, I like the idea that I have and can see lots of opportunity to grow it (cough). I will share the fact that I found the task quite hard and procrastinated for a day or two. Maybe it was just that others would be looking at the work at a point when it would not be formed enough to be impressive, maybe its that someone told me recently that I’m not ‘creative’. Maybe I just got a bit nervous. Anyway, after a chat with myself in my “Sort Yourself Out Mirror”, I got to work but had burned that time already so I accept that I could have a more fleshed out mechanical section had I got started a little quicker.

I am not really one for mobile gaming I must admit. I like bigger, more complicated games that involve multiple game systems and are pretty to look at and have a hue of realism to them. I think that I would like to explore mobile games a little but I think that I will be more at home making something for PC or console if I’m honest. But, that said, something about this idea seems to say ‘mobile’ to me. Maybe its the silliness of the narrative set up or the fact that the first game mode I thought of was adversarial, I’m not sure.

I’m gonna make you an offer you can’t refuse.

The main strength in this light-hearted set up is that it makes a lot of sense to have the two teams, Mobsters and Barbers, from which the player can choose. I have a few ideas on what would make each team unique too. The Mobsters would be overt in their actions. I thought about the order of actions being something like

1: Player goes to the family home to talk to one of the bosses and receives that contract (hit, intimidation, kidnapping, and so on).

2: Player gets to the location and/or hunts down the mark.

3: Player performs the task.

4: Player returns to the family home and collects reward, money, experience, bonuses and so on. Return to step one.

That’s a pretty solid gameplay loop. But of course, the devil is in the detail and have already, in my short experience as a developer, dealt with a few incomplete GDD’s and am very aware that this would need a lot more focus. But on the other hand, this does already start the conversation about what sort of game this would be. I was thinking open world or at least open sandbox. I like that the player could choose a variety of approaches (heavy-handed vs .22 pistol ‘hit’) There could be interesting hunting mechanics where you get to learn the marks routine so as to chose an approach. There could be a lot here.

Progression could involve upgrading the basics such as weapons and clothing but also more acute hunting skills with prompts or tools onscreen or in-ear that would help you find the right mark. Basic progression and character levelling could be represented as promotions through the family until you were the head, having taken over from the late Don L’Oreal.

Then there is, of course, the Mobster naming machine pictured above that would be easy to code and turn into a mini-game or just a good way to personalise the character for the player.

Just a little off the top sir?

This one I found much more difficult. I thought that the Barbers would have to be covert, sneaky and employ guerrilla tactics if they were to be successful. I see mobsters lured to the chair (a haircut may become a precious treat in the mobsters’ world after all) and dispatched using the cut throat razor or a soaking towel wrapped around the face and held from behind. Dark, I know, for a silly theme such as this. But, I think I like that sort of thing. I like the contrast between the light set up and the potentially brutal gameplay. It would allow a very expressive game for the creator at least. Saints Row has just sprung to mind…

The problem of the basic Barber gameplay loop prompted its own mind map.

I thought that obvious way to do it was to be inspired by (copy) Hitman and use a similar sort of disguise system that would allow the Barber to sneak into a club or restaurant and observe the mark from close up. The player would then have to wait for the Mobster to be sat down and alone so that he could create an impromptu barber chair set up and take out the target there and then. At that point, the player would have to escape and … yes, Hitman.

Then I thought about luring a Mobster to a special Dexter style kill room covered in sheet plastic, although you would have to research what materials were available in the 1930’s first. A quick Google has told me that yes, we could use plastic. In the center of the room is the barber chair and maybe the game is to get information out or the ‘customer’ that could be used to set up the next target before the poor fella expires, or a rescue is attempted. The mechanics that need to be designed would need to answer questions like ‘what are you manipulating to lure him’ and ‘how do you get caught trying to make that happen’. There are many other questions of course.

Along the same sort of lines would be the mobile barber shop where a Mobster would arrange to meet you for a precious haircut (that he would have to say his lady did for him). The player would have to get to the location without being detected by other Mobsters and then you could perform the hit while the guy in the chair. There could be an upgrade and progression system for your mobile kit that would allow the player more choice in how the kill happens and what to do with the body afterwards.

Well, that’s about as far as I got. I would like your feedback and I must say that as silly as the idea is, its mine, don’t copy it, I have put it here for marking and it is not public property. ‘Tis mine.

If you got this far, your dog needs to be let out for a wee.

Having come back to this presentation this morning I think I have landed on the most interesting part so far. I touched on TensorFlow in my last post, having just learned that there is such a thing as a TPU or that they are at least on the way for general use. This is a machine learning library developed by the Brain Team at Google that is available now for free as its open source. I will be looking at this closely, albeit with squinting ‘I don’t really get this’ eyes to start with.

The five things covered in the presentation that can be introduced into games right now…

1: Using analytics to predict stuff.

This leans on collecting data about players interacting with your game and allow the developer to understand how people are using it. Although the presented did not go into detail about this as he had only five minutes on stage, I think that what he is saying is that Tensor (I may not be using the jargon just right I’m afraid) would then be able to predict the nature of the changes that would be most beneficial in moving player behaviour towards some desired result. If I’m right, then it’s kind of like having another developer on hand or at least a researcher! Sounds great and something that I will want to look at during the course and beyond.

2: Experiment with re-trainable deep models.

This is where I stop what I’m doing and really listen. I have heard about this sort of thing just kind of ‘through the ether’ (ha!) and I would really love to know exactly how it works. I saw a great video (although I can’t seem to find it now) on YouTube where a programmer teaches an AI to play Doom but I was trapped in my other life running the clothing business. Time to have another look at this I think!

3: Use neural networks to get some style to your art.

I was very surprised by this. In fact, I had been surprised by quite of a lot of this Lightning talk and felt a little left behind in my old age but more specifically I was surprised at all the areas in which machine learning was being used. The presenter talks about neural style transfer, taking an image that you wish to affect and an image to determine the desired style and then having the AI ‘create’ a hybrid of the two that still makes sense to the human eye. Very impressive and something to be leveraged for more impressive looking game prototyping and later in the development pipeline where art assets may be hard to come by.

4: Dream up new content for your game.

Ah, seems I may have been on the right track with what I said about the system predicting changes developers could make to the game to increase some predefined behaviour. This takes it further talking about how machine learning and TensorFlow can be used to develop characters based on desirable qualities that it has learned and a few other ideas. One that caught my eye, however, was the music prediction system. I have worked on my game Serial Link for a while now and one of the most exciting things in it is the music. We were lucky enough to have an awesome guy, Simon Felix (Shift Quantum, PS4) produce the music for the demo that we hope to release. I created logic that would drive the music based on 3 states, calm, suspicious and engaged and had a right old time working out how to make sure that the music could track what it’s doing, what it needed to do and then waiting for the right time to make the transition. There were multiple challenges and in all honesty it still does not work completely all the time. But, I did very much enjoy doing it. I think that this machine learning approach would be very interesting. The next step for me is to check out Project Magenta from Google and see if there is something that I can do with it.

5: NPC control.

Although the presenter points out that players won’t appreciate complexity for the sake of it, he does go on to talk about some of the areas and NPC AI that could really benefit from machine learning. Serial Link, the game I am working on, has a sidekick character and I wonder if something like this would be able to have her learn how to handle herself in the game and in the end, produce an AI that is more interesting and flexible than a traditional coded behaviour tree and so on. This could be done by having the AI learn from a human player, or multiple human players, that could control the character for some specified training period. I am very interested in this too it seems.

I have to say, this talk has blown my tiny mind. I don’t think that I really appreciated how much emerging technology there really is. Featured in the video…

ATAP Advanced Technology and projects: I love the way that they present themselves to be literally on the edge between what’s possible and what’s impossible. And of course, every time they are victorious is one way or another the space that is possible grows and the impossible shrinks proportionately. Spotlight stories were very interesting for storytelling blurring the line between game and story just a little more. I have developed a little for VR in general with The Prison (where you play as the Wardens embodied wrath and need to terrify the prisoners in your charge) and in the Virtual Military Sim which was the final coursework and the live client brief for the BA Top Up last year. I am keen to see all the different ways in which VR can be used and this caught my eye because of that. One thing that I find a little limiting with VR is the limited locomotion options and as someone that suffers from ‘the most acute motion sickness I have ever seen’ (diving instructor in the Dominican Republic), I can only tolerate teleportation during play sessions. I wonder how much of a factor this is to VR in general although it’s interesting that Google thinks that AR and VR combined from the 4th wave of computing…

Project Soli also caught my eye with its incredibly sensitive gesture reading ability. I think that some blend of VR and this kind of tech would be the start of very high fidelity virtual experiences. It would be great to not have to hold a controller at all and something like Soli be able to read the finger input from a distance although that’s not what the technology is being used for at the moment.

Tango, an AR and VR standard was presented as the cutting edge of that technology. I was very impressed with one point in particular. The presenter explained that due to the software’s ‘area learning’ feature, it would remember where game items had been dropped in the real world. I just could not help myself but think of Breezehome from Skyrim and how it might feel to be able to actually stockpile my gold ingots in my garage!

The gentleman presenting Kubernetes succeeded in making sure that I knew that I need to learn a lot more jargon and that I’m sure that what he was doing was very impressive. I would need to look more deeply at the cloud-based multiplayer backend that he was discussing before I can fully appreciate what was so exciting about it. I will leave that until I am competent (plus a little more) in C++, my teething language of choice.

Skyrim (the second time I have referenced it I know…) is big. The world I mean although the collection of quests is also sizeable. Well, one of the presentations that really got my attention was the Virtual World Building on the Gaming Google Cloud Platform. This would allow a cloud-based, unlimited game world that would carry on engaging with it other occupants until Javaskor (my Nord character) would return. This does bring up a conversation about how such a world would be convincingly populated. Im, sure there are techniques around if I were to dig a little. The other little nugget that I got from this part of the talk was that there is such a thing as a TPU, a Tensor Processing Unit for AI acceleration! This is very exciting as I am very interested in game AI.

For me, this is one of those things that I think that I should be into but my instinct tells me that I don’t want to be. I have never really taken to mobile games and I think that I am going to have to suck it up and immerse myself to cure that. I don’t know what it is that switches me off the platform. Maybe it’s the tiny screen, maybe it’s the awkward controls, maybe I just have not tried enough games! It could also be that I like immersion and like to see myself as the character that I play in the game. While the presenter was talking about the number of hours consumed via YouTube he mentioned that he feels that the makers of Clash Royale have done a great job making the game ‘sticky’. I like that term and describes the desired customer retention perfectly and in one word. Ok, I’m downloading Slither.io and I will let you know what I think. Yeah, it’s not for me.

The last part of his talk was quite relevant to all game developers that would like to encourage people to stream their games, and let’s face it, why would you not? It’s a great compounding advertising strategy and is a genuine way to attract people to the product. After all, they have seen the software running, wort’s and all in plenty of videos and still are happy to invest money in your game? I would take that to mean that the game is solid. The fresh perspective here for me was in thinking that some of the things that streamers find attractive can be engineered into the design of a game giving it a greater chance of finding a home with popular streamers and content creators. The obvious danger is in losing the distinction between tweaking the content for these people and ‘selling out’ to them, ruining some property of your game that was important or worse, necessary.

I think I need to leave it there for tonight as my stinky little (awesome) daughter needs a bath. If you read all this, your tea is cold.

In an effort to fill one or more of the many gaps I have in my game development knowledge, I am reading a book called The Art of Game Development – A Book of Lenses. Although I have worked on a few games now (not enough, not by a long shot) and although all of them have taught me more than I have time to write and you have time to read, I can’t help but recognise the need for a structured set of tools with which to work. I don’t mean a ‘system’ of some other soul-sucking word, but more along the lines of signposts and roughly sketched maps, just something to check myself against. During the course last year (Games Design for Industry BA Top Up) I had the same experience as when I first started to learn to play the guitar. I did not have to be a musician (and never really became one either) to know a bad song when I heard it, but that didn’t translate into the ability to write great songs either. So, I was in this middle ground where I know what I didn’t want and in most cases, I kind of knew what I did want, but I just couldn’t join those two pointers (C++ slip there…) together.

I am trying to solve that problem intentionally now regarding the new skill I desire. Game developer. I think I know which denomination of developer I want to be although I am not sure and in all honesty, it’s still very early in my ‘career’. I’m a programmer. No, I’m a game designer. But I really want to be able to animate too… What about audio? And so I need guidance. And I need it fast as I am no spring turkey. At 38 father of one (more if my wife gets her way) and no real job right now, the least I can do is read up!

The thing I like most about this book so far is nothing to do with game development at all. It’s the pacing. It’s just taking its time and setting the scene, talking about why a book like this should even be cluttering up the automated Amazon warehouse now that all the book shops are closed. Well not all of them…

I am a little over 70 pages in but already very interested in what the author has to say. I think that the way that he has chosen to structure the book, using the tool of a ‘lens’ with which to view your game is very well conceived and allows a structured conversation about game design without reducing the process to a series of steps. I really like the attitude that you could/should use the lenses, and combinations of lenses, to ask the same questions you have been asking already, just in a slightly different light.

My favourite lens so far is the lens of Infinite Inspiration. It is the lesson that some of the most impressive skills or techniques that a practitioner of any discipline has may not have come from watching others practice but from anywhere. 

If you read this far, your pot noodle is overdone and you need to give it a stir…