Yesterday I reached a breakthrough in the simplification of the game. I realized that although having a single player avatar gave players a more human connection to the actions that were occurring in the game, it didn’t add anything to the sound that was being produced, and in fact the sound was not that interesting at all. I came home, and ripped out the player character, and started rebuilding something similar to an old LD game that I did, bad hotel.

The other thing that I ripped out was the audio engine that I was using. Although I love the pugs luv beats sounds, I think they were getting in the way of me imagining anything new. Martin Brinkmann has a really nice set of instruments up on his website, so I dropped one of those in.

Bad Hotel is basically a tree based system, so I am just traversing the nodes out to the leaves and then looping back to the root. This was the first thing that I sonified.

I then added in enemies, to add some entropy to the song making. These enemies are dirt stupid, they just go straight for the root node.

There is nothing in the game itself to keep me from going on forever. However, I am setting certain rules for myself, just to make the song more song like. I think that these rules could be written into game systems eventually. They could take the form of either more health on the individual hotel elements, or a certain amount of cash flow that you need to manage as the game goes on. In addition, some of the enemies could aim for the leaf nodes instead of the root node, and this would guarantee that the song would be pruned back, rather than totally falling apart at certain points.

The other thing that could be done to structure the song better would be to have certain diffusion limits as the triggers travel through the leaves of the tree. That way you would need to add new power sections to different parts of the structure. If only one of these was sonified at a time, it might give rise to verse chorus type structures.

I have been through a number of prototypes for the follow up to pugs luv beats, and I think that I am closing in on a working direction. I think this is the 5th one in a series of wildly different ideas. Where this post starts, we are thinking about doing a tower defense where the towers are step sequencers. We also are thinking about having a main controllable character for setting up the towers, like the engineer in team fortress.

I am going to do this prototype more out in the open than the one for pugs luv beats, at least until someone tells me to stop. Hopefully it will give people an idea of how we work, as well as give me a way to get feedback from a larger group. I will be updating this post as often as I can with new stuff.

Notes 14-3

Bloons tower defense 3 does not have any health bar style readouts on the state of the enemies. It displays the states of the enemies as distinct graphics. On the first 20 levels, I only saw 6 states, four of which mapped directly to health states, and two of which mapped to special states. One of these was an enemy that contained multiple enemies. The other one was an enemy that could only be pushed into a killable mode through a special attack type.

Revenge of the titans survival mode only had 6 tower types. On further review, this is controlled by the main game tech tree. The six that were opened up were due to the way that I had approached the game. A resource collector, and a tower that buffed it. It had three types of attack towers, but these were effectively all the same, just speed boosts of the base tower. The last tower was a healer for the attack towers. It also had a bunch of one time use pickups, so a bunch of the game was trying to find these on the screen. I think that once the game went past 7.5 mins or so, it stopped doing highlights of these as they random dropped, which made grabbing them a ton harder. The enemies didn’t have displayed health, but you could guess how many hits they were going to take through their general size. Also the size of the crowd was a general display of the amount of health that you were going to have to deal with. The large enemies had a health display that only showed while they were under attack. The same for your towers, although highlighting them seemed to display their health.

The prototype got this far today:

Notes 16-3

I started the prototype at this point today:

I added the enemies back in, and corresponding lose states. The enemies are a bit dumb right now, they just spawn a certain distance at a random angle from the players base, and move towards it. I think that at some point they should probably move towards towers that are attacking them, and try to take out the players towers. The enemies also drop money that you can use to build new towers, or upgrade your existing towers. I haven’t played it to an endgame yet because of some bugs, but that is my first plan of the day.

One of the biggest issues with this prototype right now is how confusing it is to look at. I think that by limiting the expansion of the towers so that they cannot overlap, it should create nice textures on the screen as well as removing some of the visual confusion.

Also, you may notice that I have a virtual control stick in the bottom left, and two big virtual buttons. I really would like to get rid of both of those things. I may try out the bit pilot controls for the moving the character, but I am concerned that they won’t map well to a game where you want your player to end at a particular spot, rather than vacate their current position in a specific vector. This may just be me waffling over details though.

I am currently using just blinking graphics to signify audio, but I noticed a few issues that may mean really boring audio. It seems that an optimal strategy may mean that the notes that are playing from the towers will just be a totally steady drum beat. I am not sure how to deal with that, but perhaps putting in actual audio may help with discovering a cure. Will probably use the pugs luv beats sound engine as I have been on all of the prototypes to try to figure it out.

Other than those two big issues, I want to add context sensitive towers, so that if you build a tower close to an on screen resource, then it harvests that resource. That may be adding too much complexity, but the only way to find out is to add it unfortunately.

Notes 17-3

I added the wave on / wave off mechanics that are present in so many tower defense games. The reason I added that was because the games were going by way too fast. Now that the pace of the game is slowed down, you can start to notice some of the issues. I have also added actual audio, and that is becoming apparent that the way the towers charge up does not make for quite so interesting audio. I have two ideas for how to fix these issues.

The first one is to cut out the main avatar. That fixes two problems, one that there are control issues that I have not faced yet, and a simple place towers directly is something that I can handle much more easily. The second is that I may be ignoring what is fun about tower defense, and making a different kind of game. Tower defense is at its core a lightweight management sim. I think what I am making is more about avoidance. I think that by focusing in on just one type of fun, rather than trying to have two types of fun simultaneously, I will be able to make something better.

As for the audio component. I like the towers being trigged by the enemies, it adds some computed variety to the sounds that they produce. I don’t like the fact that the sound outside of that is such a droning rhythm. One of the nice things about offline sequencing is that you can make really complex rhythms really easily. With this game, it seems that you are forced into making really simplistic sounds for the gameplay to work.

We took one of the things that people liked most from pugs luv beats, and spun it out into its own app. This was a pretty quick and fun project. It is a bit like a more whimsical version of bebot, an app that we all really like.

Pugs Synth on itunes store

Making games has been a long term goal of mine for a few years now, probably since I outgrew astronaut as a life goal. This year I really started making games in earnest. While I had made about 4 games prior to this year, I didn’t really have the drive or the understanding to do it. A bunch of things came together this year though, the most important of which was the funding of Lucky Frame by Channel 4 and Creative Scotland. That allowed me to stop doing my day job of making websites and dedicate myself to making games full time. What a year it has been too.

The most recent thing is that Pugs Luv Beats was nominated for an IGF award. I had entered the year with the goal of having a single game that would be worth submitting to the IGF at all, I had zero expectations of anything more than that. To be clear, I did not have any game that was even worth submitting last year.

It turns out that over the course of this year, I have made 10 things that qualified as video games. I also made a card game, which may have helped a little bit in my understanding of games, but I am not really sure. This is a super far stretch from 2010 where made 4 games, only 2 of which were arguably playable. And an even further stretch from the other 27 years of my life where I made 1 video game.

TEN SECOND FIRE JUMP

This is the third Glorious Trainwreck that I had done. Glorious trainwrecks is where you make a game in 2 hours, so it is a slightly crazier jam then the Ludum Dare, but good for testing out ideas. For this one I decided to learn flixel. Also I had not even tested out the tool chain, so I had to get that all working within the 2 hours. This game kinda sucks. But Flixel turned out to be totally awesome, and I ended up using it on a few more projects throughout the year.

I actually ended up cloning big part of flixels engine into the iphone engine that I am using, because it is so enjoyable to work with. This may also be because it is the first real game engine that I have worked with, so I didn’t know anything else. I suspect it is because flixel is really great.

Direct link to SWF - arrows to move, space to jump

Sound Track

During the month of March I started working on Lucky Frame all day, and made a few prototypes that were not released. Sound Track is probably the most complete of these.

I had been playing a bunch of trainyard at the time, and wanted to make a game that accentuated the rhythmic aspects of that game. Yann made about 10 levels for it, and we user tested it a few times before realizing that it was far to complicated for people to control. I think that we didn’t fully understand user testing at that point, nor did we understand our market, so we may have been more harsh on this prototype than we should have been.

However, it was good that we scrapped it, because the toy at the core wasn’t that fun to actually play with. It felt more like a chore to build songs by drawing roads.

This one doesn’t have a link because I was not diligent about saving working builds back in this time period. Also it is for iPhone, and I neglected to port to PC.

Bad Hotel

My second Ludum Dare entry, this one is a literal take on the genre “tower defense”.

Boy, does it have some problems. I made the controls far too losey goosey, you have to hit buttons at the exact right times to place pieces. There is a strange cool down period on the placement that has nothing to do with the actual game play, and feels really divorced from it. There is no difficulty curve, and it starts out pretty darn hard, so you don’t really get a chance to practice your skills before being bombarded by enemies.

All that said, the core toy is pretty fun to play with. Basically it is a human controlled Diffusion Limited Aggregation system, so you can make super cool looking structures super quick.

Sean, the artist at Lucky Frame keeps bugging me about this one, so it will probably have a sequel at some point. One of the things that I want to draw from is Rampart, which was pointed out to me by Tony as a pretty similar concept.

Direct link to SWF

Unislot

This one started out as a text adventure slot machine, which I did in 20 minutes as a joke. When I saw how sucked into it Yann got, I decided to add graphics to it. Still was basically a joke.

I ended up putting it in the IGF Pirate kart, so a bunch of people have played it in the context of other strange games, so that is good that there is a context for this type of thing. The Pirate Kart ended getting a bunch of news coverage, which is strange and cool. I think I am going to do more of those if I have the time.

I still think that slot machines are a great model for thinking about simple game play. Mining in minecraft is basically a big slot machine. Every time you break through a block, it is like pulling the handle.

Play on Kongregate

Droopy Lasers

Did this one with Alex and Yann in 6 hours for a dundee game jam. I didn’t do too much of the gameplay code because I was handling the bridging between the pitch tracker and the game engine. Unfortunately because of the crazy setup we used, it is pretty hard to setup. I think that it would be possible to rebuild it, but I am not sure why you would.

I also did the character art for this one. It is “OK”.

Working with Alex was awesome, it was great to have another programmer on a project. A++++ Would work that way again.

Demo Video

Martin vs Monty

A game for Ludum Dare, made with Lucky Frame. This one is half vvvvvv, and half super crate box, but not as successful as either of those by a long shot. I burned half of the compo time trying to get the collision code correct, and ended up not tweaking the gameplay enough. I should probably revisit this one now that the collision code is solid.

Also, Super Crate Box for iOS has convinced me that platformers can work on the iphone. I may revisit this one, if for no other reason that platformers are a well accepted genre, and one of the things that people like to cut their teeth on when learning level design.

OSX Download link and instructions

Pyoing

After the stress banging my head against collision detection for MvM, I decided to take it easy and do something super simple. I ended up making a clone of crossfire for iPhone. This one didn’t require much work, but turned out pretty great. There are still some rough edges, but I don’t know how much I want to fix them. We should probably make this free.

One thing to learn from this one, it is that games where you play against another player are much easier to make engaging then games that you play by yourself.

iTunes store Link

Some Faces

This game is an object lesson in not getting bogged down in systems. After spending a week or so writing a parser for BulletML, I decided to make something with it. BulletML is a programming language for programming bullets, and is used in some of my favorite shmups. It turns out that using the same systems as your heros is not the best way to make the same things as your heros.

I would like to actually make a shmup at some point in my life, but I need to actually spend some time trying to understand them first. This thing I banged out in a few hours, and it made me realize that shmups are far more complex then they seem. I am probably not going to attempt one for quite some time.

Glorious Trainwrecks Link

Rocket Gun

This one is the final outcome of an idea that I had been thinking about making for over a year. Unsurprisingly, it is not very good, as many over worked ideas turn out to be.

The initial plan was to make a game called rocket gun, where you control a gun with rockets attached to it, and have to control your movement either by shooting or firing your rockets. I made a prototype for that, and it turned out to be impossible to control. I didn’t even get as far as making levels for it.

I revisited the idea, but took out the shooting, and just made a game where you are an oblong object controlled by rockets. Turns out that this is still super hard to control.

If I revisit the idea, I think I may continue to lower the complexity of the control system, because that seems to be the real issue here.

Glorious Trainwrecks Link

Pugs Luv Beats

Not going to say much about this here other then the gameplay is an ongoing revision, partially because we really care about this one, and partially because the gameplay isn’t the best it could be yet. It is interesting to change gameplay on games that are already “finished”, but I think it will be an important lesson.

However, as a toy, it is super fun to play with and make music with. I think that fulfills where we wanted to go with it. Any further things that happened with this game are amazing. I am so excited and humbled that it got into IGF.

Website

I will probably do a bigger post on the concept and development and such, but in short, this is a music composition game. The difference is that most music games are performance games, and as such focus on the creative pursuit of generally playing other peoples music. We tried to make something that is more about creating your own music.

It was an intensly collaborative project with my friends at Lucky Frame . And at 6 months in prototyping and development, the longest game project that I have done by a factor of 10^4.

Pugs Luv Beats on itunes store
Pugs Luv Beats Website
Nice writeup on Create Digital Music