Monday, May 2, 2011

Scribble Worm iPad Post-Mortem

Scribble Worm is our new puzzle game for iPad hitting stores any minute now. It's also our second game after making Big Mountain Snowboarding for iPhone and porting it to Android, iPad, and Mac. It started as an experiment to try to make a small game and get it out quickly, but even small games seem to be much bigger than first expected. It ended up taking us a little over 5 months, and we still have to do a small screen version for iPhone and Android, and a version for Mac.

The Game:

You draw a worm with your finger and it follows your pattern to try to reach the apple. We throw various obstacles in the way such as a wall that worms bounce off of and a wall that kills the worm when touched. We then animate the walls around the screen for an extra challenge.

I've never seen a game like this before, so it's definitely not a "yet another" game. It's not a jumper, an angry animals, a flight control, or a match 3. If we can get people to try it out I think a lot of people are going to like it.

Art Style:

The art is a combination of procedural effects and hand drawn "scribble" images done by myself inside ArtStudio for iPad. It looks a lot like a doodle game, but there's actually a lot of stuff going on behind the scenes to get the crayon and paper look. Neither of us is particularly good at making 3D models without outsourcing, so it seemed like a good approach that we could pull off well.

I'm pretty happy with the end result, but it's going to be terribly hard to stand out among the millions of doodle games out there. I think it's several steps above Doodle Jump when in motion, but it's hard to sell the differences in screenshots. Our next title might have to be more visually unique in order to stand out.

Music and Sound Effects:

All of the sound in this game is created on the iPad. I wrote two songs in Garage Band for the main menu and in-game music. Actually I wrote about a dozen songs while trying to make some cutesy music, but everything kept turning out as inappropriate death metal. The sound effects are also recorded in Garage Band using the sampler feature. The eat apple sound is actually me chomping on a carrot with an iPhone microphone next to my mouth.

Puzzle Design:

Making the actual puzzles turned out to be the single biggest chunk of work in the game. It seems obvious now that making puzzles is going to be a huge part of making a puzzle game, but I underestimated it by several orders of magnitude. Fortunately Katie picked up this ball and ran with it.

We have an in-game editor that lets you drag and drop walls, and then try it out in real time and make adjustments. This WYSIWYG interface was infinitely better than trying to edit the XML files and reload.

Play Testing:

The first line of defense in play testing is yourself and then your friends, but your friends aren't going to be likely to tell you that something sucks. We're fortunate to have the Boston Indies group that meets up nearby to try out each others games. We actually went through at least 3 revisions of show it to lots of people, then tune, then show it again. It really helped us out to both reinforce that we have something cool, and to see where people struggle or do things we don't like. Always try to find some people who don't like you to try out your game before releasing.

Our first version of the game was entirely black and white, trying to look like pencil on paper. One of the big things to come out of play testing was we had to add more color to the game. People were identifying the worms as various bad things. The crayon scribble shader was born and I think the game looks much more appealing now.

Another thing to come out of play testing was we had to fix worm spamming. After a few levels people would stop trying to solve the puzzles and instead just try to make as many worms as possible until one randomly reached the apple. While this may be entertaining for a short period there's not as much gratification as actually figuring out how to solve the puzzles. Our first attempt to fix this had the number of worms limited. If there were already several worms in the world you couldn't spawn a new one. This could get frustrating if worms got stuck anywhere, so we changed it to spawning a new worm removes the oldest worm from the game.


It feels like whispering into a hurricane sometimes to try to get a game noticed. The most effective ways of marketing an iOS game are picking a good name, having a good logo, and having good screenshots. All of these are free. You then hope and pray and try to find ways to stack the odds of making it into the top charts or getting featured, and even then your game could disappear quickly. Ads on big sites run at around $600/month with a waiting list, and seem to have greatly limited effectiveness.

We did make a video as you can see in the section about the game itself. Some of the art and UI is slightly outdated now, but overall I'm pretty happy with it. However, at the time of this writing we only have 28 views of the video! Some of this might be due to embedded views, but youtube by itself is not a great marketing tool. I do plan on linking to this video or a newer one from the app description though.

Twitter is something new that I've been trying out (@DJWGoldenHammer). It's fun, but so far has not been very useful. I have 134 followers, but I think a lot of them are just following me so that I'll follow them, and don't actually care about anything I'm tweeting. It hasn't been a great place for marketing, but it has turned into a really good way for me to find other dev blogs to read.

We have a press release ready to go on Hopefully that will help some, but it's a new thing that we have not tried before. I must admit that getting on any of the big sites is a mystery to me. Even after close to a million copies and steady sales for two years Big Mountain Snowboarding has just been a blip in the blogosphere.

Our ace in the hole is the free version of Big Mountain Snowboarding, which gets 1000-2000 iOS downloads per day. We can use this as an advertising platform for Scribble Worm. The problem right now though is adMob does not support house ads for iPad and iAds does not support house ads at all. We'll run house ads through adMob, but we might need to make a new version of BMS to really push the Scribble Worm ads to the right people. The danger is that BMS Lite currently has a really solid 3.5 star average with over 200 ratings for the current version that I'm a little afraid of jeopardizing with a new version. If all else fails we can put some ads in BMS full and do a free for a day promotion.

For the screenshots, we're trying out the Chillingo approach of having images with extra taglines on them to try to sell it better.

Why just iPad?

I love to talk about how we have a cross platform engine and want to be on every device in the universe, and yet the first release of Scribble Worm is only on iPad. This was mainly in order to limit the amount of work so we can have a high quality focused version. The iPad is a dream to develop for. It has exactly one screen resolution and aspect ratio, all of the hardware supports shaders, and the dev tools are pretty good. To be on iPhone we need to make new versions of most of our UIs for the different aspect ratio, and make a fallback version of the graphics for older devices. The Android port is even more work on top of that. We're planning for OSX as our second release because we don't have to redo most of the UI.

A slightly sneakier reason is that the iPad market is much smaller than iPhone. There are about a dozen new iPad releases per day, compared with more than a hundred on iPhone. If we can manage to get any publicity off the iPad or Mac versions, then maybe it will make the iPhone release a little easier. We're planning on converting the iPad version to universal once we make the iPhone version.

What went right:

It's not a proper post-mortem without a what went right and what went wrong section!

  1. The initial idea was a solid one. I'm very happy with the end result in terms of gameplay.
  2. We made good use of our existing tech. Every project has us expanding the engine, but much of what we developed for Big Mountain Snowboarding and OverPowered applied to this game as well!
  3. Boston Indies was great for play testing, and really helped us refine the game and the look.
  4. Being on iPad exclusively really let us work faster by only having to worry about one set of variables.
  5. Between ArtStudio, Garage Band, Pixelmator, and XCode all of our tools were cheap and awesome.
  6. Somehow we managed to nail the cute factor of the worms. Adding bright colors to the worms and the obstacles seems to have helped a lot.
  7. Having a level editor was essential for making this game.

What went wrong:

  1. Underestimating the size of the game. I've had to revise my definition of a "small game" to not include anything that has proper level design.
  2. Marketing black hole. You probably have not heard of this game unless you know one of us.
  3. Dropping a video on youtube does not seem like a good way to get people to actually watch it!
  4. While the art style may look good and even better in motion, it may end up making it harder for us to get noticed than something more unique.

What we still don't know:

  1. Are our name, screenshots, and icon good enough?
  2. Will linking to a video in the app description help at all?
  3. Is our staggered release plan actually a good one?
  4. Can we get Apple to notice us and show us some love?
  5. What effect will a press release have?