• Added an AI – You can now play the computer
• Added new simpler boards & game types
• Added a new “Granite” skin
• Improved help system
• Improved interface (made buttons larger etc.)

Ipad Release 1.1 (Multi Board & Training) – ETA Aug 30, ’10

One of the things that has become very apparent is that Swarm can be difficult to learn for some players. We will be adding new board layouts to Swarm, with smaller boards that are easier to get started on. We will also be adding a step through training system where rather than “tell” you how to play, you will be able to learn by completing small puzzles as you progress through training screens. Each screen will be specific to the task at hand (such as “Capture a Drone”) and will also make use of the same framework that we will be using to create multiple board types.

Sebastien will also be working on improvements to the Swarm AI (Cerebro) so expect Cerebro to be a little smarter by this release.

Ipad Release 1.1 (Cerebro AI) – ETA Aug 30, ’10

You will be able to “Play The Computer”.

Ipad Release 1.2 (Network Play) – ETA Oct ’10

It goes without saying that being able to play other humans across the network will be a big win for Swarm.

Back in May 2006 when I first met my (now) wife I thought it would be fun to invent a board game.

I remember some of my best days were spent in the Dome Cafe in Hampstead London age 17 (1988) playing Backgammon with close friends, bartenders, and other clientele. It all felt very Greek and artisan. That stuck with me, and I think I’ve always had a hankering to create something that brings that kind of mental stimulation and fun into the world.

Hence Swarm SG. Swarm SG went through many iterations, board shapes, rule changes and playing games with my (very patient) wife before I got the courage to go to a meeting with other people in London who liked to invent board games. (Yes, there are groups all over the world of people who gather on a weekly basis with the sole purpose of discussing the board games they invent).

At the one meeting I went to I played the game with a veteran board game inventor who seemed like a very intelligent chap. He didn’t “get” it and was very clear about the fact. I left the meeting feeling like a bit of a fake. Here were these folks who met every week to discuss games they invented and here was I – a first time board game inventor – expecting to have a hit instantly. So, I shelved Swarm, promising myself I would pick it up at a later date.

Birth of Swarm SG

Fast forward 4 years, a lot of Javascript, a successful tech. podcast, & Pluggio under my belt and Apple releases the iPad. As soon as I saw Steve’s original presentation of the iPad I “knew ” that at some point I would have to find a way to get Swarm SG onto it. But it wasn’t until episode 36 of Techzing when we interviewed Jeff Haynie about Titanium Appcelerator that I realized I could build Swarm by myself with my coding skill-set of JavaScript/CSS/jQuery etc. In fact if you listen to episode 36 you can literally hear myself and Jason’s jaws drop as we ask questions of Jeff and uncover what Titanium can do.

It took me about a month to build a working version of Swarm SG in the web browser with jQuery & CSS. I didn’t think about the iPad during the time I built it (other than the width & height dimensions). Then after a month I decided to tackle moving it from the browser to the iPad.

First thing I noticed was that my original piece movement wasn’t very nice on the iPad. The main reason for this was that finger pressing and mouse usage are very different. The “mouse” way of doing things just didn’t work on the iPad. I spent about a week experimenting with all sorts of variations to find a method that would “feel” nice on the iPad and finally settled on the idea where as you drag your finger it doesn’t move the piece but it highlights the square under your finger, then you let go and the piece obediently moves to the square.

Also, on the iPad there were all sorts of interesting issues related to multi-touch. My game does not expect multi-touch, but multi touch caused all sort of weirdness none the less. So I had to figure out way of trapping “mistake” touches and NOT acting on them.

Another interesting thing was that jquery click binding was sloowwwww on the iPad. I’m not sure it’s jQuery’s fault I think it’s general web binding that is the issue. It would literally take .25 to .5 of a second for a piece to register being clicked. I discovered that if you use apples touchstart and touchend it speeds things up to the point where they feel like it’s native code.

Finally, another little thing I noticed was that apple have built in an algorithm that changes the x/y co-ordinates of the touch position to skew it to make sense when looking at the ipad from a normal angle. This meant that for the player looking at the topside of the board (i.e upside down) the touch point was different to the one expected! So I had to tweak the jQuery.touch library to work “upside-down” as well as the correct way round.

I should also mention that when I very first moved Swarm from the web environment to my first build on iPad took about 15 minutes with the help of the amazing Titanium. All I needed to do was open a webview (using the Titanum API) and load in my original HTML/javascript. Then titanium deploys the app to iPad. When you open the app you are opening it in web-kit.

The Titanium API has a way that your JavaScript can talk to the API. I used this to save the game state. Actually each time you make a move in Swarm it saves the entire App state as a JSON string. Which is quite impressive because it doesn’t interrupt game play, or animation, and it’s a big old JSON data set that it’s passing in there.

The App Store

One of the things that you don’t hear very often is just how much effort is required to start developing iPad/iPhone apps from the Apple provisioning perspective! I’m not going to get into to much detail but will say that it is a similar process to setting up an SSL certificate for a web-server. If you you’re like me (i.e a noob) and are thinking of building an iPhone/iPad app I would set aside one-two weeks I would consider setting aside a decent amount of time just to navigate the certification process, and setting up all the legal stuff that you need to have set-up just to be able to sell games (rather than give them away for free).

Actually, it is so much easier to give them a way for free that I wonder if this isn’t a small part of Apple’s comditization strategy.

Another nice feature that Titanium offers is that it will package your final product into a format that the app store expects as an upload format. I uploaded Swarm to the App store for the very first time on June 10, 2010 23:34 the app was immediately placed into the “Waiting For Review” status. There it remained until June 17, 2010 08:48 when it was changed to “In Review”. So seven days to get noticed. Then, Apple put it live in the App store @ June 17, 2010 20:44.

I was surprised that they approved it first time round, because most developers I’ve spoken to say that it was rejected a couple of times and they had to alter things to meet the Standards. The only thing I would say about that was that I paid special attention to ensure that Swarm SG would retain state no matter what a user had just done – and I also spent about one week trying to break the game in as many possible ways as I could.

Pricing

Pricing is an interesting beast. We’ve discussed pricing a number of times on Techzing. Jason is of the opinion that I should have launched with a free version and then also have an upgrade version. I think this is sound advice but I didn’t really have the time or inclination to create a DRM. So being lazy I just came up with a price for one version (for the moment).

Originally I thought I would charge a very high price (something like $49.99) in order to differentiate Swarm from the crowd and also to convey the level of quality that I believe Swarm to have. That was before I had a very thorough look through the app store and realized that $49.99 would be laughed off the shelves…

Hmm. The thing is I want to convey the message that Swarm SG is not a throwaway game (in fact Swarm SG is much deeper than Backgammon or Chess [in terms of search space] and I believe that it could even take years a very long time to master) so what I’ve been looking for in the app store is what are the acceptable high price points.

I also feel it’s a question of the amount of work that’s gone into the software. So, on some level it doesn’t feel right for me to charge as much as a team of 20 developers who’ve created a highly graphical role play adventure and invested 1 million into the game…

I note that those types of games go for around $14.99 in the app store.

So. I don’t want the Swarm price-point to be percieved as being to full of itself (say $19.99)… but I also don’t want it to be too low that it will be percieved as a flash in the pan.

And that’s how I’ve come up with my target price of $9.99 (the same as, say, Scrabble).

However, I also don’t think that people should buy an unproven game/app for that price. So, in the early days I’ve decided to sell Swarm at $4.99. I think after there has been 1000 sales and some decent feedback/fixes/upgrades that I will bump it up to the $9.99 that I hope to sell it at.

Stats

One thing to remember about these stats is that Swarm is only available on the iPad so the market is much smaller than iPhone. I think iPad is about 3m now.

Day 1 – 9 Sales
Day 2 – 2 Sales (3 ratings)
Day 3 – 2 Sales (2 reviews – 6 ratings)
Day 4 – 1 Sales (3 reviews – 7 ratings)
Day 5 – 4 Sales
Day 6 – 27 Sales (5 reviews – 10 ratings)
Day 7 – 59 Sales (Front page of iPad Games in New & Notworthy)
Day 8 – 68 Sales

For me these stats and reviews are amazing. More than I could have hoped for.

Clearly the turning point was the ratings and reviews. When it hits a certain number of ratings/reviews the app is promoted onto the new and noteworthy page and from there there is definitely a sales bump.

How to get the good reviews? That I can’t answer. All I can say is that in Swarm’s case it’s a new concept and I’ve paid a lot of attention to making the user interface and graphics as professional as possible. I also went through lots of bug testing and playing with it to ensure it all “felt” right and worked as I expected. I also compared Swarm SG to a number of other board games and apps available and made sure that Swarm didn’t fall short (as much as possible).

Where to go from here

Swarm SG still has a long way to go. It needs an AI. I don’t even know how to make an AI for it. One thing that Ken suggested (a friend of Techzing) is that I should have a competition to see which team/individual can make the best AI for Swarm and share 10% of revenue with them. I don’t think it would be very easy to make an AI for Swarm because the search space is just about infinite (well very, very, large to say the least)!

The one thing I regret is that there isn’t a very good feedback mechanism in Swarm SG. In my other two ventures Pluggio and Techzing the community has played a huge part in making them successful.

So, that’s what I will be thinking about next. How to build community into Swarm so that I can get the feedback I need in order to build better features and game play… and also how to convince some people who are cleverer than me to help build an AI.

Update 1

In case it wasn’t clear, his game isn’t just a web app that lives on the home screen as a bookmark. It uses the Titanium Appcelerator, which is a thin native wrapper around a Webkit view. The app logic is written in JavaScript, rendering is done using HTML/CSS and/or SVG or so. FWIW, this actually obeys the arbitrary App Store rules: C, Objective-C, C++ and JavaScript (if run in Apple’s supplied runtime) are allowed.

(Thx @pmjordan on hacker news for this clarification)

Update 2

Some folks on hacker news have balked at the claim that Swarm SG is “deeper than Backgammon & Chess”…

I just wanted to point out that Swarm SG is deeper because there are more possibilities in the game play. There are 84 squares. Each player has 11 pieces that can move in any direction. Players can move their opponents pieces as well as their own (effectively meaning each player has 22 pieces at their disposal per turn). Each piece can move in any of six directions. There are two ways to win. So, mathematically at least, it’s probably deeper.