Hacker News new | past | comments | ask | show | jobs | submit login
Canabalt: iPhone Game (canabalt.com)
99 points by staunch on Nov 20, 2009 | hide | past | favorite | 46 comments



hey hackers, matthew pointed me here and I thought I'd share some stats (more about the Flash version than the iPhone version but yea)

  Days spent making Canabalt: 5 (long ones)
  Hits on my server: ~2 million
  Hits on other major portals: ~2 million
  Estimated hits on crappy portals: ~4 million+?
  Hits on my server from twitter: 50k+
  Best Referrer: Stumbleupon
  Money from donations: ~$500
  iPhone sales: can't say but good enough!
  Top App Store Ranking: 37
  Bandwidth Bills: $1000+
  Top legit score I've ever seen: ~30km
  My top score: ~8km
  Resolution: 480px x 160px
  Initial Piracy Rate: 12%
  Current Piracy Rate: 25%
Inspirations: Half Life 2, District 9, Banlieue 13, Another World/Flashback/Prince of Persia, War of the Worlds (book, not movie)

Trivia: I never finished drawing the space ship that flies by since it is barely on-screen for more than a few minutes. Sadly, multiple screenshots have captured it in all its under-realized sloppiness.


Those donations were well-spent:

http://www.flickr.com/photos/flashbangstudios/4069363572/

(I'm the goon in the top-left of the frame)


"Best Referrer: Stumbleupon"

What's this based on? I've tried StumbleUpon ads before. The bounce rate is insane. Approx 99.5%. It's basically useless for anything other than real-world load testing.

BTW, Reddit's new self-serve ad system is awesome. Dirt cheap, quality traffic, commenting on the ads to talk with your customers directly, and general Internet Bonus Points™.


we didn't do any ads with anybody


Can you talk about revenue that came from the flash version? millions of hits seems like it should give some nice ad revenue. And did you consider seeking sponsorship for the flash version?


I have one request: the little guy should roll like a paratrooper when he comes down from his jump while moving fast.


iPhone has piracy? Is this from Jailbroken phones? Why would someone pirate something worth $3?


Yes. Yes. Because they are cheap bastards who won't pay for anything unless they absolutely cannot get away with stealing it.


Which is why it's ridiculous for developers to worry about it, those 25% would never have bought the game in the first place, just pretend they don't exist. Do you know how much effort it takes to pirate an iPhone app? First you have to find out that pirated binaries even exist, then you have to jailbreak, then you have to fight through mountains of spam and scams until you find the right communities, then you have to spend time learning the technical tricks, then you have to hunt down the pirated app binaries. These people are determined not to pay, they aren't your customers.


Now that I think of it, they're probably in other cultures/economies, where frugality is much more noble and $3 could be a reasonable amount to go through some trouble to avoid it. I imagine Chinese or Russian programmers would take pride in going through some effort to save $3.

So you're right: they're not your audience to begin with.


I agree that those people are not going to buy the game no matter what. What I think might be worrisome is the bandwidth charges that the pirated iPhone apps could use up if they are contacting his server for anything. That adds up. I would imagine that the lions share of the $1000 Bandwidth bill is for hosting the flash game but in other cases the bandwidth bill could be substantial if he is serving images or videos from the phone.

I think that it would be ok for an app to verify that the app is not pirated and then lock the pirated people out of the expensive use case.


It's worth noting that the normal piracy rate for iPhon games is similar to normal software from what I've heard anyways 90%+), and I definitely don't view our lower rate as havig anything to do with our sales!


Some are unable to pay (app store i18n, etc.)


wow only 5 days making the game? Does that include programming, artwork, music all from scratch?


My friend Danny made the music but otherwise all programming, artwork and original sound engineering/foley was done by me in basically two serious weekends ( one in matthew's office). But that's why I made flixel! I made space invaders in two hours this afternoon :)


Just had a look at flixel, and... well, I'd like to have your babies. Unfortunately medical science isn't quite up to it.


so 5 days to create using a sophisticated framework you created over months (years?) - that makes more sense.

I'm amazed by your team's work. you're now bookmarked and on my 'check back often' list ;)

i'm looking forward to checking out flixel when I have some time.

keep up the great work.



Does anyone have a theory on what makes this game so successful? I went through a phase where I was addicted to it myself - and yet it's so mind-numbingly simple!


I've been viewing games through the prism of how many "almosts" they have lately. Canabalt is fun because you almost fall off dozens of times before you actually do (and when it happens you usually almost made it). Because you think your death was avoidable, and only barely occurred, you immediately want to try again.

Something almost happening is surprisingly compelling. Slot machines are a very pure implementation of this idea; if they just showed you all three symbols immediately you would never have that sense of something almost happening...


I find this comment remarkably insightful. It's such a simple concept and yet that idea never occurred to me - but yet it just feels right! The slot machine example is dead on. Awesome.


Thanks! It first occurred to me when we were trying to figure out why our most popular game, Off-Road Velociraptor Safari, was our most popular game. I realized that it's essentially an "almost" generator; the game is always teasing you with cool possibilities (you'll jump your vehicle off a cliff and almost hit a raptor, and then later realize you can actually do this and the game will reward you for it). "Almost" is a pretty common trait in fun/addictive designs if you start looking for it...


You're continually testing your skills with steadily-increasing difficulty. You have an inherent yet self-imposed objective (to beat your own score or someone else's). It requires high concentration but allows you to run on autopilot. It's textbook Flow Theory.

See:

http://en.wikipedia.org/wiki/Flow_(psychology)

Also Raph Koster's book A Theory of Fun


The random procedural map must be a factor; it's like gambling.


That makes me think of the purpose of dice in many boardgames. In particular, Settlers of Catan - which is a great blend of skill and chance.


Dead-simple gameplay, retro 8-bit style graphics, nice grey-scale visuals, awesome music. There are no instructions (on the iPhone version), so the discovery of how to play is also fun (helps that this phase doesn't last long, only one thing to figure out).


I love the detail & variety of the backgrounds, & touches like the flying birds.


Ok, this game is great. It has fitting music, it's instantly addictive, and it raises your blood pressure.

I would have liked the difficulty level to remain low for a longer period of time. Also, it would be nice if there were other obstacles. After about 20 seconds of game play, it got pretty damn hard (read fast), and I wasn't able to memorize landmarks, which is crucial in fast pace, side scrolling, platformers.


The levels are randomly generated, so don't bother memorizing them.


It slows down if you run into the boxes on the rooftops. Use them to your advantage.


But if you slow down too much, I think it becomes impossible to jump some of the gaps.


Not in my experience. Since the maps are generated on-the-fly, I think the gaps between buildings are dependent upon your current velocity. I am not 100% sure, though.


My DropZap game has a high rating and great reviews in the US app store but it's not popular yet:

http://itunes.apple.com/us/app/dropzap/id338490375?mt=8

How should I market it?

(Demo video: http://www.youtube.com/watch?v=1ub6gCU03iE )


Canabalt's free Flash version has had several million plays, the majority of which have been on Adam's own site. The Flash version's success prompted the iPhone port, and also gave him an existing audience willing to buy the game.


Your fundamental game mechanic seems fine (though maybe a little derivative, as I'm sure you're aware). I think the major problem is that the app doesn't feel as smooth as drop7, etc. The winning games have a tremendous sense of style. With a few relatively minor changes, I think the game would be a lot more attractive.

1) Get a more pleasing set of sounds. The high-pitched screeching you are going with currently when blocks blow up is torture for me. I like the sound when you die, however.

2) Make a game-mode for the beginner where it is easy to figure out the game mechanic. I died within a few rounds each time I tried. (Start with all little red block or whatever...) Make this auto-start the first time the game is opened.

3) Moves are too clunky. The blocks move to the side of the screen too slowly after the user presses a button. I would make this 3-4 times faster, or just eliminate it. Blocks also maybe fall a bit too slowly-- I would eliminate the pause before they fall.

4) There is too much visual clutter on the screen. Its nice of you to show the time and battery life on top, but this clashes with your application. Also, the colors of the score, and "Menu" "Help", etc. all clash a bit.

Good luck-- I think you fundamentally have a good game, and with just a bit of a nicer interface, you could do very well.


Grats, it's pretty nice. Just spent some minutes playing it up to lvl 10 and now my wife is having fun with it.

I agree with the other poster that the bar on top with the battery/time is distracting. The menu fonts seemed a bit non-standard to me. The blocks seem a bit simple, could use some extra glow or something.

Keeping the high score on the top is a nice touch to encourage self-competition. The big combos are what make you want to come back for more.

It was kinda hard to figure out the "combo" mechanics at first (which blocks shrink when one disappears and a column falls down). There was nothing in the tutorial about this.

Regardless of the nitpicks, it's fun. I'll definitely play it again.


Does anyone know if this was created using the Flash Pro's iPhone app generator or is it a native Obj-C app? I'm asking since there seems to be a Flash version so that would have been the simplest route for this game.


This was a homebrewed port of my flixel framework, from what I've heard from fellow flash devs the CS5 beta is having trouble getting anything better than 1 or 2 fps. Canabalt runs at 30+ even on the original iPhone :)


I have noticed more iphone developers making a flash based version of their games for web promotion. I think it is a great idea, play before you pay. Does anyone have stats on how well this works as a promotional technique?

Also here is a link to the game in itunes http://appsto.re/canabalt


The Flash version predates the iPhone version.


This is how we played the Super Mario Brothers Lost Levels in university: mandatory hold down of Run+Right and only tap jump with varying pressure. You can finish all the levels like that!


Reminds me a lot of http://www.kaninkul.com/games/773/copter.html , both great games.


Also: Try Rope and Fly, I found it to be fairly similar and also quite addicting.


If you like that game try Hook Champ, same mechanic but with powerups, unlockables, great graphics and a committed developer with a major update already submitted.


addictive.


awesome game. If you make an android version, i'll buy it!




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: