Why did you build your own infrastructure instead of using Apple's turn-based GameCenter like Letterpress does? What specifically was missing, in terms of the quote "more revealing was how deficient Game Center actually was for building a complete, feature-rich multi-player game"
For banner ads we're using iAd and Admob as a fallback. For interstitial (full-screen ads) we're using Chartboost, Revmob, and Greystripe primarily.
GameCenter has a _very_ small payload, but more importantly, we wanted to be able to do polling (for "live" play), maintain custom stats, have stat thresholds (after "n" date, resigning will count as a loss), maintain whether people are online (actively playing), have custom push notifications that were more than just "you've been challenged" or "it's your turn", have chat, be able to make "random matching" parameterized (based on game variant, player ability, etc.), have Facebook and email-based auth, store invitations and player preferences on-server instead of on-device, have player nudging, player blocking, and more.
That's a pretty long but incomplete list. Most importantly, we can do whatever we want, whenever we want, without shoehorning it in to GameCenter. Take a look at the biggest deficiency of Letterpress - even Loren admits that it's GameCenter. I'm not knocking it - creating something that is universally usable is hard...
Mostly, we just have all of the control at our fingertips, and can do so much more :)
What's your revenue split between banner and interstitial? I've only been doing banner so far, but have been gearing up to try Revmob.
GameCenter's payload size got increased from 4K to 64K - you must move a lot of data! :-)
I do agree on chat, etc - I ended up making my own node.js chat server / presence server alongside GameCenter. Custom matching sounds quite cool. Nudging? Hm, could be done with APNS alone maybe, but yeah I see that then you'd be tracking match data on your server alongside GC and that's almost the same as building your own game server.
I agree that GC isn't great, but for us one-person shops the zero maintenance cost is a key factor. At least when it's working and not full of bugs :-)
We moved 1.5TB between devices and servers last month I think...
For a one-person shop, getting a game out at all is pretty amazing! I certainly see the appeal of Game Center. I'm not knocking it, we just wanted more flexibility and control. We started with a single API server and single DB server, but it was still over $200/m, so it was a cost we were willing to eat personally and experiment with - but we knew it had to become self-sustaining quickly, or we'd have to rethink our strategy.
As for ads - iAd pays about 75% as well as RevMob and Chartboost - believe it or not. That said, this month we're about 28% banner, the rest interstitials - they're worth it, as long as you don't abuse them.
4 comments
[ 3.5 ms ] story [ 24.9 ms ] threadWhy did you build your own infrastructure instead of using Apple's turn-based GameCenter like Letterpress does? What specifically was missing, in terms of the quote "more revealing was how deficient Game Center actually was for building a complete, feature-rich multi-player game"
For banner ads we're using iAd and Admob as a fallback. For interstitial (full-screen ads) we're using Chartboost, Revmob, and Greystripe primarily.
GameCenter has a _very_ small payload, but more importantly, we wanted to be able to do polling (for "live" play), maintain custom stats, have stat thresholds (after "n" date, resigning will count as a loss), maintain whether people are online (actively playing), have custom push notifications that were more than just "you've been challenged" or "it's your turn", have chat, be able to make "random matching" parameterized (based on game variant, player ability, etc.), have Facebook and email-based auth, store invitations and player preferences on-server instead of on-device, have player nudging, player blocking, and more.
That's a pretty long but incomplete list. Most importantly, we can do whatever we want, whenever we want, without shoehorning it in to GameCenter. Take a look at the biggest deficiency of Letterpress - even Loren admits that it's GameCenter. I'm not knocking it - creating something that is universally usable is hard...
Mostly, we just have all of the control at our fingertips, and can do so much more :)
GameCenter's payload size got increased from 4K to 64K - you must move a lot of data! :-)
I do agree on chat, etc - I ended up making my own node.js chat server / presence server alongside GameCenter. Custom matching sounds quite cool. Nudging? Hm, could be done with APNS alone maybe, but yeah I see that then you'd be tracking match data on your server alongside GC and that's almost the same as building your own game server.
I agree that GC isn't great, but for us one-person shops the zero maintenance cost is a key factor. At least when it's working and not full of bugs :-)
Great article, still re-reading it a few times!
For a one-person shop, getting a game out at all is pretty amazing! I certainly see the appeal of Game Center. I'm not knocking it, we just wanted more flexibility and control. We started with a single API server and single DB server, but it was still over $200/m, so it was a cost we were willing to eat personally and experiment with - but we knew it had to become self-sustaining quickly, or we'd have to rethink our strategy.
As for ads - iAd pays about 75% as well as RevMob and Chartboost - believe it or not. That said, this month we're about 28% banner, the rest interstitials - they're worth it, as long as you don't abuse them.