Less is more

For my next game I've been indecisive weather I should make a bit more complex game, or a simpler one. My main concern is that I want to make a commercially viable game, I don't want to make another "learning experience." So what makes people buy games? Well, they need to be good. That's pretty much it. Does it matter if the game is simple or not? I'd say no. But I've always wanted to build complex games. On the other hand I could always do that later. And it's far easier to make a simple game shine, like, say, Hundreds. And it's never/always too late to build something utterly complex.

So after a long vacation and allowing my mind settle on the topic I decided to run with a very simple game, but on a large scale, i.e. multiplayer. I also feel that real-time multiplayer can kill the casual part of gaming. I loath causal gaming.

The game I decided on is a game I've tried before and know rocks: multiplayer tetris.

We played this game in college and I was the worst player in our class, but I liked it the most. When you get two or more rows in a single move, you push as many "crap" rows to the bottom of one of the opponent's matrix. The goal is of course to be the last man standing. It's a fantastic adrenaline rush. Which is also to say that it is exhausting. "Exhausting" does not make successful iOS games in the long run, but I think it will get me a bunch of buyers when it goes viral. And if it's half as good as I remember, it will.

The original Xtris source code is only 3047 lines of C. Unfortunately I'll have to rewrite it, because I want peer-to-peer networking with UDP/STUN. Which means it'll scale immensely well, as all my servers have to do is to help blow some holes in everyone's firewall. I'll buy some type of Virtual Private Server for this. I'll also implement some different types of bots which will make it seem like there are actual players online (until I reach critical mass). These bots can run on the VPS as well. Hm. Or maybe I'll just run the servers on the VPS, and skip P2P. In that case I wouldn't have to rewrite much. But my minimum requirement is 15k simultaneous players, I'll let some early performance tests decide.

Perhaps I'll use Amazon's VPS solution, it's only $0.05 per "VPN Connection-hour." Hard to say what that means, but anyhoo. I'll steal controls from EA's Blitz. They're certainly better than anything I could come up with. Regarding payment model, I'll give X multiplayer matches away for free per day. If you want to play more multiplayer matches, you'll have to pay a one-time fee (tire 1=1 USD).

I looked back at my old postmortems in this blog to see if I can foresee will go right and what will go wrong. One obvious thing is of course that a lot of people hate tetris. On the other hand there are plenty of people who thinks otherwise. Myself I used to hate it before I tried the multiplayer version in college. I'll balance the difficulty through some automatic ranking system, so the common problem of making a too difficult game pretty much goes away (especially if I run some bots at both ends of the scale). The only problem I'm unable to solve is the intermission from when you've lost a match until the next one starts. Hm. I'll think about it, but apart from that I really don't think I can go wrong with this game.

I'll start working on MMO tetris in a few days. First I'll do my load test to see how many players I can support in a client/server architecture.