John Zeitler

In a discussion released recently regarding the Wii version of Punch-Out, Shigeru Miyamoto said:

Now, when young game designers make games and they are not fun, they add a lot of new material to try and make it fun. Even though they should make the game more fun using what they have right in front of them fully, they bring in new stuff.

In a nutshell, he’s pretty much summed up my main source of worry about the XNA projects. I could care less about how well the game sells, honestly (though it would be nice if the game does well enough to cover costs). What I’m worried about most is if the people who play it are having fun.

Right now, I’m in the engine-building phase; I’m designing, more or less, the tools that I will use to create a game. At any point during that later phase– the game design phase– I still have the ability to go back and say, “I need this,” and then go and implement it. This gives me a lot of freedom, almost too much freedom. This is because the more times I go back to the engine and tack on more stuff, the more complicated it gets; and the more complicated an engine is, the greater the chance that there will be bugs based on a flawed or incompatible implementation of the stuff tacked on.

When I was building XVINE, I had a dozen or so thoughts flitting through my head at any given time as to what tools I would need, what objects I needed to implement, and what situations I would encounter. When it came time to finally put it all together, though, I found that many of the things I thought were absolutely necessary were in fact either redundant or useless, because I could do it with simpler tools used more intelligently. Granted, there were situations where a single tool was only used once, but that’s because what it did could not be replicated by another tool combination.

In developing this engine, I’m running into that too. The entirety of this week’s coding was cleaning up the frenzied, hurried, slap-dash experimental implementations from last week’s initial “can I do this?” phase. It was a hassle, yes. But now the engine is designed in a cleaner, more effective manner. Similar objects now have common interfaces; inheritance and abstract classes are used to make sure things get done properly; and the engine has hooks for further optimization should it be needed. More to the point, with the new structuring of the engine, expandability is made that much easier– I’ve been able to easily start adding an “environmental effects” layer to each map layer in order to allow stuff like trap tiles or exits and so forth. (That’s going to be a harder part of the engine design, I think.)

What this boils down to is that once I’ve created the engine, I have to then find a way to make it fun. My original idea– a visual novel– isn’t not fun. At least, I don’t think it’s un-fun. However, a significant portion of the target audience– that is, Xbox 360 players– probably will find it to be un-fun. Yeah, this was pointed out to me before, but I resisted because I felt that the scrolling, tile-based engine setup was going to be too much work to properly implement. It’s not as hard as I thought, but it is a lot of work. The payoff is that I’ll have an engine which could be used to create a game that more people find fun– an engine which more or less works like any old NES game’s engine would, just in HD.

So what kind of game should I make? Satoru Iwata, in that same interview, made this remark on the subject:

It’s hard to know what to do when someone says to do whatever you want.

The tool that I have now is raw and unpolished, but its possibilities are limitless. However, in order to make it the best game that I could possibly make, I first have to know what kind of game I want to make, so that I can steer the engine in that direction. Now, there are some constraints. It’s a 2-D game, it’s top-down (so far), and I only have about 150MB of total space for code and resources. Really, those constraints are pretty minor. So, instead, I must fall back to the old adage: “Write what you know.”

I know old-school RPGs, and I know action RPGs along the lines of Zelda or the Mana series. I also know older action/arcade games, like The Guardian Legend and Pac Man. I really wish there were more games like those available these days. With that in mind, mimicking one of those would be pretty good for getting started, but there’s issues of complexity in each one: the original Legend of Zelda, as “simple” as it is, is far more complex than Pac Man. Heck, Pac Man itself is no slouch in the complexity department! Still, I have to take each project as it comes, and make a decision very soon on what project I want to work with first. My first project, then, using this new engine will be somewhat pedestrian– a Pac Man clone, just as a proof of concept. Once I’ve got that running and playable (it certainly isn’t going to be released) I can start work on the next project, which… I think I’ll keep quiet on for now. One step at a time.

Overall, though, I have to keep in mind the advice mentioned above. Less is usually more. Fun must come before whiz-bangery, and it’s almost certainly possible in this day and age to have a fun game in extremely tight constraints. Tetris is no less fun today than it was on an extremely high-latency monochrome LCD screen. And Final Fantasy IV told a compelling story in just one megabyte of space. I have 720p resolution, millions of colors, a powerful processor, tons of RAM, and 150 megabytes of storage. I envy the old days.