Intricate Surprising PlacesI enjoyed making a game. I'd like to make another one.
What I know, though, is that I can't make five more. I want to make one that I can keep making for years, adding content (for free) in addition to the base purchase. A game that can grow essentially as long as I keep working on it.
The reason I'd like to do this, of course, is Dwarf Fortress.
Tarn and Zach have been working on DF for fifteen years, roughly. They'll work on it for fifteen more, I'm guessing.
I like that approach. I can't be anywhere near as ambitious as they are, but conceptually, I'm on the same page.
I've been working on a game idea for the last few weeks, and I asked Tarn why he decided to make a game that had no ending state for the player or the developer. Here was his response, and he was kind enough to allow me to share it with you.
This is Tarn next, not me.
At the beginning, it had to do with not having an audience. Following a story or solving puzzles are two things you can't program easily for yourself, but exploring a random area or facing random challenges works, and there's no real cause for focusing on an end state. Arriving at a final score also mean less when playing your own games -- it was less about competition and more about forming a recording of events.
A part of it is the player-DNA from when we were kids. We never won any of the Ultima games and didn't really think of it as being possible, and the same was true of Rogue and the roguelikes. So our own RPGs were about wandering around and running yourself into the ground, more or less, and we slowly expanded the idea of Nethack bones files to the rest of the game. I think that had its first realization when I was about 15, working on the second version of our "dragslay" RPG. It kept track of the state of all the enemy groups that lived in caves from game to game (their leaders and size), and all of the dragons, and who they all had killed. So the world gets marginally more interesting as you play, since it tracks a few things, but you couldn't win since the monsters replaced over time as well, by any unnamed monster that killed one of your characters (reminds me of the modern Shadow of Mordor system, if you ever saw that). We still had score lists back then though -- it took us a few more years to drop those, in favor of diverse stats and records more like the DF legends mode.
There's the other influential angle, which grew out of the two expedition games (Starflight and Seven Cities of Gold). I don't know if Seven Cities had a win state or not, but we never got close. It was fun to just go out and make the discovered map grow. For Starflight, we actually did manage to win, against all odds, but that was just because the game's on a timer and if you don't blow up the crystal planet, you eventually can't play any more. Over time, we charted the fastest way to get the objects to win the game, did that, and then just filled in our lifeform notebooks -- it was cool that they let you keep playing after you win (which is another take on a "win state"). We've probably talked about this before, but the expedition vs. exploration angle is important to consider -- it's harder to make an exploration game carry itself, since it relies purely on the discoveries for metering. Having an expedition gives you a multi-faceted quantized game mechanic to work with, but it also can make the game seem a little more mercenary or, at worst, mechanical, sucking the life out of the exploration process. At its best, it can make the discoveries feel truly earned.
The whole Sim* series of games also deserves a shout-out -- some of those had win states (SimAnt, say), but that wasn't the point. It can also be handy to think of how exploration aspects and simulation aspects can work together -- discovery the rules of a simulation can be fun, and it can lead to replayable novel situations, though focusing on finding sim rules is probably a little more ripe for player burnout/gimmickness than other types of exploration.
Those are all the old reasons -- since I've been writing the same main game now forever, those are probably still the most important ones. I've started (and not finished, as you know) various side projects with and without win states, and for the ones without win states, I guess it still mostly comes from the same place: it's fun to dive into an intricate surprising place. I think that merges with the programming and design itself more now for me than it did before -- being able to adjust and add to the rules quickly is part of what makes the process enjoyable.