Devlog #2
Lately I’ve had an interesting problem with simulating work with drill and lathe for Gunsmith Simulator. Some aspects of it will also affect woodworking and other parts of the game. Let’s set it up first. My task was to nail down repair and production of parts that is done using machines. For this game, the starting point is always the real world, so what are the specifics of cutting precision parts in metal?
1. There are very few ways to do it right and uncountable ways to do it wrong.
This is not a sandbox, where you can shape and reshape material continuously. The end result can’t be as varied as anything between a sand castle and a carving of a mermaid. You can’t just drill a bunch of holes and later make it kinda work with sticks and glue. No, you need to arrive at a very exact piece. So even if you can do that with 3 different tools, or you can do some operations in any order you want - the right way to do it is very limited.
2. It’s unforgiving.
Measure twice, cut once. In many cases, a divergence of a thousandth of a millimeter is unacceptable for a gun part. So if you drill just a bit too deep, a bit off or out of axis, that part is usually ruined.
3. It’s in large part tedious.
After you’ve already planned the steps of the process, carrying them out is actually quite boring. Especially if you factor in firstly a very precise positioning before you even start and then frequent measurements which are necessary to abide by exceptionally low tolerances.
Now, just reading that list - does it sound like a fun element for a game? For the vast majority of players that’s a flat no. It has to be conveyed in a better way. We don’t want 8 hours of sweating in the workshop, we want the edited time lapse video produced from that day of work. It should hit all the key points on the way - of course, but without some of the boring bits. So, let’s try that approach. Player can play around with a machine freely, but when he decides “I’m making X” from that point the game limits his options only to the correct steps of that specific process. Click to turn on, click to move that, and so on. It’s easy and moderately interesting but that was clearly not a good enough solution.
After all, we don’t want that edited time lapse either! It’s a game, so we want to be “doing it”. So on the second pass, I’ll make it more user-driven but still keep it oriented. In the most important key steps for a process (usually these steps will be cuts) it’s now the player who has “hands on” controls. Failing and destroying the piece is still blocked, but it’s up to the user to actually do the task to succeed. That was immediately a more satisfying solution.
How about some stakes? I’m still apprehensive of introducing real life stakes, where you can just cut in the wrong place or go too far and an almost fixed piece instantly becomes junk. I’ve put in smaller partial fail states, which admittedly are a bit of gamification. It is now quite easy to overheat and break a drill or a bit. This is good from a game making perspective: a user while doing a task “hands on” has a good reason to exert some concentration while doing it. However the penalty for failing in it is not prohibitive. The downside is that this solution is stepping out of the “simulation” aspect of the game further than other features. We’ve already done some tests and they are looking promising, so maybe this final solution makes the cut for Gunsmith Simulator? Maybe I’ll have to tweak it even more. But what do you think? Is this one of these places where we have to put the game first and aim for the most satisfying user experience even if we stray a little bit, or do you think simulation should take precedence and it is better implement it in a less fun way trying to stay more true to life?
Or maybe you’ve got some killer solution that can attack that problem from a completely different direction?