Friday, February 23, 2018

Constructive Difficulty

As you know, I like construction games. I like building things.

But nearly all construction games make construction too easy!

I don't think the games are too easy. I think the construction is too easy. I don't need goblins attacking in waves or plague or economic challenges. I'm here for the building! Give me a construction challenge!

A long time ago, I fell in love with Medieval Engineers, because construction was a real challenge. At the time, there was no inventory: if you needed ten stone to build a wall, you had to get the ten stone within a few meters of the wall. If you were building a tower, you had to build a scaffold first, so you'd have a place for stone vertically close enough to the wall. In turn, the mechanical elements - carts and pulleys and stuff - were actually useful, because you didn't want to lug those stones up yourself, one by one!

In short order they added a magic inventory. Also, they made the buildings more structurally forgiving, because they decided to focus on warfare, and the idea was to see how well your building could survive bombardment instead of if your building could stand at all.

This ruined the game for me.

The problem is that Medieval Engineers is actually more about siege combat than anything else. With that focus, simplifying the construction and focusing on how the buildings survive sieges makes sense. It allows players to get to the meat - the sieges - faster and more robustly.

Of course, I don't give a single crap about sieges, so to me the game is now pointless. But I can't really blame the game devs for not having the same priorities as me.

I want to build. I want to have to figure out how to make a three-story building and use buttresses to keep it from falling over. I want to see how high I can build a tower, and what clever things I can do with internal supports to eke out a few more stories. I want to build a castle on a cliff and have to figure out how to wire it into the cliff so it doesn't go sliding down the mountain. These are the challenges I like.

This isn't limited to medieval stuff.

When I build space ships, I want to struggle to get them to work at all, so that when I make even small, functional rockets it feels great. Kerbal used to be quite good at this. However, as Kerbal came closer to release, they carefully dumbed down all the physics elements. It's now extremely difficult to get a rocket to fall apart: everything auto-welds and the forces have been much reduced. The game is much more focused on simple payload/fuel calculations, which isn't nearly as interesting to me.

Starship Corporation is a lot more my style, with intricate construction and extremely interesting testing phases. However, the game's difficulty doesn't revolve around engineering challenges, but around contractual obligations and research constraints. Once you understand the complex, nitpicky nature of things like power and oxygen, you can engineer a lot of interesting ships... except the game doesn't give you any additional parts or hulls until you've spent hours and hours and hours doing absolutely nothing. Again, the challenge isn't related to the construction, the challenge is based around some external measure of fitness that can't really be turned off.

There will always be some external measure of fitness, some external gating or rating system. Otherwise the game would be completely freeform. That wouldn't necessarily be bad, but it would fail to engage a lot of players. So the external ratings continue.

But these external ratings could be tied to the actual construction of stuff, instead of to how well you do things outside of construction, such as stabbing orcs or balancing a corporate checkbook.

A good example of this would be a skating game. The game is all about doing tricks and stunts. You earn scores based on your tricksiestuntness, but the scores are largely open ended.

The scoring is usually weighted to prefer long combos - meaning that you typically want to string tricks together rather than go for one ultimate mega-trick. This shapes the way you do tricks, but it's still fundamentally about your own stuntwork and allows for plenty of personal self-expression.

Alternate ratings systems - such as bonus points for being higher off the ground or moving at faster speeds - would result in different priorities for our tricks. Moreover, these would typically be just as easy to implement. It would be trivial to make it a toggle: join one team and get rated based on chained tricks. Another team, rated on speed. Another team, rated on height over the ground. Now the player has a ton of freedom to shape their own external ratings system to match their own stunt construction preference.

I want building something to feel the same way. I want my construction to feel like a neat trick chain in a skating game. Something I personally did, with my personal skill and priorities and artistry.

And that means it has to be hard to do, but also with enough slack that I can do a lot of different things for any reason I feel like.

Can that exist as a game people want to play?

I dunno.