You get problems whenever the material and formal constraints aren’t balanced. In puzzle/adventure games specifically, the material causes (the items, specifically) are often only minimally sufficient to “solve” the game, and often have no use besides the one place where they are needed.

The player will usually start off trying to intuitively figure out what combination of items or interactions will solve a problem, but after they’ve tried half a dozen things which “should have worked” but didn’t, they come to the conclusion that they’re just going to have to try things at random.

Part of that problem is that adventure games typically have over-specific solutions, but the bigger part is that they provide poor feedback. The player is looking to get results which relate to his intention, but all the game spits back is “I can’t do that,” which makes the player feel stupid.

For instance games typically have so few items in them, it becomes a question of using such an such an item to solve a given puzzle. This works but I think it’s poor design, since there’s not much creativity on the part of the player. In the language of this article, I think you would say that there is a poor balance between material and formal constraints. The material constraints in typical puzzle/action-adventure games are insipid!

This problem came up in testing The Real Texas, people at the start of the game often spend a lot of time looking at various litter in the parkade. This litter has little real purpose, but the player (at this point) is expecting a certain kind of puzzle. I’ve seen more than a few users try a completely illogical solution of combining two items together simply because they are both present.

I like the idea of Agency, I think it captures at least part of what makes a game satisfying (vs. just fun) to play and progress.