Many reviewers and people in the Impressions topic complained about how Puzzle Agent recycles some puzzles. At the same time, there are complaints that the puzzles are not challenging enough. At first I instinctively agreed with both, but the more I've thought about it, the more I've become convinced that recycling puzzles is not only a good thing , but also a good opportunity to increase the difficulty level without frustrating players.
I think that a puzzle game should be:
[*]not too easy;
[*]not too hard.
The reasons should be obvious. As far as originality goes, variation is key: a couple of classical problems (Perplexed Sock Picker, Find the Fuse) are OK as long as there are also some original ones, and they're not all too similar.
The last two requirements are tricky, because easy for one person may already be hard for another. However, this is where recycled puzzles can be of use.
When a new type of puzzle is introduced, the puzzle should not be too hard, because that will just frustrate the player. Rather, an easy first puzzle is a great opportunity for the player to familiarize him- or herself with the rules and figure out the basic principles used to solve it. However, if this is the only instance of the puzzle, solving this easy problem won't be very rewarding, and whatever the player learned about the puzzle is useless after it is solved.
If we allow puzzles to repeat, however, then later a new instance of the puzzle can appear that is a bit harder, so that the player will actually get a sense of accomplishment solving it. Since the player already knows the rules, more time can be spend on the puzzle itself. A third instance could really ramp up the difficulty, to give die-hard puzzle fans a good challenge. Of course, to prevent people from getting stuck entirely, this puzzle should either be optional, or come with some really good hints.
As it is, Puzzle Agent already recycled a lot less than it could have. The combination puzzles are the only ones which occurred more than three times, and some puzzles (Grickleback Baffler, Screw-up in Space, Hydraulic Lifts), could have been repeated, but were not. I would definitely have liked to solve some trickier instances of e.g. Screw-up in Space.
A final argument in favor of repeating puzzles is one of economy. Obviously inventing and implementing original puzzles takes more time and effort than recycling existing ones. Realistically, a game developer is only going to spend a limited amount of time and money on a single game. Allowing some puzzles to be re-used allows them to cram more puzzles into a single installment, giving us players more puzzle value for our money.
Note that I'm not arguing in favor of repeating one type of puzzle ad nauseam, and I do recognize that some puzzles really just work once (Room Key Confusion comes to mind). But I think allowing some puzzles to repeat is a good way to increase both the quantity and quality (difficulty) of puzzles in the game, at very little cost. Therefore, I think the re-use of puzzles deserves praise, not criticism.