The terms in the title would refer to the familiar strategies known in information processing. This post is definitely not to explain the terms but to explain some of my personal insights towards using a dual-sided approach towards game design.
Games can really be complex to piece together, especially to avoid feature creeping. Most ideas I've seen from experience are structured and visualized from a top-down approach. A most common ideation process would often go as such:
Well, that indeed went fine and has defined a lot for the game's structural system. However, the above idea didn't sound too far off from common games out there, in fact, kind of plain. Let's see how it would go if we'd go with a bottom-up approach.
No lies, the latter definitely took more time to come up with as an example. But hey, the 2nd one seems like a lot of fun even though it's lacking a bit of systems that hold it in place in which the top-down approach would help.
So what was wrong with the former top-down approach? Well, I'll put it as because of how steps 1 to 5 has already defined so much of how the game is, it'll be much harder for designers to instill creativity into the game. Weapons designed from step 6 onward may probably lean towards dealing damage and killing enemies. It sorta limits the game due to the boundaries set. Thus, a mix and match of both approach can really help merge systems and creativity together. I hope this could provide insights for anyone brainstorming for their next cool game! Ciao~
I definitely have played countless games (like you) since I have access to media hardware. Below are a couple of the games I really appreciated (even now as a game designer):
Crash Bandicoot Series
Before Naughty Dogs dived in to interactive stories, they had really great old skool' level design. These great engagements were rampant in not just their classic platformers but in their racing game, Crash Team Racing as well as their multiplayer mini-games, Crash Bash. They were the Sony console replacements of Mario Kart and Mario Party back then. These types of good ol' fun can hardly be seen nowadays since needs and demands of gamers have escalated. I am totally looking forward to the upcoming Crash Bandicoot Trilogy remake.
Directly out of my childhood memories, there are so many game design perks to learn from in these games. The series are really good at creating replayability back there, levels had crystals, purplish-gem and perfect wumpa fruit collection milestones. The levels and racing tracks are always filled with secret paths and shortcuts. Any guide writers back then must not have it easy, haha.
I remembering queuing up for the collector's edition for this after saving up my allowances. A magnificent game by Will Wright when he was still a game designer. There's no game quite like its own thus far. Different evolutionary stages of the game offer familiar genres of gameplay but fun enough at its own right because it's just so interesting to watch your creation grow and become sentient. The cell stage of the game has referential mechanics to the currently popular Agar.io. The space-faring stage of the game kinda felt like a quirky random multi-option narrative that some latest RPGs are demonstrating.
The creature creator of this game was really almost like a Google-esque UX version of a 3D modelling software. There were even modders that unlocked the complexity limit of the tool just to create crazy semblances of Pokemon and other popular animal characters. Ending off the post with a photo of my prized collector's edition below: