i played shin maou golvellius, the msx2 version (the last of three, each of which has different maps, overall game progression, and bosses) of the game mostly known in english circles as the master system's closest analogue to the legend of zelda, and i keep thinking about it. this is one of my favorite 80s Sword Action Games i've played in this recent exploration, up there with ys 1 and 3 for me, but like a lot of the non-ys ones the central conceits are pretty bizarre in a modern context. these ideas are kind of tough to get used to on their own, since it's a game that's very much about trial and error, and the game builds them up pretty slowly at first before more exciting discoveries and abilities start coming into play in the second half. the guardian legend, which released between the msx1 version and the latter two, uses the same basic structure but attaches them to a more standard exploration overworld and shmup stages where you fight with projectiles...it's definitely more recognizably "good" and i might still like it better, too.
more than anything, it's like yoshio kiya's xanadu; the msx1 box says "by pac fujishima" and even with several other artists and programmers credited i understand the game is fundamentally his work. even compared to some of its contemporaries it's difficult to ascribe any real meaning to the individual secrets in the game; they're almost all revealed by defeating enemies, walking over hidden trigger spots, or tapping other objects with your sword, but they're also only activated in an esoteric sequence as you successively discover the ones that let buy new "bibles" to hold more currency (it's called "find," at least in this version, and as the game goes on you start meeting a bunch of the land's gods, so it's less weird than that probably sounds). and the scrolling stages are similarly trial-and-error based. if a player doesnt get tired of trying to do those things on every screen repeatedly until one of them works, it's only natural that they'd start trying to understand the game, and try to catch onto those sequences and developments faster. obviously, there's still some guesswork there; there are a fair number of hints around for various things, surely more than i even read and solved myself since i was playing in japanese, but ultimately i found myself starting to try and predict in advance things like where fujishima would put objectives, which parts of a screen were most likely to be important, and how to approach the scrolling stages. and as the game went on, i was right increasingly often. though i did get stuck looking for the last item for a while...there weren't really any hints that i could tell, but i could've remembered a screen i hadn't previously found anything on faster.
this is communication, in a way. you're trying to understand the creator's thoughts, via an intangible and abstract expression of them. and lots of games designed by one person, especially in an experimental way, have tons of this. unlike xanadu, i feel like a lot of them are hard to describe...golvellius presents more of an elusive vibe than an expression of a concrete idea. and i'm fascinated by that, even if this game doesn't particularly feel like art.
it is pretty funny though. especially the ending.