After I finished Bael’s Rock, I turned to the walkthrough – a document with which I’d become well-acquainted – to see if I could solve a multi-step Easter Egg that requires finding a series of clues across the game’s eight chapters. The walkthrough didn’t spell out the answers, only providing a series of hints, and the one for clue number 4 says it is “the most obscure one, requiring a code that is commonly used.” Ah, the contemporary IF player says, just gotta type XYZZY, doesn’t get much clearer than that, right? But it isn’t, because the full hint is actually “…commonly used in immersive-sim games.” I was able to solve the puzzle nonetheless – Thief and System Shock are some of my all-time favorites, so I knew exactly what it was getting it – but it’s symptomatic of a game that’s full of enthusiasm, but to its detriment feels cut off from the past several decades of parser-game development.
I don’t mean to overstate the case, here – heck, the completed Easter Egg is an Infocom reference (Spoiler - click to show)(you get some grue repellant). And I hate to come off parochial or a gatekeeper, I can think of numerous instances of authors who, because they were largely unburdened by expectations and preconceptions, have brought new, exciting ideas to the scene. And there are definitely glimmers of that in Bael’s Rock; there’s an extended riddle-duel that puts a clever spin on a hoary scenario, for example. But for the most part this is a collection of fantasy-themed medium dry goods puzzles of the sort seasoned players will have seen many times before, and the author’s seeming lack of familiarity with the state of the art makes proceedings feel significantly rougher than they needed to.
The most obvious place I felt this friction was the parser. This is a custom-built game, and it’s cool to see such things out there in the wild, especially released on Steam! But in a world where the major parser systems offer incredibly robust, well-tested gameplay, custom parsers are inevitably held to a high bar. In terms of features, Bael’s Rock generally does OK – most common abbreviations work, the up arrow key retrieves previous commands, you can get a history of the last couple of turns of output, and it seems to be able to handle sentences of comparable complexity to Inform, TADS, and Dialog.
But there are some notable omissions, like nonexistent handling of adjectives – I got very bored with having to type PUSH RED LEVER over and over instead of just PUSH RED. And the error messages are very frequently misleading, failing to clearly tell the player whether what they’re doing won’t work, or if the parser just wasn’t able to translate typed input into action: “you can’t examine that” and “you can’t take that” are very unhelpful responses when the player’s typo means the “that” just wasn’t recognized, and the fact that ILLUMINATE ITEM WITH TORCH spits out “you can’t use this item to do that” unless you’ve picked the exact right ITEM is actively misleading. Similarly, I had to go to the walkthrough for almost the first real puzzle of the game because PUT AMULET IN LION STATUE says “you can’t put it there”, rather than prompting you to be more specific about where exactly you want to place it.
The interface also provides more of that friction. Layout-wise, outside of cutscenes most locations have two main text boxes – the first being a largely-static description of the room, the second responding to player input, plus a command line below. It’s a reasonable enough approach, but many actions lead to output that won’t fit in that second box, meaning I very frequently had to hit enter to advance through two or three pages of text responding to a seemingly-innocuous command. I also experienced a slight delay when new text was displayed, either in response to actions or moving to a new room, which made certain sequences, like the maze – spoiler alert, there’s a maze, and you map it by dropping superfluous inventory items to track where you’ve been – feel molasses-slow. That’s a shame, especially because of the way it undermined a made a couple of puzzles that I enjoyed, but required a bit of trial and error by design, like the climactic sequence of the game.
Speaking of design, the puzzles here are very traditional, with every chapter requiring you to progress past a set-piece barrier of some kind. While there are a few exceptions, most of these involve simple object manipulation, fiddling with the occasional complex device, decoding abstruse hints, and the odd bit of magic. This is a classic formula, and again, in the abstract there are some solid brain-teasers here. But unfortunately in addition to the weak parser, there are some guess-the-verb issues that quickly taught me to run to the walkthrough rather than waste time trying to figure out what I’d missed – the most egregious offender here is an early hurdle where after you’ve collected three pieces of a cube-shaped key, I was nonplussed about how to actually use them. COMBINE KEY is, I suppose, a reasonable approach, but abstract, goal-driven verbs like that aren’t typically a part of the modern IF vocabulary, and there’s nothing in the game’s help text that indicated it’d be a reasonable thing to try, nor were comparably reasonable attempts like PUT FRAGMENT ON FRAGMENT successful.
The last place where the game could have benefitted from familiarity with post-80s IF design philosophies is the story. There is one, but it’s pretty thin: you’re abducted and abandoned in the middle of the desert, forced to find shelter in a mountain crammed stem to stern with a-wizard-did-it obstacles. There is a reasonably well-done, only somewhat telegraphed, twist at the end, but there are plenty of satisfying works of IF that have similar story structures. However, where most modern players would expect that a game like this would do some environmental storytelling, weaving some backstory into passing descriptions and otherwise creating a cohesive world that rewards exploration, Bael’s Rock mostly sticks to just the puzzles, ma’am – late in the day, you do find the journal of a previous adventurer, which I was hoping would provide an opportunity for just this kind of exposition, but it mostly just lampshaded the arbitrariness and implausibility of everything that had come before.
I am being hard on Bael’s Rock here, I recognize, perhaps out of proportion to its faults – there’s nothing wrong with its bones, even if it isn’t the kind of thing that sets my world on fire. But it’s really a shame to see what’s so clearly a passion project stumble repeatedly on problems that have been solved for decades (including the lack of a transcript feature, which is one reason I’ve been more specific about stuff like parser errors than usual). The good news is that the changelog indicates that the author’s been making upgrades, and the game ends with a blatant sequel tease. So even though I got as much frustration as enjoyment out of this game, by taking feedback on board, and hopefully building experience with contemporary IF approaches, I’ve no doubt the author could make something really special in the future.