Lord Bellwater's Secret took my first place award in the 2007 Competition. I can say this because, looking back at the notes I wrote during play, they morphed from a paragraph describing the game's competence, to the admission "I'm not big on puzzles, but this is a rock-solid game", to simple proclamations about "best game I've played so far!" And there were no notes or doodling in-between these three -- a sure sign of immersion. This game simply did a lot of simple things right.
First, there's the game's opening: I know who I am, when and where I am, what I'm doing, and a strong suspicion of why it won't be easy. Though no one will cite it as one of the great openings of all time, I'm at least into the game and into my role before turn one. Since I-F has no box art, illustrations, or even a genre section from which to pick a work, I find such introductions an absolute must. Consequently, here I didn't feel those first few turns of aimlessness that much of I-F give me.
Navigation was a breeze, and is probably the single most important reason for the sense of immersion, that sense of the interface melting into the background. Even after completing the game, I'm still a little fuzzy on how the room in this one-room game was laid out, and that's the highest compliment I can give a work regarding its navigation. In Bellwater's room, we don't move from place to place so much as point-of-interest to point-of-interest, which tends to be some sort of prop or furniture. No need to clutter our minds with directions or bland names of areas like "alcove" or "hearth" when puzzles demand all our brainpower. Our imagination will fill in any blanks.
Characters were represented by the usual technique of in-absentia, with a notable end-game exception (though still non-interactive). Caring about the plight of characters is my motivation for playing I-F; countless video games of optimization puzzles and exhaustive searches have thoroughly trod that ground, with the iron-clad boots of pedantry no less. And it's a rare work when you can sympathize with your own PC; down that road lies angst and powerlessness, rather than the can-do, will-try activism of the typical PC. The only mark against this work regarding characterization was I couldn't remember my own PC's name without others present to remind me. Consequently a few things I read confused me.
The game's weakest link was, by far, the parser. The parser... It always comes down to the parser. LOOK AT/IN wasn't synonymous with EXAMINE, which is an important point in such a game where LOOK (around the room) is used so often. (I started using LOOK and its variations everywhere due to sheer habit of typing it.) TURN part-of-thing wasn't synonymous with OPEN thing, and a similar problem of parts vs. the whole emerged in the endgame. Both of these issues caused me to retreat to the hints even though I had actually solved the puzzle, but unknowingly, due to phrasing it for the parser.
One thing that authors should not imitate from this game: if Dear Player has reached your endgame, your last puzzle, your final showdown, it's a safe bet that he's fairly drawn into your world, and will be quite susceptible to danger. This makes a fairly safe bet that, once you've sprung that danger upon him, he won't exactly be thinking clearly. As a matter of fact, it's almost certain he'll be emotional, spastic, and utterly unaware of the fact that he has all the time in the world if he just stops typing. So please, Dear Author, in the end, ease up on the difficulty.