Reading Jimmy Maher's July 31 2015 instalment of The Digital Antiquarian - The 14 Deadly Sins of Graphic-Adventure Design (or, Why Ron Gilbert Hated Adventure Games) stirred my thoughts on adventure game finishability. I'm using this cod word to describe the extent to which players can finish an adventure game without heavy or total reliance on hints and walkthroughs.
I hated Space Quest and the majority of Sierra's walk-around-on-screen-while-typing games for all the reasons described in Maher's article. I hated them as a teenager when they came out, so this isn't revisionism. They seemed to combine the wasting of player time with being a jerk about the wasting of player time, and also sported a finishability of zero. I remember Police Quest in particular as a game where you occasionally had to type in chunks of the manual verbatim or die.
Sierra's attitude to testing their games on real players before release (they didn't) was a dumb one. Let alone testing them for finishability. But I don't think you can ever fully quantify finishability. The more testing you do of your game, the more accurate a picture you get. But you still can't guarantee particular results for every single player. Probably the best you can do is to envision a core player demographic for your game (whether narrow or broad) and hope that its experience, if represented in the testing sample, will be reflected at large.
Looking for something I can test, I like to subject my games to the following:
At the strictest level, I'd like at least one tester to be able to clear the game with no help from me or anyone else. Extra-game hints don't exist at this stage, so I don't have to worry about cheating.
If this first condition is not achievable or reasonable, a step down is that I'd like to see one or more testers clear the game with no extra-game help except conversation they have amongst themselves.
If it turns out that someone who isn't a total freak can prevail under the strictest conditions, that's reassuring, and failing that, having one or more folks prevail under the second-strictest conditions is pretty reassuring. Having tested these conditions means I can say, 'Yes, a human cleared this game with no hints, so I expect other humans can do the same.'
In modern times, I have yet to make a game so big or difficult as to render these tests too onerous. Maybe my philosophy will be tested when I do make such a game, but I suspect it will still be my starting point for thoughts on the topic of verifying some kind of reasonableness in my game.
I think I started fishing around for more concrete ideas on testing finishability due to my experiences with Anchorhead (Anchorhead's IFDB page). I found it hard, and the nature of its hints and walkthrough were such that they kept tipping me into cascades of needing more hints and walkthrough. I got bored, then offside, then I gave up, since I'd spent too much time just typing in commands from extra-game documents.
Anchorhead is held in massive esteem, so I seem to be a minority of one in having had this particular experience with it (and other similarly difficult games with similarly styled hints) but on the other hand I've never heard anyone say, 'I cleared Anchorhead without any hints or walkthrough,' during any of those discussions.
It must be a tough life to be a puzzle-sporting adventure game when such a game is subject to complete spoilage by hints. Players will overcome problems in the game or they will not, and the latter path can lead to a total halt.
I'm playing Forbidden Siren on the PS2 at the moment, my latest pick from my backlog of survival horror games. And I'm relying on walkthroughs from start to finish (Chris Pruett recommended using walkthroughs in his review of Siren) – a situation which I'd normally regard as defining a failed game, even in the action genre. But Siren is so cripplingly hard, but also novel and good, that I've found the experience to be worth it. Even in using the walkthrough, my ability to execute the solutions described in the walkthrough is massively tested, and the game is still atmospheric and clever and frightening.
This reminds me of why I'm so anxious about the nature of walkthroughs and hints for pure, non-action adventure games. It's because their use can eliminate the experience of playing such games, rather than just facilitate the having of one in the first place.