Fifteen Minutes postmortem

So I have seen the most excellent analysis done by the other authors regarding a postmortem, and I thought I would throw my tuppenoth’ into the mix. I don’t have a great deal to say, but I have found the experience interesting.

Whilst not outright spoilers, I guess some of the following text is spoilery(ish). So I’ve hidden it.

[spoiler]I was actually a very early poster on the old r.*.i-f boards, and was around for the first competition where I judged some of the games. I then had a 19 year IF hiatus for various reasons before happening across Jimmy Maher’s most excellent Digital Antiquarian blog which rekindled my interest. In one of those odd coincidences that only happen in the movies, it just so happened that in my professional life I was beginning to explore options around developing embedded scripting languages, and had begun to investigate the facilities of inform7, as well as a bunch of other tools.

I never really intended to enter Fifteen Minutes for the comp – or release it at all. 15 mins is my first IF game, and my ‘learning i7 game’. This is, rather than any conscious preference for IF ‘form’ or game types, why it turned into a little puzzlebox as opposed to being more narrative driven. I wanted to learn i7 at quite a fundamental level, and use most of the features in depth and the time travel concept, with all its inherent complexities, seemed to be a good challenge.

I must confess that when I started developing it, I was unaware of the leaps in critical thinking about IF and its roles, form and the innovations that had been made in my absence. So as I read more and more of the IF board and blogs and tried to catch up on 19 missing years, I did become very aware that Fifteen Minutes would be a ‘throwback’ - a traditional puzzle oriented mini-game with minimal narrative. However, the reason I entered it into the competition was that I thought the integrated nature of the puzzle itself: the multiple selves interacting to avoid paradox, the requirement to gain enough time to both learn and pass the science test, the requirement to build the environment in order to gain the information needed - was innovative and interesting enough in and of itself to make it a worthwhile entrant.

Under the surface, 15 Mins is a complex little beast. It’s also not very well implemented. Well, I say not very well, it’s not ‘buggy’ or defect ridden as such – I’ve ironed most of the bugs out with some amazing help from some excellent Beta testers (actually, truth be told, there is a bug, a big one – but none of the reviews I’ve seen have discovered it yet – it requires a pretty unlikely series of actions by the player). A better way of putting it: It’s ‘inelegantly’ implemented. I think this is down to two things: the learning curve for i7 is surprisingly steep and the structure of i7 lends itself to inelegance in i7-newbie hands such as mine.

In 15 mins, the list of ‘things to check’ every turn is……er……considerable, and requires the state of things in the game to be monitored in several ways. Since release, I’ve thought of at least three ways of implementing in a much more elegant fashion – however, I must confess I have no intention of going back and re-factoring the code.

For the game itself, from an implementation perspective, a couple of things I was pleased about (in addition to the fact that it didn’t bug out [emote]:)[/emote] ) :

The little ‘action phrase’ generator – Other You’s actions in non-critical game minutes are generated randomly from atomic phrase elements. It worked quite well I think – I needed to give the player ‘free turns’ and needed to make phrases that indicated stuff was happening, but were extremely none specific and vague about what the character was doing so the player could fill the hole without breaking time, and yet were also kind of funny, and a bit silly in keeping with the game’s voice. It also added life and interest to otherwise empty turns.

Cumulative time: The watches being the key to the puzzle solving mechanism. For something so seemingly simple, this was an absolute ***** nightmare to implement and get absolutely right. You have no idea.

I was pleased with the information delivery mechanisms – every single piece of information a player needs in order to complete the puzzle is available before they need it. In theory, this game can be completed in one play through. The combination of the watches, the book, the page of instructions, the visibility of how the other You’s are using the machine means that it’s all there……ok, there’s a lot of it……but it’s there!

Things I wasn’t so happy with:

Paradox paralysis and the lack of feedback on ‘death’. This was pointed out by a lot of reviewers. I agree. Basically, the mechanism for assigning ‘paradox paralysis’ to the player is complex and tightly wound throughout the whole implementation. It was on my TODO. I ran out of time. No excuses.

An overly complex central puzzle mechanism. As I said, 15 mins was intended to be a programming exercise. In hindsight some of the things I did (ternary switch series, binary implementation) in order to create a complex machine for the sake of creating a complex machine, should have been ruthlessly chopped out of the finished version. But I was in love with my complicated machine. Many reviewers pointed out there was no need for this and it resulted in pointless busy work for the player. Tough love is the best love. I agree.

I am still on the fence though as to whether the central puzzle was ‘too hard’. Hmmmm…this is a tough one, and from the reviews I’ve seen, reviewers seem to be split 50/50. Some seem to have given it up as a bad job quickly. Others persevered and felt a level of reward on completion.

I guess the one disappointment I had was that the true narrative hidden within the game never emerged. One of the key puzzles revolves around the relationship between the Professor and Woolpak, the author of the book. I scattered clues around the game. In order to answer one of the questions, the player must intuit and derive and understand the why of this relationship by interpreting these clues. I’m not sure, from the feedback I’ve had whether this aspect of the game was either noticed or whether it achieved the effect I was intending. I think I need to go back and rethink how central to make this narrative.

Thoroughly enjoyed being part of this competitions, and glad to be back in the IF fold. Looking forward to next year. I have an idea for a new game, but, then again, ‘Fifteen Minutes II : Hildebrand’s Revenge’ just desperately wants to get made[/spoiler]

Ade Mct

First, it’s really cool that someone came back and wrote something good like this 10+ years after visiting the newsgroup. Other people can hold out hope that they can do the same.

Second, for what it’s worth, I think people would really enjoy a director’s cut version of Fifteen Minutes. I think, once the player solved the game, you could have an AMUSING section–actually, I think you may have had it, but an additional option like “find some CLUES to the Woolpak and Professor story” would be nice–I noticed some stuff, and I think I sort of concocted my own story from what I saw, but it was probably the wrong one. But even that added a lot and was a nice reward for getting through the puzzles.

CLUES could check off things you didn’t examine, or books you didn’t read, and list out all the stuff you didn’t look at. This would add replayability to the puzzle. That would be a neat extra programming exercise, too, and given the hoops you jumped through, it would not be too bad. You could add something to the Table of Final Question Options.

I’ve also had problems where something I found neat to program wound up being a challenge for the player. Maybe you can find another puzzle game where the switches work.