Spring Thing 2020 "Napier’s Cache" post-mortem

I thought I’d jot down some thoughts on the design and development process of this parser game. The opening of the game was originally entered in IntroComp 2018, with the full, completed game entered in Spring Thing 2020.


The game sprung out of a tweet I wrote about my ancestor Sir Robert Logan of Restalrig hiring mathematician John Napier to hunt for buried treasure in the sixteenth century. IF author Victor Ojuel replied saying “You’ve got to do that as an IF game!” Which I hadn’t thought of at all till then, but it got me thinking.

Robert Logan hired John Napier in the 1590s, and the contract drawn up between them survives today. But it’s not known if the search actually took place. So that left a narrative gap I could potentially fill, combined with stories of John Napier’s occult interests and practices. Much game potential certainly.

Development system

I decided to develop it in Inform 7, which I was still learning. I figured it would be another good project to help me learn the language and system.

In design terms I thought of feeding in back story via the player collecting objects for the treasure hunt trip to Fast Castle. The player was John Napier’s servant, which I thought fitted well into the story. It also had the advantage that the servant didn’t know everything that was going on. A simple get the objects quest provided a nice opening, and was a natural fit for IntroComp. So I aimed for that, and with help of some great playtesters got it into the competition.

Following that the game had to wait a while until I was able to complete it. My neurological disease was playing up badly. I knew I couldn’t enter the complete game in IF Comp, so with timing as things worked out Spring Thing 2020 then looked appealing, especially after playtesting suggested an earlier completed version of the game needed revising quite a lot before submission. Luckily with more playtesting input I finished the changes just before my neurological disease went badly out of control in late March 2020. It was in the comp!

The game’s scenes

The introduction of the game is set in John Napier’s home castle in Stirlingshire (this was before he inherited the family castle near Edinburgh), but I knew that the main action would shift to Logan’s Fast Castle in Berwickshire. This shift led me to adopt a scene by scene approach to the game, with distinct shifts in place and/or time. Easy to implement using Inform 7’s scenes, but also made it much easier for me, a still learning author, to handle space, time and plot.

The main section of the game is set in Fast Castle which is now almost totally in ruins today, but has been excavated by archaeologists and investigated quite extensively. I based the layout on an English drawn plan of the castle in the 1540s. The castle was a surprisingly small one, which surprised some players.

I wanted to have a scene where the player and Napier met Robert Logan. I opted for a dining room scene, where the conversation played out over a series of turns. I also added another character: Francis Stewart, Earl of Bothwell. He wasn’t his more famous uncle Bothwell, who married Mary Queen of Scots. But the younger Bothwell was also notorious in his day, and outlawed at this time. He was Logan’s cousin, and a regular visitor to Fast Castle. Adding him to the game helped me write the conversation that ensued between Napier and Logan, adding an extra perspective. I did worry that people would find this part tedious, so trimmed it down in length as much as I could. I also added a couple of mini puzzles part way through to break it up. To my relief playtesters were happy with it, and really enjoyed reading this section of the game. Big phew!

After the dining room scene concludes the action shifts to the castle cellar, where Napier is trying to uncover what is going on. This is a series of mini fetch quests, where Napier asks the player to get something needed in his occult rituals. Initially these fetch quests weren’t nearly challenging enough. Even now I think I should have added more depth. I am still learning how to write puzzles though, so this was a big challenge! But generally I was happy with things as they turned out. On the downside it is rather railroad-y. One playtester described my game as in the “light-puzzle, story-focused genre of IF”, which I think is a very fair comment. But I probably need to work on my puzzles more …

Because the castle was so small there were limits to how much I could flesh things out. I also wanted to keep some areas off bounds, for example the private residential area of Logan and his family. I added some NPCs you could talk to, though this wasn’t essential. I know some players found the talking not ideal, so I need to look at revising my approach to that in future games. I also found that adding a couple of NPCs helped me up the puzzle factor in the fetch quests in other ways, which was a good thing.

The final section of the game sees the player search in an underground cave. Again the puzzle level is low, and it’s more about story and atmosphere. The final ending was a challenge to write, and I’m not sure I pulled it off totally successfully. But it did fit with the characters as written up to that point.

Developing the through line first and improving Inform 7 skills

In terms of development I followed Emily Short’s recommended technique of coding the through line first. I.e. get the core route through the game in place. However after doing this I felt somewhat discouraged from fleshing out the rest too much. I was a bit worn out creatively! I definitely needed to allow time and distance away from the game, to develop it over a longer period, to allow a fresh look at it, and then make good changes in places.

A really nice thing is that working on the game I feel that my Inform 7 coding and development skills have improved massively. Simple things like coding examining something and uncovering an object not initially visible might seem like they should be easy to code, but I didn’t know how to do that. More complex things included using CHECK rules rather than INSTEAD in a lot of places - especially the many NPC conversation responses - for efficiency purposes. And I was very pleased, after much prompting from Mathbrush!, to finally get to grips with changing default game responses for lots of verbs etc. I also have a much better idea of the long list of things I need to look at polishing in the closing sections of a game, tips that I can apply to future games I write.

Player response and looking ahead

When the game entered the competition my health was plummeting as my neurological disease went out of control. So I couldn’t engage as much as I hoped with players. But it was really nice to see generally very positive reviews and player ratings come in. I particularly liked one review where the reviewer was prompted to go and read about the history, and found it fascinating.

Looking ahead I’m encouraged to carry on writing IF. At the moment I plan to stick to parser IF, because it’s what I’ve known for 40 years as a player, and I think I’m getting to grips with writing it in Inform 7.

I will probably often stay in the historical theme, and very possibly Scottish! I definitely have a leaning towards that. Not terribly surprising given my occupation as an academic historian … I would like to write more puzzle heavy games, but I need to think more how best to approach tackling that.

I am also looking again through Aaron Reed’s book, which was particularly helpful near the end for giving me pointers to useful last minute polishing, and helpful extensions. So I will carry on reading that, and carry on creating games.


This was a fun game to test and play! I definitely look forward to any more historical/Scottish games!


Thanks for the postmortem. We had fun playing through the first part for the Seattle / Tacoma meetup.


Ooh marvellous! Thanks for letting me know.