Bio:

Jon Ingold is the narrative director and co-founder, with Joseph Humfrey, of inkle. They’re best known for IGF-winning 80 Days and the open-sourced scripting language ink. Jon was previously a lead designer at SCEE, and a writer and parser-based interactive fiction author.

Data Box:

Developer: inkle / Steve Jackson
Publisher: inkle
Initial Release Date: May 2, 2013
Platforms: iOS, Android, Steam
Number of Developers: 2 core people, 2 additional contributors, and 4 art / music contractors
Length of Development: 4 games over 4 years
Budget: ~ £10k / title, not including our own salaries
Lines of Code: Just over 60k significant lines of code (non-whitespace)
Development Tools: Xcode, Objective-C, ink
Downloads across the series so far: >1.5 million

About five years ago, Joe and I sat down in a West London pub with Steve Jackson, co-founder of Games Workshop and Lionhead, and co-creator of Fighting Fantasy (and a personal game-design hero of mine). We were showing him our early prototype of an “inklebook” -- an iPad-based choice-driven story made of “pages” which stitched together into a single flow via frequent choice points. 

It was a lovely UI for a choice-based game, even in prototype form, and we had a strong scripting language underneath it -- the first version of ink. Steve listened, nodded, and told us to come back when we’d sold ten thousand copies. 

A year later, after the release of Dave Morris’ Frankenstein, we met Steve again. The timing was perfect: the Sorcery! license, previously with another developer, had just become available. 

We promised him the world: a full adaptation, with graphics and sound; a slick, Apple-friendly UI; dynamic characterization of the player character; strategic combat narrated in natural prose which would write “as well as David Gemmell”; a map with day-night cycle effects; an intuitive yet powerful gesture-based spellcasting system. Then we set about adapting Part 1, which was released in May 2013. 

Not all of our initial promises came to pass: over the course of four years and four games, we iterated, throwing things away and developing in new directions. Overall, however, we grew in ambition, complexity and scope.

It’s a little tricky writing a retrospective of a four-game series. We’ve changed things from game to game, and we’ve been steadily back-porting features and fixes into the earlier games. The first game is much better now than it was on release. But here are five things that went well -- and went well four times over; and five things that went badly, that we never managed to resolve. 

WHAT WENT RIGHT

1) Steve Jackson, and the Sorcery! fans

When we first approached Steve, he took a chance on us. We talked a good game, I think, but we didn’t have that much to back it up with. He talked through a lot of our ideas with us before we started, but then he took a step back and let us get on with development. He didn’t micro-manage, he didn’t hassle us for updates or explanations of every decision we took.  

He waited for the first, early build, and then gave us some considered feedback, but otherwise, he let us do what we needed to do. He put an enormous amount of trust in us to handle his IP -- and his fanbase! -- with the care and respect they deserved.

Me and Steve in Steve’s library, holding a copy of the printed Spellbook.

For the first release, we played it safe. Part 1 is a tight adaptation of the original book, only expanding on hints dropped by the book and largely keeping the original prose, albeit expanded upon. 

All the same, we were terrified: would the fan-base declare our version a heresy? Steve had agreed that we needed to remove dice-based combat, but would the fans feel that was a change too far? And our version had small choices every 100 words or so; many, many more than the book had -- but some were mere flavor. Would people find it a chore to play?

Thankfully, the response was overwhelmingly positive, and that fan support allowed us to start taking more risks, opening up and reshuffling events in Part 2, and then exploding the world entirely in Part 3, where we created two overlapping time periods and distributed the original content between them. 

Both Steve and the fans got happier and happier. Here’s was a world and an adventure they loved 30 years ago, and it wasn’t just being repeated, it was being rejuvenated. It was being made into something they could get some kudos for knowing about. At least, we hoped that was how they felt!

As a Sorcery! gamebook fan myself (my first ever computer game was an adaptation of Part 4 begun when I was eleven, which I quickly abandoned when I realized I was going to have to type the whole thing in), it was that sense of making a great thing great again that made me excited to be working on the series. 

 Steve and Joe testing Swindlestones

At inkle’s 3rd birthday party, somewhere between the launch of Part 2 and Part 3, Steve came and brought with him a signed first edition of the Shamutanti Hills, one of two from his study. At the moment, it’s sitting in pride of place next to 80 Days’ IGF award.

2) The map

The single best design decision we took was “the map.” It began as a purely visual thing. We were thinking of the maps one sees at the start of fantasy novels, which offer a world to explore and discover, but which are ultimately disappointing, when the places marked never show up in the book. What if you could visit any of them? What if they all had stories to tell? 

But the map has proved to be more than eye-candy. It is our progress meter, showing how far along you are. It is our proof of branching and consequence: we don’t need “the game will remember that,” since you can see that by taking the river, you’ve avoided the village.

It’s our strategic layer: where were you heading? What risks will you take to get there?

It’s our checkpoint system: to revert the game to a previous point, you simply rewind along the line of your journey.

Finally, it’s your end-of-game achievement: you can look back, and see how far you’ve come.

The city of Kharé, with the player’s route marked out by markers.

The map became the heart and soul of the game, and as we adapted we used it more and more, moving as much choice as possible from the story layer to the map layer, and adding additional interior maps to buildings, so players could explore room-by-room. 

Then in Part 3, we took the leap that truly puts the map first, by going open-world and throwing away the gamebook’s ‘always forwards’ structure. From Part 3 onwards, players could go anywhere they liked, by any route, and we were even able to add in direct map-manipulation gameplay, allowing the player to remix the landscape itself before setting off to explore what they’d made. 

There are quite a few mechanics in Sorcery!, from the combat system, which is a mixture of rock-paper-scissors and the Prisoner’s Dilemma, to the gentle memorization required for spellcasting, to the spirit animal Gods that reflect your character. But none compare to the map. The map is what turned Sorcery! from a gamebook into a game.

3) Great collaborators

As a small studio -- two people when we started the series -- we live and die by those who fill the gaps we can’t cover. Obviously, there was Steve providing the original game concepts, the world and its curious semi-humorous, semi-dark tone, and a whole lot of words. There was also John Blanche, whose bizarre, idiosyncratic original illustrations pop-up throughout the game, delighting old-school players who recognize every image. 

But we’ve also been lucky enough to bring in some excellent people of our own, who’ve all left their mark.

First up was Eddie Sharam, who did the character art, and was given the unenviable challenge of bringing some of Blanche’s maniacal original characters to life, such as the Sightmaster:

John Blanche’s Sightmaster, and Eddie’s version mid-combat.

Then there was Graham Robertson, whom we brought in to write a few sequences for Part 3 and liked so much that he wrote a good two-thirds of the basic content for Part 4. Graham worked remotely, and taught himself the darkest gizzards of ink scripting in order to write for the game’s open-world ad-hoc layout, while at the same time mastering the Jackson prose style and coming up with a few original scenarios that are the creepiest and/or funniest in the game. Should you find yourself worshipping a whale-head, being entranced by cats, or breaking a poor Goblin’s heart, you will have Graham to thank.

Iain Merrick, an ex-Googler and a very talented coder indeed, handled the Android and desktop porting, and has been doing technical buffing on all platforms ever since. Every now and then Joe and I peer at an aspect of the game -- most recently the swirling clouds in the prologue sequence -- and remark, “Has that got better somehow? Has Iain done something?” Perhaps he has, perhaps he hasn’t, but at this point we’re convinced enough to give him credit for pretty much anything.

But the special mention has to be for Mike Schley, our cartographer. (He does illustrations in general, but maps are one of his specialties.) Mike was an incredible collaborator across the series. His work defines the game’s core visuals: when you think of Sorcery!, you think of the map, and Mike’s work in bringing our world to life. 

To give an example of what he did for us, here’s the Sorcery! 3 draft map we gave him (along with copious notes):

Our sketch map for the wilds of Kakhabad...

And here’s what he sent back:

… and Mike’s version of the same.

“Post-Mike” became a phase of the writing, in between “first pass” and “beta”. We’d write the game, sketch out the map, get Mike’s version -- and then go back through, filling in all the interesting extra locations he’d invented along the way. Post-Mike was when, for example, we decided to let the player explore the more interesting-looking buildings of Kharé, take the shortcut alleyways he’d added, and visit the additional rooms he’d included in the building interiors. 

Our final hire was composer Laurence Chapman, who wrote the tremendous theme for 80 Days. We brought him back to do themes for Sorcery!, including the parts we’d already shipped. We think the results are magnificent.

4) Four games, rather than one with unlockable episodes

When we launched Part 1, the most common question (after “Android?”) was “Why didn’t you make the episodes available by in-app purchase?”