The best thing about Derek Yu s new book about Spelunky is that it answers all the questions I ve ever wanted to ask about his game. The worst thing about Derek Yu s new book about Spelunky is that it leaves very few questions to ask him directly. Part indie development motivational tome, part technical retrospective, Spelunky (the book) is something fans of the game will devour in one sitting, just as I did.

Published by Boss Fight Books, Spelunky traces the game s origins as a free-to-play browser game built in GameMaker, through to the release of the HD remake on XBLA in 2012 and PC the following year. Back then, major indies arriving on console was still a novelty, Early Access and Kickstarter weren t yet phenomenons, and the roguelike had yet to have its unlikely revival, save for The Binding of Isaac.

Spelunky is a divisive game: those who love it sing its praises at every given opportunity, while those who dislike it do so with a belligerent, unreasonable and frankly alarming passion. When PC Gamer named it the best game of 2013, you better believe there was much handwringing involved. For those who love the game, reading about how and why the Hell Run chain was installed, or what inspired the eggplant, will be required reading. Accessible dives into the logic behind the game s randomised dungeons is also present, as is a sort-of manifesto on Yu s preference for game worlds that are indifferent to the player.

I spoke to Derek on Skype a couple of weeks ago. His book is out now through Boss Fight.

PC Gamer: Was it as difficult writing a book as it was making a game?

Derek Yu: Yeah, it was pretty difficult. I mean, going into it I hadn t written anything over a few pages long in college I remember writing papers that were about three pages long. The sheer enormity of having to write something that long was pretty daunting. Mainly, I just realised I didn t have as many writing tools as I thought I had and I don t mean keyboard. [I mean] the vocabulary, the phrases that one needs to put a book together of that length. You need to find ways of saying things that sound interesting, but you need the variety as well. The final book I think ended up being 48,000 words or something like that, so writing and having it edited by professional writers, I realised how many stock phrases and words I really rely on day-to-day. You need to come up with more interesting ways to say things. Putting your thoughts out there is one thing, but putting your thoughts into a form that reads well and is interesting is another.

Then there s the difficulty of describing game design and gameplay using words and making it sound interesting: I ve always had a hard time with that. In the book I started with how I would introduce Spelunky to people when I met them for the first time. That s always been a difficult thing for me because there s so much in the game, and boiling that down into a few sentences, even for people who are familiar with games, is hard. If you think about a lot of classic games, if you try to describe them in words for readers who have no background with them, [it can be hard]. Like Mario: you re this little guy breaking blocks. It sounds weird and totally not fun at all. That s probably why, when we describe games, we rely so much on it s X meets Y , this is the Dark Souls of strategy games, or something like that. Because it s so hard to describe in words what playing a video game is really like and how that s exciting. All of that was very challenging.

PCG: Do you still play Spelunky?

DY: No, not really. I played it so much during development and immediately after release. I d be interested in how many developers play their games. I mean, every now and again I ll play a little bit, usually just to check something. When I was working on the book and wanted to check something, I d play the game. You have to realise I ve been playing Spelunky since the original freeware version in 2008, and I played it until the release of the PlayStation versions of the game. I d been playing it non-stop.

PCG: Sometimes I think I m done with Spelunky. I leave it for two to three months, but then something happens and I m sucked back in.

DY: I think the other thing for me is that it still hasn t been long enough that I can play it for fun. When I m playing a game I m still always a bit on edge in terms of thinking about it, wondering what s going to go wrong, thinking about how I d make improvements. There s not really much I would improve with Spelunky, but the feeling that I should be looking for those things is still present. At this point if I wanted to play a video game for fun or to relax, I m not going to play one of my own games.

PCG: You say there s not much you d improve about Spelunky, but that implies there are small things. Is that true?

DY: Yeah, I don t really think there are, to be honest. I think Spelunky as it is, is pretty set. Another thing is, as time passes even things that may have been flaws become part of the game. I could say, maybe I want to improve the graphics because I m better at drawing now , but you start to get used to something the way it is and has been, and you don t want to change it because the flaws become part of its character and personality. Aside from maybe some small bugs lingering in the game, I don t think there s anything especially anything fundamental to the design.

PCG: Since Spelunky HD released on XBLA in 2012, Early Access has become popular. It somewhat mirrors the way you initially developed Spelunky in GameMaker, seeking feedback from the TIGsource community. Do you think the rise of Early Access is a good thing, or not?

DY: I think it s good. I personally wouldn t ever do Early Access. At this point I feel like I wouldn t use it unless I really had to for some reason, because I wouldn t want that pressure of having people watch the development as it progresses. It would be more difficult to make the decisions I need to make. It s the same reason I probably wouldn t do a Kickstarter: I think a lot of people do it not just for the funding but for the marketing and promotion, but I think right now, with the sales of Spelunky and Aquaria, I can fund my next game myself and choose not to go the Kickstarter route, even for the promotion. Investors make it harder, and adds more pressure it gets people s expectations going before you really want them to have any expectations.

Expectations are so critical. I think that people s expectations of the remake of Spelunky, based on the original, coloured the way that they saw the game when it first came out. I think there was a significant amount of criticism of the remixed music and graphics, because people were so used to the pixel art and the original music. If that original game didn t exist, those people who didn t like the new graphics and music probably wouldn t have felt the same way, without those expectations. Early Access can be great, but you need to weigh the trade-offs very carefully. For people who need the funding, or for people who need the motivation… I think it can be hard to make a game that no one will see for years. I even released the freeware version of Spelunky as a beta on TIGsource in part because I wanted the motivation. I didn t know if it was good. But you may not need any of those things.

PCG: Late last year there was a lot of chatter about  a so-called "Indiepocalypse", and the gist was that it s much harder to get any exposure as an independent developer nowadays, because the market is so saturated. Do you think that s true?

DY: It s hard to say for me, because I ve released Aquaria and Spelunky, and having already released some successful games really makes a huge difference. I mean, I think Jon Blow said it himself about The Witness, that because of Braid a lot of people checked it out. It is hard for me to speak on what it would be like, or what it s like for new developers right now.

Things like Indiepocalypse and this is the end of indie : those kind of phrases I don t really buy into. I think there are definitely a lot of new challenges now that indie gaming and gaming in general is a much larger and more diverse place. That said, I think there are a lot of benefits to that, the tools are better, and I think there are a lot more resources available for people to use to make their games, to distribute their games and to learn how to make games. I definitely don t think indie gaming as a whole is in trouble. I think there are some new challenges but there are some new positives. I think the competition level is definitely higher, but the tools and resources are better.

PCG: Do you think there are too many games now?

DY: It s hard for me to say I wish that there were fewer games . I just want to see more and more games. The more the better. It s definitely more crowded though, and I will say that it was probably easier to get noticed when I was working on Aquaria and Spelunky. That said, it also seemed scarier then because not as many people were doing it. It felt more lonely, it felt more like… what are we doing? Is this even a legitimate thing to be doing as a career? How are we going to sell our game, and will people even want to buy it, compared to the big blockbusters? There were a lot more questions. That said, it also felt very cosy, and when you did meet other indies at GDC or wherever, it really felt like a family. Now I think there are so many more people, it definitely feels more crowded now.

Next page: Derek's favourite Spelunky character, and why deathmatch mode never caught on.

PCG: There s a section in the book called feedback loop where you write that Spelunky was a small idea that quickly grew organically. That seems like an ideal situation for a creative person. Is that a hard phenomenon to trigger? Do you work for it, or does it arrive in a fit of inspiration?

DY: You definitely have to put in the work. Start with small ideas and try making them. What I found is that I d often hit a wall, so I d go back and think about what I m going to do again, find a new idea, mix and match some existing ideas I ve already had and go back at it. You hit a wall, and then you hit another, and eventually you get to an idea where everything seems to fit and you just start flying.

That s kinda how it was with Spelunky: I had all these prototypes before Spelunky, and they were all really exciting at the start but pretty soon you hit a wall. Once I started working on Spelunky I never hit the wall, I just kept going and going, and there kept being more interesting avenues to explore with it. I don t think I would have just come across it without working on the previous prototypes, which meant sitting down and making games. I definitely advocate actually making games, getting your hands dirty, because that s the only way to really figure out what sticks. There are so many ideas that sound good in your head, but when you start working on them they aren t [that good].

PCG: Spelunky's Deathmatch mode is one of my favourite competitive games. You write in the book that you were surprised people didn t pick it up more. Why do you think they didn t?

DY: It s a good question. I think it was because people just didn t play it the way Andy [Hull, Spelunky programmer] and I did while we were developing it, where it was just much more tactical. We didn t chuck bombs all over the place, we d wait for that perfect opportunity and try to take out the person when they were vulnerable. Whereas I think most people who tried deathmatch, and maybe multiplayer in general especially for a game like Spelunky where the deathmatch was never really a part of the core game to begin with I think expectation played quite a big part. I feel like that s on us, to get players to understand how to play the game properly. I definitely don t blame the players or anything like that. I think a lot of people do have a lot of fun with it, as a more casual thing. It may also be that adventure mode is more compelling than deathmatch mode.

PCG: We ve got a tournament going in the office. We play 1v1, 12 bombs, 4 ropes, and everything else on default except the ghost, target and AI is toggled off. We find it very tactical with those settings.

DY: It s possible that there are people out there having a lot of fun with deathmatch and it s just not really being broadcast on the internet where I would be able to find out. I tend not to watch a lot of Spelunky streaming anyways, kinda for the same reason I don t play it: it s just slightly stressful for me. I usually just watch the highlights people send me. I think it s also the case that deathmatch isn t a very well-explained mode. It s just there and it s up to you to figure out what to do with it.

PCG: There have been some impressive Spelunky runs: someone killed the ghost, someone got the eggplant to Yama, recently there was a no gold pacifist run. It feels like the community has done everything in this game, but is there any other devious challenge you could set?

DY: Oh boy. I don t think there is! Nothing that s really within reason. I joked about people playing it blindfolded but I don t really suggest that, actually. I don t think Spelunky has enough audio cues to let you do that without a tonne of luck. I said it in the book: people have already completely surpassed our expectations in terms of what can be done in the game. The limits of the game, or what we thought were the limits, aren t the limits at all they ve been surpassed by people like Bananasaurus Rex. At this point anything people do is new and surprising to us.

PCG: Did you expect Spelunky to be such a great spectator sport?

DY: I didn t, though I hoped that it would. With Spelunky Classic, I really thought the random nature of the game would limit it as a competitive game, but it s not been a problem at all if anything the random nature has made it more exciting for speedrunners and challenge runners. Someone released a speedrun of the original Spelunky and broke down exactly what they were thinking at each spot, and talked about how they used the randomisation to their advantage, how they accounted for the randomisation. There is such a thing as increasing your luck with the randomisation by doing certain things, and that gave me hope that Spelunky would be pretty competitive, despite or because of the randomisation that s involved in it.

PCG: You begin to have an innate feel for how the levels are constructed after a while.

DY: That s the thing that I didn t necessarily know would be true, working on the game. I think I had a certain intuition about it too, it just seems to work because despite the randomisation it s not complete chaos. There s enough in Spelunky that s handmade versus random or procedurally generated that from run to run, you never see the exact same formations, but you find familiar ones.

PCG: I ve found in my time playing Spelunky that the Black Market is the easiest place to kill the shopkeepers, thanks to the Ankh room. It s very easy to lob bombs safely down there. Was that deliberate?

DY: It wasn t deliberate but it s one of those things where I quickly realised that it s easy to get them there. I didn t plan it that way per se, it s just one of those things where it happened to work out, which was the case for a lot of Spelunky. Intuitively I design things a certain way without understanding necessarily why I was doing it that way, but after playing it or having other people play it and watching other people play it, I kinda understood why I did it. As I m working on a game, I don t necessarily have these strict ideas in mind, it s more like feelings. It s why you have to make a lot of games and spend a lot of time doing it, to build up that intuition. Then I think making games feels more like drawing, which also feels more right and satisfying, and you feel like you re putting more of yourself into it. It s better than thinking in terms of nitty-gritty design rules.

PCG: I know that Spelunky is feature complete and you ve moved on, but is there any potential to get Steam Workshop support for it? There are third-party map creation tools out there.

DY: To be honest I haven t thought too much about it. Andy Hull (programmer for Steam + Xbox) would have to want to do it. I m certainly cool with people using editors and stuff to mod the game with, but I don t know about Steam Workshop there would have to be a lot of demand for us to go and muck around in that code again. At this point we re pretty happy to move on.

PCG: Are you working on something at the moment?

DY: Yeah, we re working on a few different things, but unfortunately nothing is far enough along that I can really reveal what it is. The book has taken up a lot of my time during the past year, or over a year. Now that s winding down I m spending more time on games again.

PCG: Are you working alone on the prototyping and conception?

DY: A little bit. It s hard to say that I m in solitude now because I have a daughter, so I feel like I m never alone anymore. I m also working with other people on things. I m doing a little bit of the totally solo prototyping which I did when working on Spelunky, but it s definitely not the same. So much is different now. One is that, like I said in the book, Spelunky really kind of came about because of Aquaria in a way, because I wanted something that was different from Aquaria. That was such a difficult development for me and Alec Holowka [co-creator of Aquaria], so I really needed to decompress and make something very small. I don t feel that as much with what I m working on now, because Spelunky, even though every game development is very challenging, I think the development of Spelunky was a challenge that was much less exhausting for me on a mental and spiritual level.

PCG: What games are you playing at the moment?

DY: I haven t been playing as many games as I have in the past. Mostly I ve been playing smaller indie games and mobile titles, because those are the ones that fit my lifestyle more now. Being a parent, I spend a lot of time parenting and when I have free time I want to work, and so my time for playing video games is a lot more limited. I have to sneak it in. Most of last year I played a tonne of Puzzle & Dragons. I think a lot of people saw me Tweeting about that, and thought why are you playing that? It s actually a really good game. It s quite rewarding in terms of requiring a pretty high level of skill to master and all that. There s a lot of variety and the fundamental mechanics of the game are very interesting, it s an extremely innovative extension on basic match three mechanics. I don t play it as much now, but last year I played it a lot.

Other than that I ve mostly been trying smaller games out. Though I was playing The Witness for quite a bit. That s the thing, longer form games I have trouble with because I ll enter a period where I can t play it much and it s hard to get back into. What I played of the Witness was really good.

PCG: Who s your favourite Spelunky character?

DY: You know, I think I usually just play as the Spelunky guy, the main character. I actually like the little purple girl with the glasses.

PCG: She s my favourite, too!

DY: I also like the purple pirate girl. There s something about purple. Usually I play as Spelunky guy, though. But I do like the underdog characters, I like to play the more interesting or underused ones. I used to play Ken in Street Fighter 2 a lot, but I moved to Dudley in Street Fighter 4 because he was a much more interesting and underused character.