computer games design games design games industry massively multiplayer

A little game called Minecraft…

I’m an idiot. It only just occured to me today to wonder how common a name “Markus Persson” actually is.

My first reaction on playing with Minecraft was: “Ah, this (much success, this quickly, with this style of game) is what the WurmOnline guys would have loved to achieve, I think”. Oh, the irony. Not that I expect the WO guys are unhappy with what they’ve achieved – WO is an achievement in and of itself, and continues to evolve in beauty and depth – but it’s too slow / too little to become a mainstream success. It’s a technical and personal success, not a business success. Unlike MC.

Now that I know it’s the *same* Markus – Wurm’s Markus_Persson from, and the author of Minecraft – I’m revisiting that first impression. Incidentally, in checking out this authorship, I saw I still have the 2nd highest post count on JGO (!), despite almost 5 years of personal absence. Wow. I really did talk a *lot* of crap, I guess ;)…


There are two things in MC that jumped out at me early on. The first one was pure simplicity. This game/world is bursting with a sense that someone has ended up with nothing left to take away.


The art has a nice style in and of itself. This is something I’ve often talked about with other veterans of making cheap games … mostly, that means indie developers (many of the folks from JGO, for instance – especially look at the 4k Java games contest, and look at the frequent winners such as Kev), but it also includes folks from the big-budget games industry, like Thomas Bidaux and Ken Malcolm. And, of course, Matt Mihaly, whose Earth Eternal started out explicitly using an art-style to keep production costs low.

It’s a trick. It’s the avoidance of the uncanny-valley. Make your art look deliberately cheap, instead of accidentally, and you can achieve the same level of pleasure in your audience as if you’d built photo-realistic graphics. As far as I recall, it’s even been tested (in movies and visual art) and found to be literally true.


Here’s one key point where Wurm fell down … there was a vast amount to do, but it was damned hard to know WTF you could do, how, when, why, where.

And at this point, I’m going to pull out and dust off Runescape, circa 2001. Back when it was a few thousand players, and was already taking off, but long before it became famous. RS has changed plenty over the years, and has finessed their interface, but it started off with Andrew’s unrelenting insistence that the UI must have “no menus”. Everything had to be achievable with the LMB or the RMB. There were already examples of places where the interface was crippled and made complex because 3 (or 4) buttons were needed … and eventually Andrew relented and allowed for RMB to be a menu, with LMB being the “most likely option from that menu”.

The point being that if he’d not clung so hard to that desire for ultra simplicity in the GUI, then he’d probably have ended up with a series of rich menus – or more likely a many-button interface. Sure, it had to be modified as the game grew, but IMHO it was one of the greatest attractions to RS, and helped enormously with RS’s growth and success among the school-age market (where RS thrived).

Now back to MC. In MC, your actions are very narrowly limited. Indeed, they reminded me a lot of early RS. Everything relies upon context. In MC’s case, the compromise with game-richness comes in the form of the crafting interface. Here you have to step beyond the LMB/RMB setup, the “purely contextual” actionset, and move to a new UI element. It’s new, it’s extra UI, but … it’s still brutally simple, and yet (so far) proving more than adequate to the enormous demands of variety in MC.

… and logic

This is the second one that struck me, and it took a bit longer, a bit more playing around (and watching other players in their more advanced worlds), for me to spot.

Wurm Online was always one of the richest “physical universe simulators”: there was a huge amount of physical laws implemented and underlying everything you saw on-screen. This is not a good path for a developer to take: it’s a steep slope into exponentially large CPU and content-production costs … and even worse in terms of balancing and game-design.

Oh, you think that a world with “full physics” has no content production cost? Ha! How much time do you think it takes to implement each of the laws of physics? Even with a rigid-body physics engine to start from? May seem like there’s not many of them around, but just try coding it…

But many of WO’s laws were barely noticed by the majority of the players, while others were smack in their faces a large amount of the time. MC very nearly appeared to cherry-pick only the “frequently significant” laws, and focus on those. MC’s world is breaktaking in it’s depth, and yet if you analyse it closely, you quickly notice it’s lacking some very basic essentials of a world-simulator.

And, back to the irony, I felt that last point made MC feel very much like a direct sequel (in spirit) to Wurm Online: a “lessons learned … and acted upon” when it came down to the most addictive and engaging (and unique) part of WO. And yet I was still too dumb to connect the names together. Doh!

2 replies on “A little game called Minecraft…”

I’m thinking of MC as Lego. It is really that simple, sandbox as it is. Like, you have one universal main rule, to rule it all.

But I don’t think that MC is a some sort of home work on mistakes done right. More like really huge, unconscious insipiration. Without it, we’d have just another one WO (haven’t played it, but heard about, and I understand what you talking about).

And as another one side of success, is Notch’s sight of how to do stuff: openness, honesty, etc. Yeah, while we can dispute about that one, the MC community will just continue to hail his name.

Hey Adam; Mike aka Catharsis from JGO here… I’ve been enjoying catching up with your posts. I’m very glad to see a long time JGOer / Markus succeed. And um yes you posted a lot back in the day though definitely were more insightful than most “post-a-holics”.. “What’s b^3 up to today,” turned into a daily spectacle.. ;P If you make it back out here for GDC make a post on the blog and I’ll try and get in contact. I’m about to launch the long running Java middleware / engine / component oriented platform for real time app / game dev called TyphonRT I’ve been working on for like ages way back in JGO days when I posted more there. I should have it out by GDC. I never released it due to no real ROI for desktop Java engines, but have heavily worked on it over the last 2 years with Android’s emergence as now there is ROI potential and even gasp a business model perhaps; something I’ve struggled to find over the years with my long term tech ambitions. BTW (the large scale 3D sound studio / demo room is complete now! You last stopped by before things were finished). TyphonRT provides a runtime compatibility layer over Android and J2SE allowing write once run anywhere possibilities at least for GL/ES apps. Android really pushed me forward to work on latest architecture techniques. TyphonRT has a component oriented entity system very similar to your articles and things are structured in general in a massively COP direction at the platform level (it’s split into ~400 discrete components) and with a lightweight component API for app dev. I also plan to put out an extensive tutorial series on real time app / game dev along with a lot of content on modern software architecture dumping my last 10+ years of experience. We’ll see how all this goes… Cheers.

Comments are closed.