IRL Portal at Atlassian

So my friend Wes Walser, who works for Atlassian in Sydney, Australia, posted about this amazing thing they just installed at their offices on his Facebook page:

When Atlassian opened up a second office across the street they wanted an efficient way to communicate across locations. What better way than a portal inspired by the Valve video game Portal? Unfortunately, it only transmits light and sound at this point, but it’s still very cool! Check out Atlassian’s blog for more details.

Ceilings and Foundations (On Executing Well, part 2 of 2)

My favorite story about innovation vs. execution starts with two writers in a web forum. Since it was a web forum, an argument was raging. One writer claimed that a novel needed to have an interesting and creative premise to be successful. A second writer clung to the opposite point of view: even a poor or overused premise can lead to a great novel if it is well enough written.

The argument got more and more contentious until the second writer, whose name was Jim Butcher, offered the first a challenge. He said, don’t just give me one bad idea, give me the TWO worst and most played out ideas for a novel that you can think of, and I’ll write a great novel from them. The first writer said, okay, here are my ideas:

  • Lost Roman Legion
  • Pokémon

Jim Butcher took these two ideas and wrote The Furies of Calderon, a very successful novel that grew into a seven book series.

Codex Alera Series by Jim Butcher

Why was the novel successful? Well, he did find creative spins on those lame ideas. But mostly it was great characters, well-crafted plotting, spot on pacing, fun plot twists, and very readable prose. In other words, superb execution. I’ve read the whole series, and it’s one of my favorites.

But what about software products? Can you be successful without innovation? One great example is the game Angry Birds.

The Angry Birds app hit the iPhone in December of 2009 and now, a scant two years later, it’s hard to imagine a world where toddlers aren’t going crazy for them and the founding members of Rovio Mobile aren’t lighting their cigars with €100 euro bills. Was it innovation or execution that led to their success? Well, ask Liam Bowmer, who in September of 2008 released a web game with the exact same game play called Castle Clout. The game was renamed to Crush the Castle and released on the iPhone in September 2009, three months before Angry Birds. It would not be an exaggeration to call Angry Birds a reskin of Crush the Castle. And similar gameplay has been in various games for years, even before Crush the Castle (Scorched Earth, for example, first released in 1991).

Angry Birds

The critical difference that caused Angry Birds to be a world beater and Crush the Castle to be a footnote was execution. Angry Birds’ successful execution can be seen in the details. The graphic design clicked with their audience, especially younger children. The music and sound effects are polished. Small touches like the structures shaking slightly when a new level loads (showing that they can be knocked down, tempting you) excites your interest in a way that Crush the Castle just doesn’t.  Adding to the four principles from my last article, this brings us to our fifth principle:

Principle 5 – Great execution can make up for lack of innovation.

But what about the big innovators? Think about iPhone from Apple. Or sticking with games, what about the success of Valve with the game Portal? Or what about Pixar’s digital revolution in animated movies?

Companies that successfully innovate and then reap the benefits of that innovation have one thing in common: amazingly polished products. The iPhone is a great example. It is innovative, but like Angry Birds, it also shines in the details. The interface is clean, minimal, and intuitive. A lot of thought and user testing was obviously spent on how to make a 3.5″ touchscreen be the window into a satisfying user experience. Valve’s products, besides being innovative, are bug-free exemplars of great pacing and level design. You see that quality in the details with Pixar as well. It’s true they created a revolution of 3D animated movies, but the pure movie making craft that Pixar demonstrates would have been successful in a more traditional type of animation (just not as successful). Watch the first ten minutes of Up or the first twenty minutes of Wall-E and try to disagree. This leads us to the sixth principle:

Principle 6 – Companies who successfully innovate absolutely nail execution.

What’s interesting is how at least one of these companies accomplishes this level of polish. Valve has a strongly empirical approach to game development. They keep games in development much longer than a lot of their competitors and put even early versions of their games through tremendous amounts of user testing.

Quoting Gabe Newell, co-founder and managing director of video game development at Valve:

At Valve, we see our game designs as hypotheses and our playtests as experiments to validate these hypotheses. In addition, we always want to iterate and improve on our work, and we are constantly seeking feedback – through playtests and other means – in order to do so. By making use of a wide array of user research techniques, we will make better decisions and as a consequence, better games.

“We start playtesting as early as we can—as soon as we have something playable. We’ll start with internal folks and then bring in external folks soon after.


This leads us to…

Principle 7 – The best way to assess how well you are executing and to get better is through user testing.

Does an innovation have to be polished right out of the gate? Well no. One interesting example is another game, Minecraft. Released by Markus Persson initially as a public alpha, it certainly had some rough edges and limited gameplay which undercut its innovative game idea. But people saw the promise in it – enough to purchase the for-sale beta at a reduced cost (which included access to the final game when released) – to fund the development of the game. Playing the release version, for all its quirky graphics and sandbox environment, it feels like a polished and complete game. The initial alpha was released in May of 2009, and the 1.0 release became available November 2011. To date, it has sold over 4 million units. It’s interesting to note that Markus Persson had a huge base of players essentially playtesting his product and paying for the privilege! This certainly helped Minecraft become more successful.

Since I’ve been harping on execution being more valuable than innovation, let me step back affirm that innovation is good. It will basically do two things for you. First, it will get you noticed. You’ll have a chance to follow through on the promise of your idea. It will buy you time – not forever – but you’ll have a window. Companies get that chance and mess it up all the time, but it’s a tremendous opportunity that many others have made good on. Second, as long as your execution is sound, innovation will amplify your success – sometimes quite dramatically. Polished execution is the price to get in. But superb innovation can multiply your success.

I like to think of execution as the foundation that will keep the house standing. But innovation, combined with things like business drivers, will determine the heights of your potential success. Superbly execute something that isn’t very interesting at its core, and your potential for success will usually be quite limited.

My last principle, which I’ll allow to sum up this article:

Principle 8 – Innovation raises the ceiling on how successful a product can be, but solid execution is a bedrock requirement for realizing that potential.


Italian Plumbers and Perfection in Non-Linear Time

It goes without saying that there are some pretty interesting subcultures on the Internet. Within the video gaming culture, there’s a subculture of people who try to do speed runs of various popular video games. Another subculture is the ROM hackers: people who modify levels on various older video games like the SNES Super Mario World (often making them a lot harder). When those two subcultures meet, forming a sub-sub-culture, you get something like this:

Seriously, stop to watch the video if you haven’t. You owe it to yourself.

Wow, can you imagine a more perfect and amazing performance? Would it change your mind if I told you that the person who created that video was using performance enhancement?

Welcome to the world of TAS: Tool Assisted Speedruns. TAS is performing a speedrun like the one you see above by continually saving and restoring the state of the game in an emulator. Every mistake you make gets undone, you back up to your last save state and try those last 5 or 10 seconds over again, repeating until you finally succeed at that piece and go on to the next. At the end of 10+ hours of try and retry over the course of several days, you have 8 perfect minutes of Super Mario virtuosity. You can post a video of that 8 minutes, and it looks like a single, uninterrupted narrative. You aren’t lying to anyone about how it happened, everyone knows how you made it, but it is nonetheless considered as a single performance for the TAS community.

Here’s what the process looks like behind the scenes, with nothing edited out:

This is very interesting to me, because I’m not sure how to assess the final product. Should I look at the whole exercise as smoke and mirrors that creates the illusion of perfection? “Yes, it looks impressive and does involve skill, but it lacks purity. Isn’t it like saying a pitcher pitched a perfect game because he had nine perfect innings in different games scattered across the season?” Or is this process of creating genuine perfection from imperfect inputs impressive in its own right?

The Super Mario World game can be looked at as an enclosed system with its own rules of cause and effect. None of those rules are being violated in these TAS performances. They are just not being performed straight through in our perspective, if we watched them being done, but rather spread out in small pieces over a larger amount of time with all of the unwanted bits removed like noise to signal. In fact, when a performance is complete, all of the inputs are played back into the game, which occurs in real-time. This works because the game is deterministic – Miyamoto does not play dice with his universe.

This reminds me a little bit of Dust Theory, which is a concept I recently read about in the novel Permutation City by Greg Egan. The idea, very briefly, is that each pattern provides its own frame of reference, and multiple patterns can be applied to the same data. Neither is intrinsically more correct that the others. This is somewhat analogous to frames of reference in Special Relativity, where different people could experience time at vastly different rates. Only in Dust Theory the events could also by nonlinear or occur out of order. For our TAS, the events making up the performance are peppered across several days, interleaved between the dark matter of much less interesting gameplay.

This concept of undoing mistakes by reversing time isn’t just in TAS, it’s intrinsic to at least a couple of video games. Prince of Persia: Sands of Time had this gameplay as a core conceit, as did the pretentious indie hit Braid. Sands of Time allowed you to do it until your magic hourglass sand ran out. Braid, on the other hand, allowed time to be rewound without any restriction but with a twist: certain objects would not be affected.

Rewinding back from failure is an inviting concept to lure in a broader audience. In a game where any mistake can be taken back, there is no losing – just “haven’t won yet”. That’s probably a better approach for life in general now that I think about it.

This is an idea with mainstream appeal. It’s no accident that the movie Groundhog Day was a huge hit. Bill Murray takes hundreds or thousands of attempts at the same day and finally executes it perfectly. Tell me that wasn’t a TAS.

Personally, I find something a little bit magical about playing with time to become retroactively infallible. Maybe we should just enjoy that magic rather than find a reason not to.