ARGh! GLaDOS@Home & The Collapse of Reason

I, like many, was late to the Portal 2 ARG party, so allow me to briefly catch you up. Over a couple of weeks, clues and challenges hidden within thirteen indie games on Steam were accumulated, potatoes were awarded on people’s profile pages and finally a URL was deciphered. This URL contained a countdown and when it reached zero, anxious participants across the world were greeted with…

…a countdown.

This time, there was a surprise. Here was the deal: the more we play those thirteen games, the earlier the projected release of Portal 2 would get. So began the brilliant marketing ploy that was GLaDOS@Home. What followed, for anyone watching the more dedicated of participants, was a masterclass in the human capacity to spot patterns in noise, identify conspiracies where there are none and engage in the most spurious wishful thinking.

At first, people took to the task diligently, hoping to gather data on what the effect of filling up the progress bar of each game would be. Some of the technical aspects that sprang up so quickly were truly brilliant and anyone who was interested enough could find a host of graphs (see below, taken from cpucount.zapto.org) and statistics on how each game was progressing. Chrome plug-ins, websites and even an application which carefully controlled which Steam games were being run on your computer while away; it was an impressive effort in many ways.

Then, the dissatisfaction began to spread. The countdown was not progressing as quickly as people had hoped. Expectations continually lowered, starting with a hope that it might be ready to play on the weekend and dwindling to doubt that it would be released more than a couple of hours early.

Largely gathering on a few IRC channels, as people got more desperate for good news they began to cling to anything they could find or, failing that, just make up in their heads. At the bottom of the GLaDOS@Home page was a count of the total ARG potatoes earned, which people watched like hawks.

First it approached 200,000. “Surely something will happen at 200,000!”

Then 300,000. “Surely something has to happen at 300,000!”

Then it got a bit weird.  “Look, the total number of potatoes per person is 36, so if you multiply that by 10,000 you get 360,000. That seems like the most likely number for something to happen.”

360,000 passed, but there was another, even stranger hypothesis hot on its heels. The numbers 88 37 54 66 were seen in a couple of locations. Through a series of mental gymnastics of truly Olympic standard, it was decided that because 88 was the ASCII code for X, 375,466 was the potato count to aim for.

I felt like I had to say something, so I pointed out along with others that the logic being employed in these guesses was so tenuous that it was becoming ridiculous. It brought to mind that apt statement from the Sherlock Holmes stories, that “It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.”

The response? “Well, what’s your theory, then?”

No, no, no. Just because I don’t have an idea doesn’t make your flawed thinking any more likely. Reasoning doesn’t work by clinging to the first idea you like the sound of with blind faith, because you can’t come up with anything better. I felt like pulling a dozen equally plausible hypotheses out of thin air to demonstrate how silly it was to build up so much hope on so little data. Then again, that might have just encouraged them.

This continued past 400,000 potatoes (“Surely something MUST happen at 400,000!”) and represents but a small sample of what was being bandied about. By the final stretch, things were taking a turn for the truly bizarre. A not-insignificant portion of people declared that they had figured out that we all have to stop playing in order to win, as if this was at all plausible in concept and achievable in practice.

At first I thought it remarkable that these were the same people who had so successfully deciphered the clues leading up to the weekend. On reflection, perhaps it was the case that the more sensible, measured voices in the crowd were being drowned out in the noise. While there were people in charge and trying to make the best of that position, opportunities for intelligent discourse were throttled in the attempt to get thousands of people to coordinate their game-playing efforts.

In the end, as anyone who knows Valve ought to expect, they had a plan. Those potatoes everyone had earned were consumed at a particular time in order to boost the power and fill those bars up faster. All but the most jaded were roused from their frustration, while others grumbled that if we’d gone along with their plan we would have been a lot closer to winning already. But the gears were in motion; it was too late.

Admittedly, this gave a sense that the outcome was partially rigged from the beginning and we didn’t quite get the significantly early release we were dreaming of. Nevertheless, I confess to being rather relieved to find Valve patiently pulling the strings. It’s a familiar feeling and I’m not too proud a person to be their puppet, at least for a while. Somehow, it’s comforting in comparison to the mess I had experienced in the days before.