This ARG Hub has been archived and locked to preserve historical integrity. Go back to the Main Dashboard

From Portal 2 ARG Wiki - Valve ARG Network

GLaDOS@Home is a spoof of other distributed computing efforts like Folding@Home. The site at http://aperturescience.com/glados@home became live on April 15, 2011 at 9 AM PDT - the time predicted by the Bathysphere puzzle and the Countdown timer from previous ARG steps. The page's stated goal is to reboot GLaDOS, which would take a considerable amount of time to complete (by default, ending on Portal 2's official launch time, April 19 at 7 AM). By using "distributed" resources - Steam users - playing the thirteen games in the Potato Sack, and by collecting the hidden Potatoes within them, players purportedly will be able to reboot GLaDOS earlier, effectively unlocking Portal 2 some time before April 19th.

Contents 1 Potato Idling 2 Potato Countdown 3 How the site works 4 Tactics 5 Priority list 6 Countdown trackers 7 The story so far 8 Theorised formula for timers 9 Auxiliary Power (previously Challenge Status) (The Potato Multiplier) 9.1 Possible Overclock Potato Numbers 9.2 Theories 10 Number of active test subjects 11 GLaDOS@Home Community Edition

Potato Idling

At the moment the potato count keeps decreasing from ~63,815 to ~61,815 and then jumps back to ~63,815.

Potato Countdown

On April 18th at 10:50 GLaDOS engaged "starch-based power cells" meaning the CPUs got overclocked which resulted in a faster computation process. The potato count started to drop from around 463,000 at a rate of about 700/min, projecting that they would be depleted in ~9 hours 40 minutes from the start, or 21:00 PST.

How the site works

The red bars next to each video game represent the total time people have played that game for; this fills up over time.

The number under each bar (current CPUs) shows how many people are currently playing that game, and hence the rate it's filling up. [Confirmed via consulting the Steam stats page]. 9 lucky people are already playing Portal 2

The main bar at the top represents the average of the length of all the red bars. Once this is full (when every red bar is full), we get Portal 2!

The potato number at the bottom shows the total number of potatoes that everybody has found. The number is now decreasing as GLaDOS overclocks the CPUs and makes the bars rise faster. If you get all 36 potatoes by the time Portal 2 is out, you also get a special prize.

When a red bar fills up completely, it turns blue. This acts as a checkpoint of sorts, knocking off a guaranteed chunk of time from the release date. This will only matter though if we fail to fill the main bar up in time. Note that once a bar is full up, there is no longer any point playing that game; switch to another.

More popular games require more people in order to fill up at the same rate; it is the proportion of game owners playing a game that matters, not the number of people by itself. [Also confirmed by Doug].

We have word on how the progress bars work from Dougley:

 The individual game progress bars are based on sales made. The more times a certain game has been bought, the more you have to play it to finish its bar.
 For example, 1 2 3 KICK IT! was released at (nearly) the same time as the pack, so it has the fewest sales making it the easiest to complete out of the whole pack.
 Killing Floor and Audiosurf were out for longer and as such have had way more sales, so they require a ton of more hours to complete.
 According to Dougley, Potatoes still matter and are of importance (but don't neglect playing the games). Get them.
 The Store might be keeping a record of how many times a certain game has been bought, so we need to go find that and order which games we should play based off of that.

Tactics

In order of priority;

1. Constantly run a game in the potato sack for as much time as you can (leave it on when you're away from your computer!), preferably one at the top of the priority list below.

2.Collect as many potatoes as you can (guide)

It follows from the above that we should be doing the easiest (least popular) games first. Firstly, this means we get the guaranteed checkpoints. Secondly, and most importantly, by leaving the hardest games until last, we ensure that we have the maximum number of potatoes when we get to that hard game, making things a lot faster.

Priority list

This is based on which games have bars that are easiest to complete, which we should focus on first. If you have the current focus game, play that; if not, the next highest game.

For reasons why we have the priority list, and why we switch before they are complete, see Switching Games Before Completion.

1. The Wonderful End of the World - finished on 4/15/2011 19:55 PDT --> 50min off the projected launch timer
2. 1... 2... 3... KICK IT! - finished on 4/16/2011 11:55 PDT --> 40min 28 sec off the projected launch timer
3. AaAaAA!!! A Reckless Disregard for Gravity - finished on 4/16/2011 21:00 PDT --> 58min 53sec off the projected launch timer
4. RUSH - finished on 4/17/2011 04:00 PDT --> 42min 49sec off the projected launch timer
5. Cogs - finished on 4/17/2011 10:15 PDT --> 50min 51sec off the projected launch timer
6. Toki Tori - finished on 4/17/2011 15:35 PDT --> 55min 8sec off the projected launch timer
7. BIT.TRIP BEAT - finished on 4/17/2011 20:20 PDT --> 32min 7sec off the projected launch timer
8. Audiosurf - finished on 4/18/2011 12:40 PDT --> 84min 30sec off the projected launch timer
9. The Ball - finished on 4/18/2011 15:35 PDT --> 54min 28sec off the projected launch timer
10. Super Meat Boy - finished on 4/18/2011 20:17 PDT --> XXmin XXsec off the projected launch timer
11. Killing Floor - finished on 4/18/2011 20:25 PDT --> 5min XXsec off the projected launch timer
12. Defense Grid - finished on 4/18/2011 21:25 PDT --> XXmin XXsec off the projected launch timer
13. Amnesia - finished on 4/18/2011 21:26 PDT --> XXmin XXsec off the projected launch timer

Countdown trackers

See also: Countdown Aids

Here is some data, and some calculators predicting when Portal 2 will be out based on current rates. Bear in mind these might be inaccurate because while GLaDOS reports the guaranteed time that has been "knocked off", these pages produce a time which is an estimation of when we will finish all games (based on current rates). This rate is constant around 10 hours before launch time, which is why Potatos are so important, they are our only hope of releasing it more than 10 hours ahead of schedule.

• HasPortal2LaunchedYet.com
• GLaDOSatHome.com
• gladosAtHome unofficial twitter
• http://www.gamingmasters.co.uk/portal/
• http://www.themcgoverns.info/portal/

The story so far

At 17:00 on April 14th, 2011 Audio Clues led to a QR code. It leads to an ApertureScience.com URL with a countdown to 4/15 9:00 AM.

At 09:00 on April 15th, 2011 the countdown page redirects to here, confirming an early release provided enough people log in to Potato Sack games. Track a current estimated completion time here. Note that this is a running estimate based on the current progress rate and not an exact countdown.

At 10:50 PST on April 18th, 2011, new lines of text appeared on the console at the GLaDOS@Home page:

10:50 - Engaging starch-based power cells

11:00 - Reboot safety test protocol initiated...

11:00 - Relaxation chamber locks released...

11:00 - Involuntary hazard mitigation associates have assumed testing positions...

11:00 - Pre-release lethality assessment initiated...

At that time, the potato count started to drop from around 463,000 at a rate of about 700/min, projecting that they would be depleted in ~9 hours 40 minuts from the start, or 21:00 PST.

Theorised formula for timers

This is a Mathematical Theory that attempts to explain how the GlaDOS@Home system calculates the game time.

Assuming that when we reach 100%, Portal 2 is released-

• Originally slated time of P2 release = b
• Actual moment of P2 release = V
• Man Hours of a single user = m
• Number of concurrent Players (NOTE: One game per unique User ID) = c
• Potato Multiplier/Auxilliary Power = p
• Game completion bonus = g
• Time Reduced = t

• thus: t=mcp+g, V=b-t

(Not valid since the potatos countdown started)

Auxiliary Power (previously Challenge Status) (The Potato Multiplier)

This email from a member of the team at Valve confirms that the Potatoes act as a multiplier to each user's Man hours. But we need to reach a certain milestone of potatoes for that multiplier to activate. The actual milestone amount, and how much of a multiplier this will give us is unknown. Please update the Wiki with this information when the milestones are achieved.

Possible Overclock Potato Numbers

• 375,466 Potatoes
• 419,011 Potatoes (4/19/11 being the original date of Portal 2's release)
• 500,000 Potatoes (There is no evidence to support this. It is hoped only because it's a large, round number)

Theories

• It could be possible that we have already surpassed the Potato Milestone, but the overclocking won't activate until a specific progress percentage. That percentage could possibly be 75%. If Valve wanted Portal 2 released within a certain time frame, they would have to put a safeguard against a possible early potato overclock wouldn't they? Thoughts?
  • This is possible. However, Valve is all about service, and to give the illusion fans can release the game early without having that truth would be bad service.
    • The fans would be none the wiser. They can arbitrarily make the overclock potato amount whatever amount there happen to be at the precise moment they want to start overclocking.
      • Not to mention Valve has given no hint as to what the number might be, so it is entirely possible they will simply activate it when they want without a specific figure in their head.
        • If the overclocking starts at a weird number such as 478,391 potatoes, then we will know that it was activated manually.
          • Seems like it was activated manually.
• It is possible that the large number of fake potatoes (obtained through the website without loading the games), is so large they are still figuring out how to handle the large decrease in potatoes when they remove them (because it will be a very large number).
  • So some of the crossed out numbers above may still be the actual milestone, but it hasn't actually been achieved yet due to the cheating.
    • Update: Cheated potatoes will disappear from profiles; however, they will not be removed from the overall GLaDOS@Home count. Removed potatoes that are re-earned legitimately will not increase the GLaDOS@Home count, either. There will be no sudden decrease due to removal of fake potatoes.
• Or it could just be possible that more potatoes == a faster rate. There's nothing in that email that suggests there is some magical number at which the rate suddenly increases.

Update: The overclocking has started (at about 463,000 potatoes). It's possible that it was always going to trigger at this time, with the amount of potatoes collected at this point being used to determine how long the overclocking would last.

Number of active test subjects

This displays the number of people currently playing Portal 2 (confirmed on Steam Stats page). Originally the page displayed "ZERO". Just ten minutes after the overclocking started it changed to "NINE".

GLaDOS@Home Community Edition

Download it and distribute that link like crazy.

• This application solves the problem of needing to get everyone to switch games within a short period of time; by letting GLaDOS@Home Community Edition manage executing the games, we can get the games completed faster.
• It'll automatically run the current focused game, or run the highest priority game that you actually own, without you having to do anything at all!
• The priority list on which it is based of can be found here.
• The source code is available at the Google Code page. Bug reports should go there too.
• A few notes: Close any existing instances of games you have and remember to set the game availability list correctly (it appears on the main form, just untick what you don't have). It also does not auto-update, so check back every now and then in case of a new version.