World of Goo Vs. Piracy

A mature and helpful addition to the discussion.

2D BOY have posted about the piracy World of Goo has suffered since launch. After Ron Carmel mentioned the figure – 90% – in a comment on RPS this week, the story was picked up across the internet. And it’s shaken a lot of people.

Carmel explains how they reached the dramatic figure.

“first, and most importantly, how we came up with this number: the game allows players to have their high scores reported to our server (it’s an optional checkbox). we record each score and the IP from which it came. we divided the total number of sales we had from all sources by the total number of unique IPs in our database, and came up with about 0.1. that’s how we came up with 90%.”

He then goes on to point out the possible inaccuracies: people installing a legitimate copy on multiple computers (I know I have), dynamic IP addresses, multiple pirated copies behind the same firewall, and people not checking the high score box, all of which could push the figure in either direction. But of course, it’s unlikely it would be that far each way.

Edit: It’s pointed out below that dynamic IP addresses could have a dramatic effect on this figure. I’ve no idea at all, but it would make sense that this could skew the results.

You might assume that the response to this – by any developer, let alone two guys on their own – would be to conclude that some manner of DRM was necessary. 2D BOY used World of Goo as an experiment, deliberately releasing it without any form of DRM, and said they would publish the results. And here they are: a 90% piracy rate. It’s got to hurt. As hypocritical as it may be, I find myself feeling far more angry that people have chosen to acquire this game without paying, than I am when it’s from a large corporation. It shakes my (personal – not the opinion of RPS) belief that DRM is both pointless and damaging. But amazingly, not 2D BOY’s.

Carmel links to Russell Carroll’s article about the 92% piracy rate of Ricochet. This is where that famous, controversial figure appeared, arguing that 1000 pirated copies of a game only represents one lost sale. Carmel continues,

“in our case, we might have even converted more than 1 in a 1000 pirates into legit purchases. either way, ricochet shipped with DRM, world of goo shipped without it, and there seems to be no difference in the outcomes. we can’t draw any conclusions based on two data points, but i’m hoping that others will release information about piracy rates so that everyone could see if DRM is the waste of time and money that we think it is.”

A remarkable response. But also a rational one. It’s so very impressive to remain rational in the face of learning how many people are taking the game you worked so hard on without giving you any money. It deserves enormous kudos.

Meanwhile, Spike TV’s Video Game Awards show has added a Best Independent Game category, in which World of Goo is nominated. It’s open to a vote, and along with Goo you can vote for Audiosurf, Braid, and PixelJunk Eden.


  1. MeestaNob! says:

    It’s a shame they cant provide more accurate figures to make their point, although (sadly) I don’t doubt it’s more likely around 65-70%.

    Either way, if you played it and like it, pony up some fucking money you tight arses. It’s only $20.

  2. Feet says:

    Piracy! Booooo! DRM! Booooo! Games! Yaaaaay!

    Comments thread complete.

  3. The Sombrero Kid says:

    i don’t know anyone who bought the game who doesn’t run it on more than 1 pc i also don’t know anyone not running it at a static ip i would expect legit copies to average at 3 pcs with an escalating number of ip’s say after a month an average of 3, so i’d say there should be 9 ip’s for every user as a very rough guess, bearing in mind that this game of course runs on laptops.

    this method however is leagues ahead of the add up all the downloads of the torrent and multiply by dream pounds approach as far as attempting to track this kind of thing goes i can think of few cheap better options

  4. Gap Gen says:

    Yeah, I mean it’s pretty pathetic to pirate a £10 that’s downloadable from the maker’s website. It’s not even like (I imagine) people pirated it to try.

    Has Steam been cracked? Obviously it’s the less attractive option for both consumers and for the makers (as income per sale probably goes down) but if 90% of people are indeed utter cocks, then it may be the safer option.

  5. bobince says:

    It’s a shame the “90%!!!” headline has got so much traction and will be breathlessly quoted as fact without the caveat that we still really don’t know.

    My own hunch is that the dynamic IP issue is likely to be a much bigger factor than the other sources of uncertainly mentioned and won’t be ‘cancelled out’, but there is no way to tell given only a list of IP checkins.

    If the checkins also logged which installation they came from (either by unique install ID, or profile setup maybe?) it might be possible to exclude that factor and come to a more accurate (still inaccurate, but better) figure.

  6. Feet says:

    I agree with Jon, it’s admirable that 2DBOY have remained rational about these numbers, despite the nasty picture they paint. Whether it’ll make any difference to the way DRM is employed, who can say.

  7. wien says:

    Yeah, I think it’s important not to get too carried away with the 90% figure. With all the variables involved in their method of measurement the 90% would be the absolute worst case. It’s not really likely to be the true percentage.

    Still sad though…

  8. Quirk says:

    I’m tempted to use strong language here.

    This is Internet 101, people. A very large proportion of internet users are on dynamic rather than static IPs – their ISP provides them with a new IP every time they log on. Get 10 registered users on such ISPs to log into your servers 10 times each, and you’ll quite possibly get 100 unique IPs (provided there are no duplicate IPs being handed out). To deduce from your 100 unique IPs and your ten registered users that you’ve got a 90% piracy rate is to fail spectacularly. The prevalence of dynamic IPs makes any correlation between unique IPs and users all but impossible to make.

    It reads to me like they’d forgotten about dynamic IPs or something and published based on those results, and then realised that what they’d said was utter nonsense; but instead of taking it back, they said “dynamic IPs might be a factor”.

    In other news, people with small feet tend to earn much less money per annum than people with larger feet. Possible confounding factors: children tend to have smaller feet than adults, and tend not to be in full-time paid employment.

  9. The Sombrero Kid says:

    forgot to mention this, totally agree 2D Boy are awesome and an inspiration for aspiring indie developers.

  10. Jerricho says:

    How many unique usernames are there on the leaderboards? What’s their ratio against total downloads, unique ips and purchased copies?

  11. A-Scale says:

    To be honest after playing the WoG demo I really don’t think it’s worth the 20 dollars. It’s a very quaint and fun game, but you can get similar gameplay in free flash games. For a game of this low level of complexity I hoped that a 10 or maybe even 15 dollar (US) price tag would be in order. It’s a lot easier for pirates to justify stealing a game if it is overpriced. I’m certain the author put a lot of hard work into this game, but I think people would be a lot more apt to buy if it cost half as much.

    This should NOT act as a deterrent to publishers who are considering putting out no DRM games. ALL popular games, DRM or not, will be stolen. The trick is to put enough quality into a game at a low enough price that people can’t help but feel bad about stealing it. Also, I don’t even know who the author is, and I read RPS 5-10 times daily. If the author is unknown, it doesn’t even feel like stealing. Put a face to your game, and let people know that you’ve got a family to feed.

  12. cyrenic says:

    Kudos to 2DBoy for keeping things in perspective.

    At first I thought they should have released the game through Steam only. But now I think if they would have released only on Steam it would have hurt their sales, since it would not have been as widely available (region issues particularly). If you adhere to Stardock’s principles, you want to ignore the pirates and release the game as widely as possible. And it would have been trivial to crack the Steam version and put it on torrents like it is right now.

    I do think they should only offer the 6th chapter to paying customers though. Might be a bit late to implement that however.

  13. Quirk says:


  14. Aldaris says:

    I guess you have to look at it the other way, at least 10%, probably 30%+ bought it.
    And for 2 people, that’s still a lot of money, as well as the satisfication of helping so many people enjoy their time.

  15. Bobsy says:

    Ha. In that Spike TV awards thing Spore is in the category for best original score. No-fucking-way.

    “Blip, bloop. If I keep this up I can convince people that this isn’t midi!” – Brian Eno, 2007

  16. Ian says:

    Even if that number is greatly exaggerated for reasons beyond their control what if , say, 40% of copies are pirated. Would that number, less than half of the “90%” number but still nearly half of all copies, not still be outrageous?

    And I don’t mean in comparison to anything, I just mean the fact that as many as half of the copies could be pirated.

    Incidentally, I only have World of Goo installed on one machine but I don’t know anybody else who has it so I don’t know whether I’m an anomaly. :p

  17. suibhne says:

    That’s a pretty sensationalistic claim. At least Stateside, almost everyone has dynamic IP addresses via their ISPs, so I’m a little surprised to see 2D BOY list that as some sort of minor fringe case.

    Obvious caveat: Not saying piracy doesn’t suck, especially with such a wonderful, deserving indie title as this.

  18. theleif says:

    My router has a dynamic ip, as, i think, all ADSL users have (at least here in Sweden). That number can change after only a week sometimes. Now, i havn’t bought (or played) this game yet, but if i had, i would have contributed 3-4 ip adresses so far to that stastistic.

  19. The Sombrero Kid says:


    you make it sound like it’s justified for adults to ‘steal all the good jobs’ and only let children play tokenistic roles in shaping society and pay them BLOODY PITTANCE for it! no wonder so many of these poor out of work children are hanging around street corners, wearing hoodies, drinking in parks and generally being misunderstood by their elders!

    BAN AGISM NOW (or wait and maybe the problem will go away on it’s own :S)

  20. Asbestos says:

    Don’t forget that you can also upload your high scores from the demo. So lets hope they only counted high score uploads from chapter two and onwards (unless they can discern between retail and demo uploads). I haven’t bought or pirated the game, but I sure did upload my scores!

  21. Heliocentric says:

    Remember, these guys are indie devs so you may not challenge their maths!

    Hell…. 10%,1%.. Whatever! how many people did buy it? Are these people wearing money hats yet? They should be they made a lovely game :)

  22. The Sombrero Kid says:

    you can get static ips for more monies of your isp & it changes every time you turn your computer on if you don’t have a router, which is silly but does happen sometimes

  23. Toziel says:

    Surely Piracy can be a thorn in the flesh of the developer but the question that also should be asked:
    Would everyone that has pirated a copy of a game buy the game the if there was no piracy? Surely there are some that would buy the game but personally i think: I have no problem buying a game that is worth the money but I hate those companies that bring out a game in early beta or late alpha status (fallout 3 PC version….) and then they complain the game does not sell well.

  24. Quirk says:

    Seriously, guys. Can’t you put some caveat in the article to the effect that these figures are blatant pseudo-science? Taking their outrageous claim at face value is, to draw an odious comparison, Daily Mail-worthy. I’m seeing comment after comment from people who presumably don’t have any technical background at all and are willing to trust your authority on this. Use it for good! That way, when you decide to use it for evil later, they will trust you!

  25. cyrenic says:

    Another thought (find a fix for the edit button!):

    At first I thought their 90% figure was wildly inaccurate because of dynamic IPs, but after thinking about it I’m not so sure. It’d be quite a coincidence for them to come so close to the number Russell Carroll came up with if it was just a random, inaccurate number.

    Also, if dynamic IPs were skewing the numbers so much, I’d expect the number they came up with to be closer to 100% than it is.

    Just a thought.

  26. green says:

    The question is if using no DRM created more buzz, and thus more sales, than going with some easily cracked copy protection…

  27. The Sombrero Kid says:

    it’s because dynmic ips skew figures over time and they use a relatively small time slice thus limiting thier impact, in a year piracy from these figures will seem like a million billion to 1 legit copy

  28. The Sombrero Kid says:

    a.k.a. stardocking

  29. Bowl of Snakes says:

    I bought the game and have played from at least 4 IPs, according to their data, I am 75% Piracy.

  30. Jerricho says:

    It’s because people like Hobbits.

  31. nikos says:

    ISPs who use dynamic IPs usually do so in a persistent manner. Also, you missed the point of the original article if you think it’s about statistics and “pseudo-science”.

  32. Downloads_Plz says:

    I’ve installed my legally purchased game on 2 computers, and being extremely conservative, have had at least 5 different IP addresses while playing on each of them.

    So I’m at least 90% Pirate, probably higher.

    Seriously, that number is complete rubbish. I’m not saying the game isn’t pirated at all, but I would bet every penny I own that the real number is nowhere near 90%, and anyone taking that number and running with it should feel a bit foolish.

  33. screeg says:

    You laud them for having a rational response and not implementing some sort of DRM… What, they’re supposed to get outraged at game piracy and punish the people who give them money? DRM is worse than useless. It doesn’t slow down pirates and pisses off people who actually pay for games. It’s irrelevant whether the game was 90% or 99% or 100% pirated, putting a DRM scheme in place isn’t going to help, is it?

  34. Colthor says:

    Dunno how often other people restart their routers, but ours tends to get turned off overnight. A new IP every day means someone like me could quickly rack up a lot of ‘pirate’ copies…

  35. The Sombrero Kid says:

    wow’s drm is near impregnable, so it depends on the drm and how much people want to pirate it

  36. Myros says:

    The anti-DRM folks just got another clip of ammo for their cause … football manager 2009. This has to be the worst day 1 DRM cluster-f*** I have seen so far. Servers not working, phone numbers not working or hanging up before you can finish, people who managed to activate their game late last night try to play today and get an ‘expired’ error and can no longer play, the serial number font is so bad that it is impossible to tell 1s from L or 0 from o or 2 from Z etc, leading to having to try a massive amount of combinations to attempt to guess the code. Bet about right now they wish they had not used any DRM :) The old thing about pirates being able to play the game easier than legal owners seems to be proven true in this case.


  37. Quirk says:

    You’re going to have to explain what you’re talking about by “in a persistent manner”. Do you mean that the ISPs giving out dynamic IPs usually give out dynamic IPs, and thus the dynamic IPs can be factored out? If they had factored the dynamic IPs out though, they would not be giving it as a possible confusing factor. If, on the other hand, you mean that dynamic IPs are not changing, I think you may need to look up what a dynamic IP is.

    And the original article goes into shock and awe at a piracy figure of 90%. It’s echoed in the “so many people” who’re not paying, in the rethink on the necessity of DRM – all of which ends up built on a false foundation if, in fact, piracy represents a very much smaller percentage of users than is being suggested here.

  38. Dusk says:

    Four of my friends have pirated this on PCs which have dynamic IP, using i suspect 100+ independent IPs since release.

    Dynamic IPs will skew the results of BOTH PIRATED AND LEGIT versions. So the effects may cancel.

    $20 for a non-re sizable flash game is way too much. Hell crysis warhead was only $48 here, and has better textures. I would pay $5, except I never buy thing off the internet.

  39. Stuk says:

    It appears some people here is denying the 90% because of dynamic IPs. It also appears that these are the people who didn’t click through and read the actual blog post from 2DBoy.

    They explain why they believe their estimate to be true. The main counterpoint to dynamic IPs is the little checkbox which enables score submission. Which is unchecked by default. And if I were I pirate I would leave unchecked.

    So for every 100 IPs someone here has had, I bet there are people (both legit and pirated) who didn’t submit any scores, balancing out the number.

    (also most ADSL packages in the UK come with a router, meaning you essentially have a static IP until the router goes off)

  40. Quirk says:

    You’re not talking about sticky IPs, by any chance? Dynamic IPs connected to cable modems etc which only change when the modem is switched off or the DHCP server is reset? Firstly, these still change, just a little less frequently, and secondly many ISPs forcibly reassign IPs anyway to prevent users of their cheapest option from running websites off a local computer.

  41. Quirk says:

    Sorry, but that’s not even remotely a “balancing factor”. That’s another factor entirely, and is essentially unquantifiable.

    You can click that checkbox, and submit a dozen towers under a dozen IPs, and be counted a dozen times. Your “balance” lies on the assumption that eleven pirates have downloaded the software and not clicked the box to have their score submitted. What evidence have you for that assumption? None.

    Note well that they are not going off the IP at the time the checkbox was clicked, but at the time the score was submitted.

    They’re two separate factors, and dynamic IPs are likely to play a much, much more massive role.

  42. cyrenic says:


    I too use texture quality as my baseline for determining the monetary value of a game.

  43. Ian says:

    Isn’t the real point of the article actually quite a positive one? In comparison with a game shipped with DRM, the piracy rate remains about the same — rather than being higher — therefore there seems to be no business justification in implementing a ‘feature’ which does little to protect revenue and much to irritate certain users.

    You can argue all you like about the methods of reaching these numbers, but surely the conclusion that 2DBoy are drawing is that developers are better off finding other mechanisms of protecting their revenue than by using the hated DRM? A good thing, angry internet men, no?

  44. The Sombrero Kid says:

    that’s stupid that doesn’t cancel 1 legit copy and 5 pirated copies each with 100 ip’s makes it look like 1 legit copy and 599 illegal copies thats 599:1 instead of 5:1 this is not the traditional definition of ‘canceling’ this is dream money!

  45. Blah says:

    Dunno if anyone’s mentioned this yet, but people who pirate stuff aren’t prone to letting the software phone home anyway (e.g. using the high score table).

  46. D says:

    “for simplicity’s sake, we just assumed those would balance out. so take take the 90% as a rough estimate.”


    Making an assumption that:
    *Legit multiple dynamic IP’s* will balance out *Pirated versions never connected to server*

    Is already making the implicit assumption that for every legit purchase, there will exist 10* pirated versions.

    *(insert number of average unique dynamic IP’s assigned to each legit purchase over the sample period.)

  47. Gap Gen says:


    It’s related to penis size.

  48. YogSo says:

    @ Dusk: Four of my friends have pirated this on PCs which have dynamic IP, using i suspect 100+ independent IPs since release.

    Dynamic IPs will skew the results of BOTH PIRATED AND LEGIT versions. So the effects may cancel.

    I’m afraid it’s quite the opposite: Instead of only adding +4 to the total amount of pirated copies (that would be the real amount), your four friends would appear as if there are 4×100 = 400 more pirates playing the game for free… And that is absolutely ludicrous.

  49. DSX says:

    I was very disappointed to read the 90% figure in the previous posts, but now I’m even more disappointed in how they’re calculating this figure.

    “The prevalence of dynamic IPs makes any correlation between unique IPs and users all but impossible to make”

    Exactly. As a former network admin, I believe this to be true, it slants the percentage significantly: more then any other statistical variance. For anyone to throw this number around and make a serious claim is *VERY* irresponsible. It undermines tremendously a very real, controversial issue.

    It’s exactly like if I stood by housing complex with a single entry road and counted cars going by all day as a means of judging the population, then published it as argument for community zoning laws.

    2D Boy: Congrats on World of Goo, Shame on you for hyping a volatile issue with nonsense statistics.

  50. Jerricho says:

    Quite the point, Ian. It potentially exsposes the truth of a number of rationalisations.

    It’s been stated elsewehre that publishers are terrified of not using DRM for fear of the financial loss. This experiment adds to the weight of FAIL of that arguement.

    Equally, Cliffski and others have raised this point, many who use DRM as a justification for piracy have also been invalidated. I say many and not all.

    I think 2DBoy are ultimately doing the right thing here even if we disagree with their subsequent measuring techniques, by showing that a DRM free casual title suffers, at most, an equal degree of piracy as one with DRM: thus completely negating any support anyone could have for said DRM.