Spielend programmieren lernen

Programmierkurse für Kinder, Jugendliche und Erwachsene

User Tools

Site Tools


en:blog:2010:1121_how-to-fail-building-a-mmog

How to fail building a MMOG

  • date 2010-11-21 10:55
  • author Horst JENS
  • tags report, event
  • googleadsense True
  • slug 20101121_how-to-fail-building-a-mmog

On 19. Nov 2010 i heard a very interesting lecture:

How to fail building a Massively multiplayer online game

please note: i write all from memory, incorrect statements and false quotes are to be expected. If videos or papers of the event become availably, you will find them in thelinks* section at the end of this article.*

The lecture was part of a series of computer-game related lectures organized by Jogi J. Neufeld of the Subotron retro-game shop in Vienna, Austria. Uncommon for such events, the lecture was sponsored by Zit and there was free wine and food.

Three veterans of a 2003 pc game startup company met and discussed the reasons of failing to build a MMOG. [caption id=“attachment_286” align=“alignright” width=“640” caption=“Dipl.-Ing. Peter Reiterer, Markus Hajek, L. Aaron Kaplan, Roland Gratzer”]image0[/caption]

The discussion, moderated by artist Roland Gratzer from monochrom, was held by Dipl. Ing. Peter Reiterer (coder, physic), Markus Hajek (coder, tools), and Aaron Kaplan (coder, network). Except of Roland Grazer all three participants were for some time hired by the (no-longer-existing) startup Wootsoft and talked about their experience there.

history

According to the participants of the dicussion, around 2002 the 2 founders of Wootsoft had a self-made, state-of-the-art graphic engine and decided to build a Science-Fiction Massive Multiplayer Online Game. The business plan was to sell subscriptions (around 10 $ per month) and allow the player to do “everything possible” inside the online world. The graphic engine was designed to work on a PC. After finding investors the team rented a office in Vienna and started hiring people. Lacking a clear vision, necessary financing and a product to sell the company ultimately collapsed.

Tech-Demo video of the graphic engine:

[googlevideo=http://video.google.com/videoplay?docid=2960015888661616574&hl=de#]

moments

During the discussion, the 3 veterans searched “the reason” why the project failed or at least “the moment” were doom was unavoidable:

  • Markus: “The project was doomed right from the start. The team was young and unexperienced. We should have build some smaller projects first. For the scope of the project (building a World of Warcraft -challenger in a SciFi universe) the financing was too low by a factor of 20.”
  • Aaron: “Each programmer was busy building his own Ivory Tower without much communication to the other members of the team. I worked very long hours there, sometimes sleeping in the office, optimizing the communication between servers into the nano-second range. Finally i leaved because of bad health. I decided to quit after i was told to not work on server communication (for what i was hired) any more but instead help with the graphics.”
  • Peter: “I was very happy being paid for coding there and did not paid much attention to other tings. I made and optimized a math library that i felt would be useful (partly in assembler). In the end, maybe 10% of the library was necessary, and i could have done the job in one week instead of 6 weeks. I realized that it was over when one of the 2 founders told me the company will go in insolvency.”

tech

In retrospect, there was no lack of innovative technology or concepts. The online world was designed to be “seamless” with no pausing for loading new areas. The graphic engine had very cool effects for it's time. There was no major SciFi online game at the time. Papers were produced about some of the more innovative aspects of the developement (distributed, shared, scalable memory) and high-skilled people were hired.

signs of failure

A document was presented and discussed, listing some signs for startup company doomed to fail. Some of the points:

  • typical dotcom life cycle: start small. expand too fast. die.
  • No experience in the team. No smaller “beginner” projects successfully done.
  • No marketing research “what is my core customer group” influencing the product. The game was designed for P.v.P. hardcore gamers, a ridicule small percentage of worldwide customers.
  • Not coordination between the team. No shared vision. teams working without much contact or control
  • No process to communicate between the sub-teams and make decisions was etablished
  • The guy in love with his big vision of the game was the same guy who should axe costly features (conflict of interest)
  • visionary goal instead of concentration on creating and selling a marketable product
  • no cooperation with financial strong publishers for fear of loosing control

conclusion

Markus: “The project was doomed to fail. But it should have been possible to fail in a far less painful way.”

afterlife

From the 3 participants, 2 remain still in the (very small) game industry of Vienna, but in changed roles. All concluded that they learned a lot during the work for wootsoft and feel that future projects profit from the experience gained there. Even if some of the follow-up projects also failed, they failed at a higher level, impossible to reach without failing at a lower level before.

/var/www/horst/spielend-programmieren.at/data/pages/en/blog/2010/1121_how-to-fail-building-a-mmog.txt · Last modified: 2017/10/12 09:43 (external edit)