Jump to content

Time To Consider A More Robust Public Test Server Strategy?


1 reply to this topic

#1 tuffy963

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • 208 posts
  • LocationSan Francisco

Posted 08 October 2014 - 06:55 PM

I think that the recent patch again highlights the need for PGI to consider a more robust public test server strategy. My observation about this patch is that many of the issues might have been avoided if this patch had been pushed to a semi-persistent public test server instead of straight to the production game. A few examples that jump out at me are:
  • ELO convergence data vs player "happiness" could have been analyzed to determine the effects of the game mode preference change. This is the kind of change that could have sat out on a public test server 2-4 weeks without causing ANY community backlash and PGI could have collected some real data on the topic instead of the 24-48 hours of data before the patch is rolled back
  • Related to the first point, there is also less drama and urgency from the community when these feature changes are first seen on a public test server. I think players prone to provide more objective feedback as they feel they are influencing decision making over time. Instead the change feels rushed and players feel like they have to convince PGI and the rest of community with ultimatums and threats to see their agendas protects. Public Test feedback can be collected without disrupting the game's business model.
  • These strange bugs that seem to crop up with every patch that appear to be a function of inadequate testing of the patch (replicated modules, status page not updating, etc). Does the production community really have to catch these bugs, while are minor, add anger and confusion to a community that is trying to provide feedback a the larger, more impactful changes in the patch?
In short, turn on a partially persistent public test server. Push key (not all) patches to this server two weeks in advance and let it run day and night for those two weeks. Review the feedback and analyze the data provided players willing/interested in testing the features/changes. PGI make adjustments based on what they are seeing on PTS. PGI can then choose to push another version of the patch to the PTS or deploy the patch to production.




Is this process more costly? YES
Does it slow down the development process? YES
Is the extra effort worth avoiding needless community drama, damage to player base, and embarressing redaction of poorly deployed patches? I THINK SO!

I believe it is time for PGI to put on their Big Boy Company Pants and institute this long standing, industry best practice.

Edited by tuffy963, 08 October 2014 - 07:14 PM.


#2 tuffy963

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • 208 posts
  • LocationSan Francisco

Posted 10 October 2014 - 05:37 AM

Bump for justice!
Bump for love!
Bump for a little attention!





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users