Comments on: The Agile Prisoner’s Dilemma http://blog.karmona.com/index.php/2007/07/16/the-agile-prisoners-dilemma/ Pragmatic Software Management, Internet Trends, Life and more... Fri, 30 Jul 2010 19:50:36 +0300 http://wordpress.org/?v=2.9.2 hourly 1 By: Pasha http://blog.karmona.com/index.php/2007/07/16/the-agile-prisoners-dilemma/comment-page-1/#comment-14 Pasha Thu, 02 Aug 2007 21:03:00 +0000 http://blog.karmona.com/index.php/2007/07/16/the-agile-prisoners-dilemma/#comment-14 Moti,<BR/>That's not the prisoner's dilemma, that's the "measurement dilemma" (as <A HREF="http://www.pashabitz.com/PermaLink,guid,b80e5877-b9cc-4679-af6d-cf149fd12754.aspx" REL="nofollow">posted here</A>.<BR/>As to iterative prisoner's dilemmas - the "eye for and eye" strategy seemed to always win. What does it mean to us?? That once you fuck up your sprint, we give you the worst tasks in the next one?

">

Moti,
That’s not the prisoner’s dilemma, that’s the “measurement dilemma” (as posted here.
As to iterative prisoner’s dilemmas – the “eye for and eye” strategy seemed to always win. What does it mean to us?? That once you fuck up your sprint, we give you the worst tasks in the next one?

]]> By: Moti Karmona http://blog.karmona.com/index.php/2007/07/16/the-agile-prisoners-dilemma/comment-page-1/#comment-15 Moti Karmona Tue, 17 Jul 2007 05:50:00 +0000 http://blog.karmona.com/index.php/2007/07/16/the-agile-prisoners-dilemma/#comment-15 Kalish, you are more than right, the basic prisoner dilemma is simplified and way too abstract for real life (e.g. no communication is allowed between the prisoners) but the concepts are still valid in the iterative model even when you are adding real life compositions; e.g. Assume your organization is rewarding for “bug fixing ratio” will you find yourself fixing as many bugs as you can - choosing the simple bugs and ignoring priorities if possible with very little verification or testing (assuming regression are not measured)? Kalish, you are more than right, the basic prisoner dilemma is simplified and way too abstract for real life (e.g. no communication is allowed between the prisoners) but the concepts are still valid in the iterative model even when you are adding real life compositions;
e.g. Assume your organization is rewarding for “bug fixing ratio” will you find yourself fixing as many bugs as you can – choosing the simple bugs and ignoring priorities if possible with very little verification or testing (assuming regression are not measured)?

]]> By: Uri Kalish http://blog.karmona.com/index.php/2007/07/16/the-agile-prisoners-dilemma/comment-page-1/#comment-16 Uri Kalish Mon, 16 Jul 2007 21:44:00 +0000 http://blog.karmona.com/index.php/2007/07/16/the-agile-prisoners-dilemma/#comment-16 If I’m not mistaken, choosing the Nash Equilibrium option in The Prisoner’s Dilemma is a preferred choice only if I’m playing a single one-turn game (and with a rational opponent). I’m almost certain nobody will be fired after raising his first red flag (and I’m quite sure some of my teammates are crazy), meaning we’re dealing with a multi-stage game here, where cooperation can be rewarded and betrayal can be retaliated, hence the shift in the Equilibrium point.

">

If I’m not mistaken, choosing the Nash Equilibrium option in The Prisoner’s Dilemma is a preferred choice only if I’m playing a single one-turn game (and with a rational opponent). I’m almost certain nobody will be fired after raising his first red flag (and I’m quite sure some of my teammates are crazy), meaning we’re dealing with a multi-stage game here, where cooperation can be rewarded and betrayal can be retaliated, hence the shift in the Equilibrium point.

]]>