Tuesday, June 2, 2020

You can't run a project with dice


Is this a fair die with a roll-string of heads and tails like this?

HHHHHHHHHHHTHHHHHHHHHHHHTHHHHHHHH

It doesn't look fair, but it well could be.
50/50 heads and tails is a so-called limit outcome requiring, in theory, nearly infinite rolls to achieve

What's going on here?
  • The die has no memory!
  • The next outcome (roll) is not dependent in any way on the last roll
  • There are no lessons learned!
  • There are no political, business, or project pressures which could affect outcomes
  • There are no biases; there is only objectivity
  • The rules of chance are fixed, and can't be changed, by anyone
In a real project:
  • You've got memory! You'd better remember what happened last
  • Whatever comes next is dependent on what just happened; no vacuums allowed
  • Lessons are learned, whether you do a formal inquiry or not. Only the truly ignorant ignore the past in all respects
  • There are pressures!
  • Of course there are biases; everyone has an attitude about risk that's not entirely objective
  • And, your job as PM is to change the rules, circumstances, culture, or whatever it takes for project success! (*)
Conclusion:
  • Dice games don't work in projects 
------------
(*) For roughly the same reason, Earned Value doesn't work well in projects either. The linear equations upon which the methodology depends are artifacts of fixed rules. But in projects, the rules are always subject to circumstances, and so, like the rules of dice, the linear equations are obsolete almost as soon as they are written.
HOWEVER, EV methods are good for setting up a plan; just remember: all plans change the moment reality is contacted.



Buy them at any online book retailer!