Wednesday, November 30, 2011

The Base Case

"Inequality of outcomes is okay, I just think we need to work on equal opportunities."

If you haven't heard that sentence, you've never been to a SWPL dinner party.  Good for you. =)

Really though, there's a theory that where you end up is a function of where you start.  I amend that theory by saying that where you end up is a function: outcome = f(start, effort, luck).

Anyhoo, follow these threads for a while and you get to a oft-asked question: If we were to set everyone to zero, what would outcomes look like?

Well, here's World Of Warcraft to help.  (tl:dr: WoW has a pretty high degree of wealth inequality)

WoW has some nice properties.  Everyone starts out at the same point.  Luck is a factor but over the timelines that folks play, not a major one.  So we basically "control" for start and luck and take a look at "effort."  Maybe "effort" isn't a good word, but the thing that remains depends strongly on the actual actions of players.  What sort of actions separate folks?  Some ideas:

  • Number of hours played
  • Cooperation with others (finding and maintaining membership in a good guild)
  • Desire to earn gold in the first place, compared to other goals such as PvP success, world exploration, raiding, etc.
  • Initial character build choices (n.b., a limited amount of rebuilding can occur even for "older" characters, and many players play with multiple characters).
It should be obvious how well these factors map to real world considerations.  If you asked me to guess someone's career earnings, I could do worse than to ask these four questions:
  • How hard does he work?
  • How well does he work in teams?
  • Is he motivated by money?
  • What is his education and work history from college through his current job?
Of course this is the real world, so "who are his parents?" and "does he have an uncommon surplus or deficit of luck" enter into things as well.

Maybe that explains why the US has an even higher degree of wealth inequality than WoW does...?

No comments: