Original text at http://members.aol.com/acockburn/riskcata/progtrai.htm. Originally called ProgressAndTrainingTeams, but recently informally renamed to DayCare. I haven't decided yet whether DayCare is suitable for polite society. --AlistairCockburn

----

JoeDavison pointed out I oversimplified the model in the paper. He added:

"1. Static analysis, which is what you've already been doing. Here, assume the productivity of the novices is n without the mentor, and m*n with the mentor, with  m>1 maybe as large as 2 or 3.  
 
2. Time series analysis. I did some stuff like this when I was playing  with discovery costs. I found it was important to recognize that the  efficiency of novices changes over the course of a development, in order to  explain some of the phenomena we actually see. (And because it's  obviously true.)"

I add to this the CornFieldEffect, which says the experts also get better working together.

Net result is an equation something like:
   X*c  +  N*m*n(t) 
for X active experts (not the mentor) working at c>1 productivity, and N novices improved by mentoring m>1, at productivity n(t) improving noticeably over time. The mentor's contribution is carried in m and n(t).  
  -- AlistairCockburn

----
The pattern name is somewhat derogatory - it shouldn't be, but it is. How about "Incubator"? Turns eggs into chickens, maybe it'll work to turn novices into experts.

''Agreed. I also don't like it for that exact reason and don't use it anymore, preferring it's original  name "Progress Teams and Training Teams". "Incubator" is also a good name. My apologies for publishing it under the DayCare name. --AlistairCockburn''