What is Semi-Random Selection?

The semi-random selection process for assigning cubes to projects that have requested funding is described below.  In this description, a pre-cube is a filled project that contains three tokens and has requested cube status (i.e., has requested funding).

 

The goal of this process is to ensure that all participating units (schools, colleges) in MCubed have representation in the cubes before shifting to a completely random lottery system.  (Note:  This goal can only be achieved when faculty from all participating units in MCubed invest their tokens in projects that request a cube, and when those units have sufficient numbers of tokens that can be funded.)  In accordance with the “no traditional peer review” principle of MCubed, the semi-random selection algorithm does not evaluate the quality of projects in any way.

 

  1. At the close of the cube-request period, the list of pre-cubes is sorted into a random list.
  2. The first project on the list is cubed.
  3. If the next project (a) contains faculty from a unit that is not represented in the cubed projects thus far and (b) the units of all the faculty in the cube still have tokens remaining, the project is cubed; otherwise, the project is skipped.
  4. Subtract one token from the units of each faculty member in the cube.
  5. Repeat step 3-4 until all the pre-cubes have been examined.
  6. Start at the top of the list again and cube the projects in order if (a) the project has not already been cubed and (b) the units of all the faculty members in the cube still have tokens remaining.
  7. Subtract one token from the units of each faculty member in the cube.
  8. Continue down the list, cubing projects that have not been cubed until the total number of cubes formed in the particular cubing even has been reached (e.g., 50 for the first cubing event).
  9. Keep any projects that aren’t pulled for cubing in the pre-cube state in anticipation of next cube distribution.