PS Simulations
Tests & Simulations - trust but verify
1. Stress factors [TCM + iCL]
- Total Community Members (TCM) spike & drop - The simulation includes periods with different numbers of community members, ranging from 1 to 100,000. 
 
- iCL random fluctuations - Fluctuating : The average individual Commitment Level varies across periods to simulate changes in overall community commitment. 
- Changing : Each member () has a different individual commitment level, that keeps changing and being updated each period. 
 
2. Fixed Parameters
- 10 periods 
- 3 individual members ( ) 
- a stable performance ( ) set to a constant value of 1.05 (105%). 
3. Parameters used
- TCM (Total Community Members): The number of members in the community for each period. 
- (Average Individual Commitment Level): The average commitment level of all members in the community for each period. 
- (Individual Commitment Level): The commitment level of each member, ranging from 1 to 10. 
- (Fractional Commitment Level): The fraction of the total commitment level attributed to each member. 
- TCP (Total Contribution Points): The total number of contribution points available in the community for each period. 
- ECp (Expected Contribution Points): The expected number of contribution points for each member based on their fiCL, calculated as . 
- GCp (Given Contribution Points): The actual number of contribution points contributed by each member, calculated as . 
- PS (Participation Score): The cumulative participation score for each member, calculated as , with . 
4. Approximations for convenience
- TCP: The total contribution points available in each period. Here we constrained using the formula "heavy tasks” worth 10 Contribution Points. 
- PS: The participation score for each member. Calculated as , where , to simulate cumulative growth based on consistent over-performance. 
- iCL: The fraction of the total commitment level attributed to each member. Here calculated as This way we’d set up one single average iCL ( ) for the entire community, instead of generating millions of individual iCLs for each member. 
5. Results


Last updated
