# Dynamical Simulation for Reputation Score

###### tags: `Retrievability Pinning`

*Authors: Irene Giacomelli (Protocol Labs), Danilo Lessa Bernardineli (BlockScience)*

:::info
Up to date by August 2022
:::

Base Reference: [DRC Reputation Score Brainstorming](/8CSLDfKNSAyyl-OfEzOdDQ)

- Simulation parameters
- Time
- Range: 12 weeks (84 days)
- Granularity: daily
- Metrics
- a
- Arrival process for new providers
- $P(N_P+ \Delta N_P) = Binomial(N_{p, max} - N_P, p_P)$
- $N_{p, max} = a_1 (1 - e^{-a_2 t}) + bt + c$
- Guesses:
- $a_1 = Unif(20, 20k)$
- Unita
- $a_2 = Unif()$
- $p_p = \beta(1, 1)$
- Arrival process for new clients
- $P(N_C+ \Delta N_C) = Binomial(N_{C, max} - N_C, p_C)$
- $N_{c, max} = a_1 (1 - e^{-a_2 t}) + bt + c$
- $a_2 = \frac{ln(2)}{HalfLife}$
- Guesses
- $a_1 = Unif[10, 10k]$
- Unit: # of agents
- $a_2 = Unif[3, 30]$
- Unit: day
- $b = \mathcal{N}()$
- Unit: # of agents per day
- $c = \mathcal{N}()$
- Unit: # of agents
- $p_C = \beta(1, 1)$
- Arrival process for new deals
- $\Delta N_D = N_C * \mathcal{N}(\mu_D, \sigma_D)$
- Guesses
- $\mu_D = 0.1$ Deals-Day Per Client
- $\sigma_D = 0.1$ Deals-Day Per Client
- $\tau_D = \mathcal{N}(\mu_{D, \tau}, \sigma_{D, \tau})$
- Guesses
- $\mu_{D,\tau}=30 d$
- $\sigma_{D,\tau}=20 d$