6b6f78885f
Some checks are pending
Sync Graph Data to teleo-app / sync (push) Waiting to run
Pentagon-Agent: Ganymede <F99EBFA6-547B-4096-BEEA-1D59C3E4028A>
38 lines
No EOL
2.1 KiB
Markdown
38 lines
No EOL
2.1 KiB
Markdown
---
|
|
type: claim
|
|
domain: internet-finance
|
|
description: "Using max or average rate instead of time-varying rate leads to chronic under or overstaffing"
|
|
confidence: proven
|
|
source: "Liu et al. (NC State), 'Modeling and Simulation of Nonstationary Non-Poisson Arrival Processes' (2019)"
|
|
created: 2026-03-11
|
|
---
|
|
|
|
# Constant rate approximation of time-varying arrivals causes systematic staffing errors
|
|
|
|
Replacing a time-varying arrival rate λ(t) with a constant approximation—whether the maximum rate, average rate, or any other single value—leads to systematic capacity planning failures. Systems sized for maximum rate are chronically overstaffed during low-demand periods, wasting resources. Systems sized for average rate are chronically understaffed during high-demand periods, creating congestion.
|
|
|
|
This is not a minor efficiency loss but a structural mismatch: the constant-rate approximation discards the temporal structure of demand, making it impossible to match capacity to load.
|
|
|
|
## Evidence
|
|
|
|
Liu et al. explicitly state that "replacing a time-varying arrival rate with a constant (max or average) leads to systems being badly understaffed or overstaffed." This is a direct consequence of nonstationary arrival processes where demand varies predictably over time.
|
|
|
|
The paper demonstrates that "congestion measures are increasing functions of arrival process variability," meaning that even if average load is manageable, temporal concentration of arrivals creates congestion that constant-rate models cannot predict.
|
|
|
|
## Implications for Pipeline Architecture
|
|
|
|
For capital formation pipelines with session-based arrival patterns, this means:
|
|
|
|
1. Sizing capacity for peak (research session active) rate wastes resources during quiet periods
|
|
2. Sizing capacity for average rate creates backlogs during research sessions
|
|
3. Optimal capacity must be time-varying or must use queueing/buffering to smooth demand
|
|
|
|
The MMPP framework provides tools to size capacity for the mixture of states rather than for a single average state, enabling more efficient resource allocation.
|
|
|
|
---
|
|
|
|
Relevant Notes:
|
|
- domains/internet-finance/_map
|
|
|
|
Topics:
|
|
- core/mechanisms/_map |