Methodology
A layered probabilistic framework — from information aggregation to calibrated simulation — designed for explainability, not black-box rankings.
System Layers
1
Data Layer
8 weighted mock sources · prospect database · betting markets
CBS FinalBleacher ReportTankathonNBA Draft NetHoopshqNetscoutsSB NationUSA Today FTW
2
Consensus Layer
Source-weighted robust trimmed-mean aggregation
3
Market Layer
Betting odds integration at picks #1–#2
4
Archetype Layer
Creation · shooting · size · defense · upside · NBA-ready
5
Team Fit Layer
Prospect archetype × franchise need profile
6
Monte Carlo Layer
20,000 calibrated stochastic simulations
7
Probability Outputs
Board · volatility · confidence · scenarios
Weight Matrix
Consensus60%
Betting18%
Reliability12%
Team Fit10%
- Simulations
- 20,000
- #1 Overall
- AJ · 99.9%
Source Reliability
| Source | Weight | Coverage |
|---|---|---|
| tankathon | 1.5 | 46 players |
| bleacher | 1.3 | 22 players |
| nbadraftnet | 1.2 | 14 players |
| cbs final | 1.0 | 30 players |
| hoopshq | 1.0 | 14 players |
| netscouts | 1.0 | 14 players |
| sb nation | 0.9 | 14 players |
| usa today ftw | 0.9 | 14 players |
Information Conflict · AJ Dybantsa
When sources disagree, volatility rises — even for consensus #1 picks.
bleacher#1
cbs final#1
hoopshq#1
netscouts#1
sb nation#1
tankathon#1
usa today ftw#1
espn#10
athletic#20
nbadraftnet#20