Consensus spine
Ingest 16 ranking sources, normalize names, schools, and positions, and build a weighted consensus baseline per prospect and per position group.
Season-scoped to 2026 with source and position canonicalization baked into the pipeline.
A deterministic engine that turns 16 sources, measurables, and film notes into one draft board you can defend after the picks are made.
APEX OS 2026 · Season-scoped, migration-safe, and versioned for auditability.
APEX OS is not another mock draft. It is the spine that keeps your evaluation process consistent, deterministic, and auditable through the full draft cycle.
Ingest 16 ranking sources, normalize names, schools, and positions, and build a weighted consensus baseline per prospect and per position group.
Season-scoped to 2026 with source and position canonicalization baked into the pipeline.
Score every prospect against archetype libraries, measurables context, and failure modes to produce one composite, tier label, and capital band.
No stochastic sampling in scoring; every run is versioned and reproducible.
Render decision cards with tier, capital, divergence, FM tags, historical comps, and a draft day call ready to drop into your deck.
The same card layout powers the in-app right pane and exported PDFs.
Each layer has exactly one job, from raw CSVs to a draft-ready snapshot. The pipeline is deterministic, versioned, and season-scoped.
APEX ingests rankings and measurables from 16 sources, normalizes names, schools, and positions, and filters to an 861-player universe. Legacy rows are retained for audit, but never leak into active queries.
Build a season-aware consensus table, weighting sources by tier and excluding model-only feeds so algorithmic scores never contaminate scout rankings.
Run each prospect through position-specific archetype libraries, measurables gates, and failure modes to generate a single APEX score, tier, and capital range. Same inputs, same output, every run — all tied to a model version.
Compare APEX vs consensus to create structured divergence flags, compression tags, development bets, and scheme-dependent markers, each with a text explanation.
Lock pre-draft snapshots, export boards and PDFs, and ship a packet that makes your decisions explainable after the draft, not just before it.
Every disagreement with the market is gated, labeled, and logged — or it doesn't ship. Divergence is a first-class object, not a side effect.
APEX OS ships multiple surfaces — board, detail panel, and reports — all driven by the same deterministic engine and tags.
Big Board and APEX Board tabs with unified filters, tag legends, tier coloring, and click-through to a right-hand detail panel.
Built for quick scan and triage — tag filters, column guide, and APEX vs consensus columns clearly separated.
Six-tab right pane with summary, traits, risk, comps, notes, and report views, plus a decision card pinned at the top.
Custom HTML/JS, not Streamlit tabs, so layout and height estimation are predictable.
Prizm-style PDF generator with trait bars, historical comps, divergence narrative, and FM legend that match the app exactly.
One-click exports designed to drop straight into draft decks and post-draft audits.
APEX OS is designed to survive regime changes. Every source, score, tag, and override is traceable.
If you just want another mock draft, APEX OS is overkill. If you want an operating system for how you evaluate, it's exactly the right size.