Decision models & simulation

Fantasy draft strategy lab

Make a pick with the next round in mind.

An interactive draft board that connects player projections, positional scarcity, roster fit, and a snake-draft schedule to an explainable decision.

Open the draft boardExplore source

Uses the repository’s September 1, 2026 projection snapshot and stated draft assumptions. It does not connect to an ESPN league or execute picks in ESPN.

Historical draft-planning toolDraft decisions · Concept illustration
projection snapshot
2026
pick-order scenarios
Snake draft
roster & scarcity factors
Explainable
Contribution
Personal model & decision interface

01 / The problem

Make a pick with the next round in mind.

The highest projected scorer is not always the most useful next pick. A draft decision depends on roster needs, positional replacement value, the players likely to remain, and when the next turn arrives.

I turned a quantitative draft strategy into an interactive board. Visitors can mark players as drafted, track their roster, and inspect why the model values a candidate under the current draft scenario.

02 / In practice

Follow the workflow.

Open the draft board ↗

  1. Choose a draft position and inspect the available board.

  2. Mark picks, track roster needs, and compare candidates.

  3. Inspect recommendations, revise the scenario, and export the board.

Fantasy draft strategy lab interface with example inputs and results
Actual interface · Public research snapshot. Open the demo to change the inputs and inspect the results.
A working interface, ready to explore.

Uses the repository’s September 1, 2026 projection snapshot and stated draft assumptions. It does not connect to an ESPN league or execute picks in ESPN.

Launch demo ↗

03 / Under the hood

How the pieces fit.

  1. 01

    Snapshot

    Read the public projection, tier, value-over-replacement, and market-ADP data.

  2. 02

    Track

    Maintain picked and available players alongside the snake-draft schedule.

  3. 03

    Evaluate

    Apply roster-fit, tier-drop, and candidate-scoring rules to the current state.

  4. 04

    Explain

    Expose recommendation ingredients and assumptions instead of a single unexplained ranking.

04 / Engineering judgment

Decisions & tradeoffs.

Separate projections from outcomes

The source data is a dated expectation of player performance, not a record of realized results. The interface keeps that context visible.

Make opportunity cost tangible

A player’s value changes as other players at the same position disappear. Tier and roster context make that change inspectable.

State uncertainty honestly

Any reconstructed uncertainty proxy is labeled as an assumption. The tool does not claim the precision of a calibrated draft-outcome model.

05 / Inspect the work

Evidence, with context.

Strategy model & public board

Original Python strategy, projection snapshot, and interactive draft implementation.

Keep exploring

Quantitative researchEquity factor risk modelMathematics & signal processingFourier signal lab