AFFL Auction Lab

AFFL Auction Lab is an interactive history of how a long-running ESPN fantasy football league spends a $200 auction budget. It turns ten seasons of winning bids into explorable roster-construction, efficiency, nomination, and position-return stories.

The project is intentionally separate from AFFL Wrapped. Wrapped tells one season's story; Auction Lab is the research workbench for comparing teams and strategies across seasons.

What you can explore

• Average dollars allocated to QB, RB, WR, and TE by season or franchise
• Position-investment trends across the 2016-2025 auction era
• Player-level $200 roster stacks, separated by price-implied starter and bench roles
• Spending concentration versus drafted-player fantasy points per dollar
• A reconciled price-tier heatmap for $1, $2-$5, $6-$15, $16-$40, and $41+ players
• Own-nomination trends by franchise and season
• QB, RB, WR, and TE spending versus exact starter fantasy points
• A searchable, sortable ledger of all 1,804 auction purchases with filtered CSV export

Open the dashboard

No build step or server is required. Open index.html in a browser. The repository root is also ready to publish as a static GitHub Pages site.

Run the public-data and interface checks with:

npm test

Metric foundations

Auction prices are the winning bids retained by ESPN. Starter investment is not based on ESPN's draft-room autofill order. Instead, Auction Lab selects the highest-cost legal starting lineup under each season's roster rules, treating price as the better signal of draft intent.

The return charts use AFFL's archived standard scoring settings:

• 1 point per 25 passing yards
• 4 points per passing touchdown
• 1 point per 10 rushing or receiving yards
• 6 points per rushing or receiving touchdown
• 0 points per reception

Starter points are credited only when an auctioned player started for the team that drafted him. Full definitions and known limitations are in the metrics guide.

Data boundaries

This public repository contains a de-identified dashboard bundle. ESPN session credentials, owner IDs, owner usernames, and the private extraction archive are excluded and guarded by automated tests.

ESPN's retained post-draft data does not include losing bids, second-highest bidders, or a bid-event history. Nominating team is therefore shown only as a strategic signal and is never presented as runner-up interest.

Project structure

.
|-- index.html              # Dashboard shell
|-- site/
|   |-- app.js              # Charts, filters, table, and interactions
|   |-- data.js             # De-identified analytical bundle
|   `-- styles.css          # Responsive visual system
|-- docs/metrics.md         # Definitions, scoring, and caveats
`-- tests/                  # Privacy and data-contract checks