How AI works in LeaderLodge
Computed by LeaderLodge. Explained by AI.
The idea
LeaderLodge measures performance from your game history. Cloud language models turn those measurements into readable summaries. The model does not calculate ranks or invent scores. It interprets the data that LeaderLodge records and computes.
Two ways to ask
- Contextual insights
- Player detail, leaderboards, Hall of Fame, matchups, and game detail pages include an Insights panel beside the data you are viewing. Choose a preset (Strengths, Overview, Dynamics, and others) or ask a short custom question about that snapshot.
- Ask the Lodge
- A dedicated page for natural-language questions about your game data and players. The model requests validated queries; Python runs them and iterates to find meaningful answers.
What we send to AI providers
- Computed stats: rankings, placements, trends, and group summaries
- Player display names from your group roster
- Your question or preset choice
We do not send your password or account details. We do not let the model run arbitrary database queries or raw SQL.
Models and control
Insights use server-side API keys only. Signed-in users can pick among a small set of low-cost models. Answers are cached when appropriate, and rate limits can be enabled site-wide.
Every insight panel includes a disclaimer: AI summary. Rankings are computed by LeaderLodge.