ATS Dashboard — Against The Spread by Team
Loading model-cover% per team across 7 sports…
📋 How this works
For every historical game, we back-compute the model's implied "spread" from ELO diff (ELO_diff / 25 = pts). Compared to actual margin, we tag each game as a team COVER, FAIL, or PUSH.
Teams that cover ≥65% of games signal MODEL UNDERRATES. Teams covering ≤35% signal MODEL OVERRATES. Use this to lean toward / fade specific teams.