Modeling xPass as a two-stage problem: first predicting the intended target XY for the ~5% of passes where the recipient is unknown or blocked, then computing completion probability against that imputed target. This eliminates the short-distance completion paradox where elite players appear unable to complete 2-yard passes (because blocked/intercepted short passes get logged as failed completions with very short distances).
Stage 1: For passes without a known recipient, predict intended target location from passer position, body orientation (if available), and 360 context. Stage 2: Compute completion probability against the imputed target. This two-step approach prevents the model from seeing impossibly short "failed" passes and inflating difficulty estimates for trivial passes.
When a pass is blocked or intercepted, it registers in the data as a failed completion with a very short distance (the distance the ball actually traveled before interception, not the intended distance). Standard xPass models see these as "failed 2-yard passes" — which should be 99% completion — and conclude the player can't complete trivial passes. The real story: the player attempted a 20-yard progressive pass that was intercepted after 2 yards. Without imputing the intended target, xPass models systematically penalize the best progressive passers.