Pitch Sequence Effectiveness
Search for a pitcher to view sequence analysis
Try searching "Skubal" or "Yankees" to find pitchers
3D Pitch Trajectories
Sequencer
Pick a pitcher and an eligible hitter (100+ PA since 2025), then place two pitches on the zone. We score the sequence against historical 2025–2026 outcomes using a Bayesian blend of the pitcher's stuff and the hitter's splits.
Select a pitcher above to begin.
Plate Appearance Analysis
Select a completed game and plate appearance to compare real pitch outcomes against the Sequencer's probabilistic model.
Pitches the pitcher is throwing well below their own typical rate in this count. Refreshes every minute.
LiveSeqer predicts the next pitch type and location between every pitch using the pitcher's empirical tendencies (count, batter hand, prior pitch) blended with the hitter's pitch-type weakness. Predictions refresh every 10 seconds while a game is live.
Game Plan shows each starting pitcher's likely approach against the opposing lineup, batter-by-batter and grouped by times-through-the-order. Lineups poll every 20 minutes starting 3 hours before first pitch, and the SP status checks every 60 seconds once a game is live.
Pick a live game to see inning-by-inning bullpen projections for both teams, ranked by Matchup, Role Fit, and Fatigue. Auto-refreshes each time the batter changes.
Pick a game to score both bullpens 1–100. Each reliever's composite is split into a transparent Fatigue sub-score (recent workload from MLB StatsAPI) and a Matchup sub-score (arsenal vs. the opposing active hitters). Click a row to expand the sub-score breakdown.
Select two pitchers to compare their sequencing fingerprints
Try "Skubal" vs. "Cole" or "Wheeler" vs. "Burnes"
Search for a pitcher above to browse plate appearances.
Simulate the terminal outcome of a plate appearance against the selected pitcher using an absorbing Markov chain over all 12 ball-strike count states. Choose a backend: Empirical uses exact-count Statcast outcome distributions (pitcher-only, fast); LiveSeqer marginalizes the full prediction engine's pitch-type mix over the scorer (slower, hitter-conditioned).
Roster construction, player value, and opponent scouting — powered by Sequence's Bayesian arsenal engine.