01 / The count
Select any of the twelve legal counts
Play through baseball situations, choose your pitch and target, then see what happens. Work a single at-bat or call every pitch in a complete game—then study the strategy behind your decisions.
Select any of the twelve legal counts
Lineup spot is a starting assumption, not a scouting report
Walks, contact and execution risks shift by context
Compare available pitches and targets
The scores and outcome probabilities are from an illustrative coaching model, not measured MLB probabilities, medical advice, or individualized player projections. Higher is a relative utility score within this model—not chance of success.
Experiment with a different zone location
The previous pitch changes the next pitch's intent
Good tunneling means comparable early trajectories before later separation. Different eye levels can disrupt timing, but excessive repetition or poor command can undo the plan. The drawing is conceptual, not a ball-flight model.
A decision drill with context, feedback, and multiple defensible answers
Objectives change with ball-strike leverage, command and hitter tendencies
Two pitches may look similar during the early flight that informs the hitter's swing decision, yet reach different final locations or arrive with different timing. Shared release/early paths and late separation can create deception, but pitch-to-pitch execution matters.
Use up–down location variation to force recalibration. A high fastball may pair with a lower breaking or offspeed pitch. The highest target is not always a strike, and a chase pitch is not always justified.
A ball on 3–1 can concede a walk, whereas 0–2 provides more room for a competitive chase pitch. Two-strike fouls extend the at-bat; they are not strikeouts. Early counts still carry contact risk.
Lineup spot offers a scouting prior only. In one lineup the ninth batter may have elite contact; in another the leadoff hitter may have middle-of-the-order power. Editable chase, discipline, contact and power ratings should override generalized roles.
Some pitch shapes and movement directions play differently against right- and left-handed hitters. Matchup rules are not universal: pitch shape, command, release, hitter strengths and individualized splits must govern.
Repeated pitches can be useful when executed well, but repeated speed–location combinations may become easier to anticipate. The tool gives a modest illustrative benefit to meaningful vertical contrast and useful pitch-pair changes.
This is a working instructional simulation, not a validated prediction system. The five outcomes are mutually exclusive and sum to 100% for a specific candidate:
The display also shows a separate conditional risk: estimated hard contact given that the ball is put in play. This is not a sixth outcome and must not be added to the five-category total.
Inputs are assumed: target-specific zone probability, model swing likelihood, pitch whiff tendency, hitter traits, command and sequencing bonuses. The app scores the candidates using an adjustable conceptual utility formula, not measured expected runs. Use scores comparatively, never as a real-world success rate.
Known limitations: no pitcher's real tracking data, no batter scouting records, no velocity/spin/movement calibration, no umpire zone, no true pitch trajectory, no validated probabilistic training set. The simulation is not a substitute for coaching judgment.
To publish calibrated outcome probabilities, import pitch-by-pitch data, separate training from validation samples, account for count, pitch type, trajectory, movement, handedness, hitter, game situation and execution, then report sample size, uncertainty and model reliability. Do not infer a real hitter's ability solely from batting order.
For true tunneling metrics, use flight-coordinate data to estimate release, decision-point and plate separation. The illustration in this demo does not calculate those distances.
Planned areas—not capabilities claimed for this prototype