Run Support vs. Defensive Breakdown Counterfactual Analyzer
Toggle defensive miscues to simulate counterfactual routine outs and observe recalculated win equity.
Use this toy to examine an assumption, then separate the arithmetic from what an actual inning would have required. A displayed lead is the result of fixed subtraction, not evidence that removing a defensive play would guarantee victory.
The starting score is Toronto 6, Tampa Bay 7. The two assumed costs are two runs and one run. With both active, the remaining defensive cost is 3, the margin is −1 and the toy score index is 0%. Remove the two-run cost only: Tampa Bay becomes 5, remaining cost 1, margin +1 and index 84.6%. Remove the one-run cost only: Tampa Bay becomes 6, remaining cost 2, margin 0 and index 50%. Remove both: Tampa Bay becomes 4, remaining cost 0, margin +2 and index 84.6%.
The score index is a branch-selected constant: any lead gets .846, any tie .500, the baseline 6–7 gets zero, and other deficits get .18. No historical probability table is queried. The separate 78.2% and84.6% cards are fixed illustrative values; they do not update as measured probabilities. The nine chart points also use authored values and multipliers rather than plate-appearance data. They do not include outs, runners, batter count, team strength or alternative scoring after a play changes.
The “Remove One-Run Miscue” preset keeps the two-run cost and removes the one-run cost; it does not alter the starter/bullpen split. Toronto’s six runs remain fixed in every state. Adding the two assumed costs presumes no overlap, interaction or downstream change. A defensive misread is not automatically an official error, and a presumed run cost is not an official earned-run ruling.
MLB’s error glossary explains official-scorer judgment. For empirical probabilities, MLB’s Game Strategy Explorer compares comparable historical situations and documents its dataset and extra-inning conventions. These sources explain concepts, without validating this page’s invented events or coefficients.
Toggles and presets save the current state in this browser. Reset restores the authored baseline. JSON exports offensive splits, both miscue flags and recalculated opponent runs; its legacy historical_context field contains example constants, not observed rates. CSV is a summary and omits individual flags. Neither format provides an import workflow. No real game attribution, betting edge or causal defensive valuation has been established.