Votes allocated by demographic reach each turn. Final 4 turns = 25%; earlier turns = 75%.
A single seat, decided on plurality. Jamal Martinez took 69.5% against 30.5% for the nearest rival.
Why New York voters lean toward Jamal Martinez over Borat Sagdiyev. Each driver is signed: + = lift, − = drag. Drivers are relative within this list and only move the persuadable slice of each party's vote, so their real effect is small. Coattail rows are direct share tilts in %. A raw vote lead does not equal the sum of these bars.
Ceiling this cycle: at most 20.0% of Borat Sagdiyev's vote can be persuaded across to Jamal Martinez, and 20.0% of Jamal Martinez's vote can be persuaded away. Registration lean sets the starting split and how hard a party is to move, not the final share. These drivers add up to 3.0 pts of that movable slice.