#hashcodesolved
Self-driving rides
More than 48 million score.
Main idea - find best ride for each car when car becomes empty. To find it we calculate some rate for each ride. Biggest rate - better ride.
how_much_score_I_will_get_if_car_took_this_ride_right_now
basicRate = f(car, ride) = -------------------------------------------------------------
how_much_time_car_will_spend_to_finish_this_ride
progressiveRate = r(car, current_ride) = f(car, current_ride) + Max(f(CAR, ride0), f(CAR, ride1), ... f(CAR, rideN))
where CAR - state of car after finishing current_ride
ride0, ride1, ... rideN - rides that have sense after finishing current_ride
basicRate rate is bad in cases where we have long rides. Like in metropolis test.