New combinatorial perspectives on MVP parking functions and their outcome map

TL;DR

This paper explores the MVP parking problem, analyzing the outcome map, its fibers, and special subclasses called Motzkin parking functions, revealing new bounds, enumerations, and connections to combinatorial models like Motzkin paths and the Abelian sandpile.

Contribution

It introduces new combinatorial insights into MVP parking functions, improves bounds on fiber sizes, and connects the outcome map to Motzkin paths and the Abelian sandpile model.

Findings

Linked fibers of the outcome map to subgraphs of the inversion graph.

Provided bounds on fiber sizes of the outcome map.

Derived a closed formula for MVP parking functions with complete bipartite outcomes.

Abstract

In parking problems, a given number of cars enter a one-way street sequentially, and try to park according to a specified preferred spot in the street. Various models are possible depending on the chosen rule for collisions, when two cars have the same preferred spot. We study a model introduced by Harris, Kamau, Mori, and Tian in recent work, called the MVP parking problem. In this model, priority is given to the cars arriving later in the sequence. When a car finds its preferred spot occupied by a previous car, it "bumps" that car out of the spot and parks there. The earlier car then has to drive on, and parks in the first available spot it can find. If all cars manage to park through this procedure, we say that the list of preferences is an MVP parking function. We study the outcome map of MVP parking functions, which describes in what order the cars end up. In particular, we link the fibres of the outcome map to certain subgraphs of the inversion graph of the outcome permutation. This allows us to reinterpret and improve bounds from Harris et al. on the fibre sizes. We then focus on a subset of parking functions, called Motzkin parking functions, where every spot is preferred by at most two cars. We generalise results from Harris et al., and exhibit rich connections to Motzkin paths. We also give a closed enumerative formula for the number of MVP parking functions whose outcome is the complete bipartite permutation. Finally, we give a new interpretation of the MVP outcome map in terms of an algorithmic process on recurrent configurations of the Abelian sandpile model.