Explicit　development　orientations　of　ports　are　critical　to　promoting　port　clusters　from　chaotic　competitions　to　a　hierarchical　collaborative　development.　To　investigate　the　development　orientation　and　hierarchies　of　coastal　ports　in　Mainland　China,　a　visibility　graph　network　was　constructed　in　this　study　for　a　time　series　analysis　of　China＇s　port　throughputs　from　2000　to　2020.　In　this　model,　the　four　throughput　indexes　of　cargo,　domestic　trade　cargo,　foreign　trade　cargo,　and　container　were　used　to　measure　the　scale　of　ports.　The　topological　structural　features　of　the　throughput　network　of　ports,　including　its　average　degree,　diameter,　and　average　clustering　coefficient,　were　calculated.　The　ports　were　then　divided　into　four　types　based　on　their　network　characteristics.　Results　show　that　the　throughput　time　series　network　of　coastal　ports　in　Mainland　China　has　an　obvious　hierarchical　structure　and　generally　has　the　characteristics　of　small-world　networks.　These　ports　are　closely　connected　and　have　a　strong　transshipment　capacity.　However,　an　＂absolute　center＂　does　not　exist　in　this　network,　and　these　ports　are　in　a　state　of　disorderly　and　chaotic　competition　with　serious　homogenization.　The　proposed　model　provides　an　appropriate　method　for　the　future　planning　and　design　of　port　clusters.