# Account centrality This is categorised as an [account metric](/xnUSdbbaQS-wn0KmIYRh-Q). [Wikipedia article](https://en.wikipedia.org/wiki/Centrality): 'indicators of centrality identify the most important vertices [nodes] within a graph.' There are multiple definitions/metrics of centrality, all based on the [degrees](/YLHBTc1gTqStlIe9GOBEAQ) of nodes (club members). For a directed network (in which relations have a direction), 'indegree' and 'outdegree' centrality can both be determined. Following the notation established [here](/2iRH2t-QRLOYL3cHkbCgTw), indegree for a given member is associated with exchanges for which that member receives credit, and outdegree with exchanges where the member spends/issues credit. The simplest indicator is the 'degree centrality,' which is just the number of trading relations a club member has. This is clearly an aspect of engagement. 'Betweenness centrality:' how many times a node acts as a bridge along the shortest path between two other nodes. This may be useful for analysing supply chains - if for a given sector (e.g. food) there are a number of businesses operating as distributors between producers and retailers but only one of them has a high betweenness then it is likely that they are dominating that niche. The above consider only network topology, but nodes can change *state* following an interaction, e.g. if a rumour is spreading through a network a node can change from 'not received' to 'received,' which will influence its subsequent interactions with other nodes. This may be worth exploring in the context of club failure modes; if a member maxes out their available credit this limits their interactions in the immediate future to selling, which may have knock-on effects. 'Percolation centrality' can be used to measure the importance of a node in terms of how much it aids the spread of such phenomena (see also [percolation theory](https://en.wikipedia.org/wiki/Percolation_theory)). '[Node influence metrics](https://en.wikipedia.org/wiki/Node_influence_metric#Accessibility)' are intended to overcome the limitations of centrality measures, which identify the *most important* nodes but do not necessarily provide a universally applicable ranking of the relative importance of *all* nodes. 'Accessibility' quantifies how accessible the rest of the network is from a given node. In this context a 'walk' is a series of adjacent trading relations where each club member is only included at most once. Accessibility can be calculated for a given walk length (i.e. number of trading relations followed) and probability of trade at each step in the walk. This suggests several ways to increase accessibility: * Increasing the number of possible walks by connecting the member of interest with other members such that they begin trading * Increasing the probability that the member of interest will trade with its existing partners * Connecting other members with each other - whether or not this results in a new possible walk will depend on how closely connected they already are to the member of interest. * Increasing the probability that members along the possible walks will trade with each other. Being direct, the first two interventions would be the obvious place to start; see [stimulating trade](/xgggmZ4qSiyn_WGsqoj49g).