Positioning systems in self-organizing networks generally rely on measurements such as delay and received signal strength, which may be difficult to obtain and often require dedicated equipment. An alternative to such approaches is to use simple connectivity information, that is, the presence or absence of a link between any pair of nodes, and to extend it to hop-counts, in order to obtain an approximate coordinate system. Such an approximation is sufficient for a large number of applications , such as routing. In this paper, we propose Jumps, a positioning system for those self-organizing networks in which other types of (exact) positioning systems cannot be used or are deemed to be too costly. Jumps builds a multiple coordinate system based solely on nodes' neighborhood knowledge. Jumps is interesting in the context of wireless sensor networks, as it neither requires additional embedded equipment nor relies on any node's capabilities. While other approaches use only three hop-count measurements to infer the position of a node, Jumps uses an arbitrary number. We observe that an increase in the number of measurements leads to an improvement in the localization process, without requiring a high dense environment. We show through simulations that Jumps, when compared with existing approaches, reduces the number of nodes sharing the same coordinates, which paves the way for functions such as position-based routing.