Greetings, That definitely seems wrong to me also, unless the network is so dense that it isn't getting spread out at all, or the normalized values are being swamped by some outliers. You may want to take a look at the range of your values to see if you have a couple of extremely strong (or extremely weak edges). You may also want to consider applying a cutoff. To directly answer your question, though, if you are using the defaults (e.g. heuristic interpretation) the weight value is interpreted as a distance.

-- scooter