Probably not. Nine games of data is not really that strong of evidence. The algorithm probably gives each player a prior such has assigning one fake loss in order to limit the maximum ranking of players with few games and no losses. This prior is discussed in the paper on the ranking algorithm used here: http://www.remi-coulom.fr/WHR/WHR.pdf