Many symmetry breaking methods have been proposed so far. Previous works have shown that these methods could be combined together under some conditions. We use a different angle : we compare the pruning power of two symmetry breaking methods. The first one is the rather classical SBDS method. The second one is the recently proposed dynamic lex constraints (DLC). We theoretically show that DLC prunes more nodes than SBDS if values are tried in increasing order. We also show experimentally that DLC can be more efficient than SBDS.