This, roughly speaking, shaves off the spaghettis, as the weighted profile of digital search trees is identical to that of the core.
In such a case, the contribution of the spaghettis is 1/b times that of the core.
They graphically term this a 'spaghetti effect', and 'utterly unrevealing' (p.272), yet this is not wholly the case.
It is interesting to note that for this case, one can obtain tight bounds on the height without distinguishing the core from the spaghettis.
The spaghettis are the trees remaining after removing the core from the trie.
In particular, the decomposition of the tree into a core and hanging spaghettis yields new connections between tries and the digital search tree.
By figuring out what the core looks like, we can determine when the spaghettis take over.
The constant c is the sum of the contributions of the core and the spaghettis.
