When an archaeologist is confronted with thousands of pieces of pottery all jumbled together, it is next to impossible to reassemble all the broken pieces. We would however like to know about how many objects are actually represented by all these fragments. Life at a site with many small bowls would have been very different from a site with many big jars. Big heavy jars obviously weigh much more than small bowls, so we can't use weights of different kinds of materials to estimate the relative numbers of jars versus bowls in a particular site. The same applies to numbers of sherds. The way out of this problem is to compare all the bases or rims of a particular type, and figure out what the minimum number of bowls or jars had to have been to produce those bases or rims. This will produce a minimum number, because five small fragments of a similar bowl could have belonged to the same bowl, or to 2,3,4, or even five different bowls.
This form of analysis has been performed for the SCC site. These estimates are approximate, but they can be taken as relatively precise guides to the numbers of bowls, jars, etc. which had to be broken to produce the sherds at SCC. This is more difficult for earthenware, because most earthenware vessels had no bases; they had round bottoms, so it is impossible to determine whether a piece comes from an earthenware base or body. The solution in this case is to use rim sherds.