1-M and M-1 with Subset Match Rules

Summary: We need to understand how the 1-M & M-1 with subset match rules is trying to identify transactions on the Many side. For instance, say there are 100 transactions on the Many side, does it start checking for 99 transactions to match, then 98, 97.. or does it start by trying to find 1 and then build up to the 100? The reason why we are asking is we are trying to build a 1-M rule that only looks for 2 transactions to match on the Many side, but on days where there are a lot of transactions it gets