If some choice produced a graph with largest component of a graph smaller than , the balanced component cut result with packing parameter one would give an empty graph cut whose sides both have at least vertices. For any fixed such vertex set partition, a uniform edge of the complete graph crosses with probability
To avoid crossing in row , at least one of its offered edges must be noncrossing. The row's probability is . The rows consist of independent random variables, so at a union bound over at most vertex set partitions gives
For example, the positive integer
makes this bound at most . Therefore every choice has a component of order at least , with probability tending to one. The simultaneous giant for fixed random-edge choice estimate includes choices made after seeing the entire array; it does not assume a particular online selection rule.
Let , and . A graph cut with no crossing edges is obtained by assigning whole graph components to its two sides. Among all such assignments, maximize the smaller side's size , and write for the other side.
Suppose . Then . If a graph component on the larger side had size , moving it to the smaller side would make both and greater than . That contradicts maximality. Hence every graph component on the larger side has at least vertices. Since each has at most vertices and their total exceeds , there must be at least of them. Consequently
a contradiction. Thus , proving the required empty cut with both sides at most . This balanced component cut argument works for arbitrary positive component weights as well; it does not require divisibility of .
Use the standard Achlioptas process with two independent uniform candidate edges per step and a rule choosing one. The component-band vertex count counts vertices, not the number of graph components. Integer time rounding changes the arguments below by steps.
First, at the eventual graph component of order at least is assembled from graph components already present at . Contract those initial graph components. During steps, at most new edges are added, so the eventual connected contracted graph contains at most initial graph components. Initial graph components of order less than can therefore contribute at most
vertices, for sufficiently large and fixed , since . For their contribution is zero. Consequently .
We now use forced merging of large components. If at a time the graph components of order at least contain at least vertices, then after at most steps there is a graph component of order at least , except on an event of probability exponentially small in . One explicit choice is . The proof is a union bound over set partitions of the at most initial large graph components: if all final graph components were small, a balanced component cut would split their initial vertices into sets of size at least , and both candidate edges cross that fixed cut with probability at least in every step. The rule cannot avoid such a forced crossing. The failure bound is , and a union bound makes the estimate simultaneous over all starting times at most , for each fixed .
Set and choose a fixed integer with . If , this lemma would produce a graph component of order at least by time
That contradicts the assumed . Thus , and subtraction gives
The same works for every fixed ; it depends only on , with the number of offered edges fixed at two.
Offer independent uniform edges of the complete graph per row and permit any one to be selected from each row. If a chosen graph has no graph component of order at least , the balanced component cut provides an empty cut whose sides both have at least vertices. Each candidate edge crosses that cut with probability at least , so a row can avoid crossing with probability at most . The union bound over at most cuts gives failure probability at most . Any fixed integer with suffices. The guarantee is simultaneous even for choices made after seeing every offered edge.