Picture buying 2% of a company at the seed stage. The company then raises a Series A and issues new preferred shares to the incoming investors. You didn't sell anything and your share count is unchanged, but the pie now has more slices, so your 2% might fall to 1.4%. Raise a Series B and it drops again. Each round you don't participate in chips away at the percentage, even as the price per share climbs.
The thing to internalize is that dilution isn't inherently bad — owning 1% of a company worth $500M beats owning 2% of one worth $20M. What matters is whether each round raises the value of the company faster than it shrinks your stake. Two things make dilution worse than it needs to be: oversized option pool top-ups, which come out of existing holders before the new money arrives, and down rounds, where new shares are issued at a lower price than the last round. Angels with pro rata rights can defend their percentage by investing more in later rounds, but most simply accept dilution as the cost of the company growing.