I'll just point out that switches aren't always reliable, especially when you need them to change rapidly. It may take a thousand plays through a map with a hundred switches all being toggled non-stop with hyper triggers running at the same time for you to experience it once, but when a switch fails to set or clear, you'll know it.
A much more reliable substitute for switches is death counts, and they are much easier to work with, since they can be assigned player-specifically with a single trigger. I use both, myself, but use death counts about a thousand times more often than switches.
Basically, instead of using the action "Set Switch 1," you would use the action "Modify death counts for Player 8: Set to 1 for Jim Raynor (Vulture)." Instead of clearing Switch 1, you'd set Player 8's death count for Jim Raynor back to 0. Instead of using the condition "Switch 1 is Set," you would use the condition "Player 8 has suffered exactly 1 death of Jim Raynor (Vulture)."
Not only can this be done on a player-by-player basis, but it also allows you to use the same "switch" in more than two modes (instead of "Set" and "Clear," you have 0, 1, 2, 3, 4, etc).
Just something to keep in mind. As for your original question, a switch is nothing more than a tool to be used in trigger conditions. One trigger turns a switch on, another trigger performs actions repeatedly while the switch is on. Another trigger turns that switch off, and the repeating trigger turns off.
Not too bad when you get used to using them.