Well, playing the same Mage all the time will make you predictable to a certain degree. It allows your opponents to tailor their spellbooks to have answers to your most likely threats. So, unless you create spellbooks that are very unusual for the type of Mage you're playing, you're basically playing with a handicap.
That alone would be sufficient reason for me to regularly switch to a different Mage. (In the past, the Wizard was a possible exception due to its flexibility, but now that they're only trained in the Air Elemental school, they've become just as predictable as the other mages.)
Totally agree with that. While learning the game, we used to play the same mage over and over again, while changing cards. In the first game I was blown to pieces by the other mage. In the second game we had the same cards, but I changed my game play and it became a longer game - but I still lost. So in the 3rd and 4th game I changed some cards I felt I needed. I won both of these more and more convincing.
Then my partner changed his mage and I lost again with that winning deck.
So what you need to do when building a deck is make a decision:
a) Do I go full risk and go for a play style that is very powerful, but if the wrong mage is on the other end, I might lose?
b) Do I imagine all kinds of mages and come up with an allround mage that is able to counter an aggressive opening, a creature spammer, a hex spammer, etc.
I haven't figured out the right strategy yet. Following a) can lead to spectacular wins, but also spectacular losses. Following b) leads to longer games, but not always a win.
I guess the trick is to find something effective in b) (a base set of cards) to allow you to survive any mage for the first rounds and then try to still implement something like a).
Deck building is an exciting part of the game I think.