There is nothing in the structure of the tie-break that gives advantage to the servers more than regular sets. The only reasonable and defensible reason that TB's are 'better' for big servers is that they have more practice in them (but even this is barely borne out by statistical evidence). The serve being easier to perform under pressure is another plausible hypothesis, but it turns out to be false: http://heavytopspin.com/category/tiebreaks/
. This blog (not mine) has some perceptive analyses about tie-breaks but the 'tie-breaks structurally favour big servers' hypothesis does not gain support from the myriad statistical evidence the author examines. It really irks me when TV commentators pull out this line, the perpetuation of myths isn't good for understanding a game.
Not sure about this. To me it makes sense the best servers would have an advantage. They are far more likely to win service games to 0 or 15, after all. In a normal set that doesn't make any difference as holding your service games is all that counts, whether it's to 0 or to 30. But in a tie-breaker, if you're less likely to lose points
on serve, that does
make a difference as the breaker is a matter of points. Of course you could make the same case about the best returners, but even in today's conditions, the serve is more powerful than the return, and the best servers win more points on serve than the best returners on return.