Whenever we ply a tournament (especially a large one), we often try and estimate our likely or target score. What this score is usually depends on the strength of the field, how we are feeling, and an inflated sense of confidence. However this is an inexact science, often made harder by the vagaries of tournament pairings.
A slightly more scientific approach (and I do mean slightly) is to estimate your 'bounce' number. This assumes that at some point is a large swiss you will begin to 'bounce' between higher rated and lower rated opponents. For example, if you are seeded just above the half way point, you should win your first round, lose the second, win the third etc. Of course you may score an upset win (or unlucky loss), but +1 is your target score for an event with an odd number of rounds.
Calculating your bounce number is fairly easy (if you ignore boundary cases). Just take Log base 2 of the number of players/2 and Log base 2 of your seeding number. The difference between the two should be the number of games you will win (or lose) at the start of the event, before you go win, loss, win, loss for the rest of the event. So in a 128 player field being seeded 20th should see you win your first 2 games (against seeds 85 and 53) before losing to seed 4.
Of course this assumes that the higher seeded opponent will always win, but at least it is one way of judging how well you might do in an event,