Pats = 3.99
Cowboys = 3.6
TB = 2.5
Take the # of quarters that Eli plays (well) and if that is greater than the opponent's team rating then Giants "can" win. In last week's game, for example, Eli played for 3 quarters, 3 < 3.99 so the Pats win. (Eli had to play 4 quarters for the Giants to pull off the win.)
Eli is simply inconsistent, so if you know whether he is showing up and for how long, that imo will dictate success of the team or not. In this weekend's game vs TB, if Eli plays 3 quarters vs TB (2.5), 3>2.5 so the Giants win. If Manning plays 2 quarters it rates to be insufficient. Obviously there is no magic formula, but the point of this is to highlight the acute nature of Manning as the weakest link in the Giants chain for winning and losing. His play is so volatile, and the team is dependent upon him to come through for success. It also articulates why predicting Giants performance this year is hit and miss. If Eli is there, the Giants compete; if he is not, the Giants are not there either.