But you have to start somewhere. Do you simply take the previous seasons rankings and resort based on your opinion?I take the past 3 years offensives stats for each team—total plays from scrimmage and calculate the percentage of pass attempts, sacks, rush attempts. If a new coach has been hired during that period I would place more emphasis on those percentages than the 3 year average. Things such as Oline talent impacts the sack % and overall number of plays I project a team to run.Once I project the total number of plays, pass att, rush att I then allocate rush attempts and targets for individual players based on their 3 year averages. If the player is young I put more emphasis on the previous season. After I have all the projections done I use a SOS and travel to tweak individual projections.I then have a ranking I can start from. At that point I can move players up and down the list based on my preference, however I am very conservative about moving players up and down too dramatically.I enjoy doing crunching the numbers and it allows me to find out about certain players that may hold value later in the draft. For instance, I can recall crunching the numbers a couple of years ago and Fred Jackson really jumped out at me because in limited touches he was very productive. I am not saying FJax will wion you a championship, but he makes a very nice injury/bye week option.