I agree with what's stated above, my rosters are geared towards game play. I use a system of ratings and combine results to gauge basic att's like speed, agilty, strength, jumping, and other atts are given by season statistics and a ton of deliberating among 7 guys who are 2k lifers and play an official 2k season with me each year. The big rating we use is consistency, we use this as our main att meaning we have a scale for ex, the top is 105 for elite star players, 100 super stars, 95 stars all the way down to 70 being the bottom. So if you want to know how good a player is in our opinion just look at their consistency rating and that will give you a general idea. We also spent around 2 hrs per team soley on players skin tones and equipment. If I feel the game needs a tweak to accomplish something on the field then its done. I like to have my player ratings to be were they should be, but I'd rather have the game play how I want to, rather than worry if a players rating is too high or low. Oh yeah I HIGHLY suggest using the -25% method if using my rosters. For those of you who like my rosters but want the overall rating to reflect a rating not so high, use the -25% on all teams rather than soley on your own, and the defensive ratings will be more traditional. In the end I suggest what dan stated give them all a try. There all fun in their own way. Wow thats the most I have typed in a while on this site, I'm now going to go get my butt kicked at Call of Duty 5. Dang xbox live 12 year old kids, that own me.