The computer ratings become more accurate as the season progresses. The computers also begin with speculated positions to start with. I wouldn't go with computer ratings for the final rating, but I believe the strength of schedule would be a valid variable to consider for the raters. It just puts into context what everyone already knows Such as if you play in the Nike tournament your strength of schedule goes way up.
I don't know how the strength of schedule works. I am figuring that they tally the total of the teams played. I That probably works for teams in the mid to lower tiers, but for top level teams not so much. To them anyone but the top tier teams are no threat so it doesn't matter to them if they play someone ranked # 500 or someone ranked #3000. I can remember when SMS played in their local league and had to really work to stay under 100 points. It made no difference to them who they played 2500 rated team or a 150 rated team. They still blew them our with the only difference being on how long the third team played. The league was so weak that it surely affected their computer rankings, however. They still finished in the top three but those 3500 rated teams really pulled them down. This is just an example.
Now lets say a team plays nothing but say #500 teams and has one loss. While another team plays some teams around 500 but also some teams rated around the 350mark Then it plays a number of teams in the top 50 and ends up with one loss. Now using the computer rankings the first team would have a higher strength of schedule ranking and thus all things being equal would be rated higher by computer. You could also get blown out by the top rated team in the country and your strength of schedule goes way up. Why should you get more credit for getting blown out by # 1 rather than a 1pt loss to say number 10.
This is why teams need to be rated by people not computers.
