​

I’m using the Colley Matrix iteration process (https://www.colleyrankings.com/method.html)

To summarize, you take the win percentage of every team and correct it against the win percentage of their opponents. Then you have a new win percentage. And I repeat that process until the correction factor is below 0.01%. Essential this is what I would expect the win percentages to be if the teams played an average team every week.