A ranking systems' accuracy is measured by its "Predictive Power" (how often a higher ranked team beats a lower ranked team). A Predictive Power of 0% is equivalent to flipping a coin. 100% means the rankings always predict the result of a game (impossible given the random nature of the sport). Some ranking systems only rank the top 25 teams in each age group. That is not helpful if you are trying to flight a tournament. "Coverage" is the fraction of all games where both teams are ranked by the ranking system. Here is how the online ranking systems compare today:
This site ranks 84,000 youth soccer teams in the USA based upon the results of more than 450,000 games played since August 1st 2016. The results come directly from more than 350 different tournament or league websites and are the most accurate and complete set used by any ranking system. The rankings are based solely on team results. Every goal in every game is considered when calculating the ranking of each team. Each team is assigned a score that is measured in goals. To predict the goal difference when two teams play each other simply subtract the score of one team from the score of the other. For example, if Denver Rovers (35.6) plays Parker Stars (33.5) then you can expect Denver Rovers to win by about 2 goals. This works across age groups and even when girls play boys!
The rankings are recalculated almost every day, so please check back often. Teams must play about 10 games before before they are ranked. Sometimes games are split between duplicate teams. I try to fix as many of those as possible. If you spot anything that looks wrong please drill into the team details and click "Report Errors" (at the top right of the screen).
I still need to load results from the following sites: NPL Virginia, www.illowa.org, www.bays.org, www.mnyouthsoccer.org, SCSA, Sports Logic, www.futbolitobimbo.net, www.eddoa.org, Iowa Soccer League, Charlotte United Cup, cpysl.net, New England State Cup, www.lijsoccer.com.
Let me know if you spot anything else!