Updated DRAFT 2013 boys ratings
9-21-2013
The tabs on the blog now point to the 2013 ranks. Soon I will add links to the archived 2012 ranks. I don't have the league table done yet nor column re what fall league the team is in.
I've entered the first 1-2 league games. Leagues included are RCL, PSPL, OPL, NPSL, DIII, DVI, and PML.
Email corrections and additions to lastplanetranking@gmail.com.
Some things to keep in mind for the 2013 ratings:
* I have changed the scaling for teams. All teams are scaled against the median U13 team. The median U13 team is set to a strength of 800. This makes the ratings look a little more like chess ratings. 2000 is a very strong HS team and the top teams won't hit this till U17-18.
* Check out the FAQ on the blog to understand the meaning of the strength rating. This is a predictive model and the DIFFERENCE in strength ratings between 2 teams gives you a prediction of the RATIO of the goals scored. If rescaled the ratings, so now a difference of 100 is a prediction that the ratio of goals scored will be 2 (so a 2:1 or 4:2 or 8:4 score, which of those depends on the strength-attack characteristics of the teams). A difference of 200 is a prediction that the ratio of goals scored will be 4 (so 4:1 or 8:2 etc score).
* The ratings are based on 13 months of matches but the ratings are weighted (heavily) towards the more recent 2-3 months of games. On the pdf, you can see how a team changed after try-outs (or after key players added or dropped).
* The "All-xx" tables you see have teams that are from 2012-2013 and don't exist anymore, but the state tables only have teams that played a game in the last 3 months.
* Only teams with 7+ games are listed on the main tables. You can see all teams on the All_Bxx_ranks.html file. Don't make much of the rating of a team with only 3-5 games.
* I have the B02s listed to help me figure out 02 teams that played up in a tourney but didn't list their age. I won't be rating the B02s past this stage. I start with U12s when games are 11v11.
* I have B94s listed, but I won't include those either in the final.
* Oregon ratings will be sketchy until the fall league starts. There were many name changes and reorganization of teams, so more data is needed to get these sorted out.
--LP
9-21-2013
The tabs on the blog now point to the 2013 ranks. Soon I will add links to the archived 2012 ranks. I don't have the league table done yet nor column re what fall league the team is in.
I've entered the first 1-2 league games. Leagues included are RCL, PSPL, OPL, NPSL, DIII, DVI, and PML.
Email corrections and additions to lastplanetranking@gmail.com.
Some things to keep in mind for the 2013 ratings:
* I have changed the scaling for teams. All teams are scaled against the median U13 team. The median U13 team is set to a strength of 800. This makes the ratings look a little more like chess ratings. 2000 is a very strong HS team and the top teams won't hit this till U17-18.
* Check out the FAQ on the blog to understand the meaning of the strength rating. This is a predictive model and the DIFFERENCE in strength ratings between 2 teams gives you a prediction of the RATIO of the goals scored. If rescaled the ratings, so now a difference of 100 is a prediction that the ratio of goals scored will be 2 (so a 2:1 or 4:2 or 8:4 score, which of those depends on the strength-attack characteristics of the teams). A difference of 200 is a prediction that the ratio of goals scored will be 4 (so 4:1 or 8:2 etc score).
* The ratings are based on 13 months of matches but the ratings are weighted (heavily) towards the more recent 2-3 months of games. On the pdf, you can see how a team changed after try-outs (or after key players added or dropped).
* The "All-xx" tables you see have teams that are from 2012-2013 and don't exist anymore, but the state tables only have teams that played a game in the last 3 months.
* Only teams with 7+ games are listed on the main tables. You can see all teams on the All_Bxx_ranks.html file. Don't make much of the rating of a team with only 3-5 games.
* I have the B02s listed to help me figure out 02 teams that played up in a tourney but didn't list their age. I won't be rating the B02s past this stage. I start with U12s when games are 11v11.
* I have B94s listed, but I won't include those either in the final.
* Oregon ratings will be sketchy until the fall league starts. There were many name changes and reorganization of teams, so more data is needed to get these sorted out.
--LP
No comments:
Post a Comment