Last Planet soccer ranking

[Unregistered]

how often does it get updated?

[Unregistered]

When team Onyx requests that it be updated.

[Unregistered]

how often does it get updated?

Frequently. Usually after a significant tournament or the end of a season.

[Unregistered]

Frequently. Usually after a significant tournament or the end of a season.

Basically, when time permits. Haven't figured out how to turn it into a full-time paid endeavor quite yet :)

[Unregistered]

Basically, when time permits. Haven't figured out how to turn it into a full-time paid endeavor quite yet :)

ok cool! thanks!

[Unregistered]

Newb here...what is it?

[All for One]

Newb here...what is it?

http://lastplanetranking.***********...ontpage_5.html

[Unregistered]

Basically, when time permits. Haven't figured out how to turn it into a full-time paid endeavor quite yet :)

Any chance of getting G97 on there?

[lastplanet]

West Coast G01 LastPlanet Ranks updated

April 4, 2014

West Coast G01 Ranks updated based on AllForOne's G01 data updated to matches up to end of March 2014
http://lastplanetranking.************/p/g01.html
Includes WA, OR, CA, ID, NV, AZ, UT

Email lastplanetranking@gmail.com with any corrections and thanks to those who have emailed in corrections and links to tourneys we missed.

Enjoy!

-LP

P.S. The ranking tables are updated monthly.

[Unregistered]

I like that ranking system in principle. I guess we are stuck with garbage in garbage out though. The source data must not be too hot.

[lastplanet]

The G01 source data is very extensive. We make an effort to include most G01 select games in the Western States. All the data is public (just click on a team name in the tables and you'll all the match data for each team). We welcome corrections to the data (lastplanetranking@gmail.com).

Keep in mind that the purpose of the rankings is not to identify the top-10 (say) but to identify teams of similar strength so competitive matches and brackets can be devised.

Here's one of the recent tests using the SoCal Governors G01 prelims. This is a interesting case because it involves many teams (20+ brackets) with teams that play across many different leagues AND this is lower level so teams play fewer games, so there is limited information about the relative strengths. X-axis is prediction re finishing top-2 in bracket and y-axis is what actually happened. You can see a strong correlation between prediction and actually making it through.
[IMG]http://2.bp.************/-waxbZReXAyg/UvGrf-kYJWI/AAAAAAAAT3w/zmZnoMhYJJo/s1600/Rplot.png[/IMG]

However, Oregon has had a lot of mergers lately and that might make Oregon G01 info more uncertain. Certainly the boys U15-18 Oregon strengths estimates are uncertain---need to get match data from the HS boys spring leagues.
Run the match predictor for your upcoming games and see how the model does. If you are doing G01s, click GIRLS at the top (default is BOYS).
http://lastplanetranking.************/2013/05/match-predictor.html
This uses the updated statistical model to generate a prediction of goals for each team. In my experience, this statistical model is pretty good at identifying close matches versus blow-out.

Cheers,

LP

[All for One]

I like that ranking system in principle. I guess we are stuck with garbage in garbage out though. The source data must not be too hot.

Curious why you think this? For G01 alone the database contains almost 25000 games and includes all non-rec teams from WA, OR, ID, AK, UT, NV, AZ, CAN, and CAS. Every league, tournament, and state cup game is included. The data for other ages isn't as all-encompassing but is still pretty complete for WA and OR.

[Unregistered]

Curious why you think this? For G01 alone the database contains almost 25000 games and includes all non-rec teams from WA, OR, ID, AK, UT, NV, AZ, CAN, and CAS. Every league, tournament, and state cup game is included. The data for other ages isn't as all-encompassing but is still pretty complete for WA and OR.

The rankings are fantastic. The guy that thinks otherwise is one of those parents that has a warped sense of reality on his DD's team. That's what is beautiful about last planet. It's not subjective. It's math. Of course every now and then I have found an error in the data. When that happens I have emailed the last planet guy and he corrected it after verifying the accuracy of my input. Good guy too.

[Unregistered]

Curious why you think this? For G01 alone the database contains almost 25000 games and includes all non-rec teams from WA, OR, ID, AK, UT, NV, AZ, CAN, and CAS. Every league, tournament, and state cup game is included. The data for other ages isn't as all-encompassing but is still pretty complete for WA and OR.

If you are familiar with the area, you seed a problem immediately. 1 and 2 are an A and B team. Other teams in the top 10 are second division teams. Completely absent are top teams like FC Portland, Wash Timbers, THUSC, and the like.

So something is wrong. Head to head results would not even support this considering many of the top ranked teams don't even play each other due to many of them being second division teams.

[Unregistered]

It's a pretty cool system. Not as much for the local rankings, we already know those teams pretty well, but it's very interesting when you are traveling out of the area to play in tournaments and know nothing about the teams.