Register | Login
Attackpoint AR - performance and training tools for adventure athletes

Discussion: Top performers analysis

in: US Sprint & UltraLong champs (Sep 19–20, 2009 - Rochester, NY)

Sep 23, 2009 10:24 PM # 
O-scores:
Following runners were exceptionally good on ultra long relative to their average performance : Source
------------------------------------------------
Runner Improvement
Ian Jones 37.4%
Joseph Sackett 35.9%
Larry Zygo 24.9%
Adela Danciu 23.7%
Jeff Eames 23.4%
Jerome Young 19.6%
Ole Bergset 18.6%
Dancho Hristov 17.3%
Molly Kemp 17.2%
Brian Sutherland 16.8%
Fred Reed 16.7%
Nathan Collinsworth 16.1%
Sandy Fillebrown 15.9%
Sergey Velichko 15.6%
Dave Kostansek 15.3%
Sam Reed 15.1%
Carl Underwood 14.5%
Cathy Hayhow 14.4%
Dayle Lavine 14.0%
Valery Doverov 14.0%
Scott Thatcher 13.5%
Mary Jo Childs 13.3%
Gregory Balter 13.1%
Maricel Olaru 13.1%
CHARLIE DEWEESE 13.1%
Gail Gagarin 12.9%
Andrew Eck 12.6%
Nolan Miles 12.6%
Clinton Morse 12.5%
Vadim Masalkov 12.1%
Charity Masaitis 11.4%
Peter Dady 11.2%
Larry Pedersen 11.2%
Addie Nolan 11.0%
Julie Keim 10.4%
Nick Duca 10.0%

===========================
Similar data for Middle
---------------------
Runner Improvement
Elitsa Healy 53.2%
David Levine 34.6%
Ed Hicks 29.9%
Mary Underwood 25.2%
Charles Houk 22.0%
Hannah Burgess 21.2%
Graham Ziemba 20.4%
Luke Hutchinson 20.1%
Jeff Teutsch 18.5%
David Welsh 18.5%
Meg Parson 17.9%
Tom Nolan 17.5%
Jamie Despain 16.5%
Paul Caston 16.4%
Clive Hayhow 16.0%
Lucas Lyons 16.0%
John Klaben 15.9%
Bob Bundy 15.1%
Nathaniel Lyons 14.7%
Richard Gaylord 14.6%
Emma Waddington 13.8%
Eric Kemp 13.1%
Michael Hendricks 13.0%
Ben Parson 12.6%
Laura Teutsch 12.3%
Judith Dickinson 12.3%
Walter Siegenthaler 12.2%
Clinton Morse 12.2%
Ekaterina Orekhova 12.1%
Charlie Shahbazian 12.0%
Timothy Dobretsov 11.7%
Vadim Masalkov 11.6%
Phillip Hawkes-Teeter 11.6%
Robert Huebner 11.5%
Kevin Culberg 10.6%
Dean Sturtevant 10.6%
Kathy Bannister 10.5%
Dancho Hristov 9.9%


===========================
Top gainers at Sprint Champs
---------------------

Runner Achivement
Louis Pataki 27.0%
Charity Masaitis 26.4%
Susan Hawkes-Teeter 26.2%
Hannah Burgess 25.0%
Katherine Ebright 23.8%
Melissa Trout 22.9%
Joseph Sackett 22.5%
Andrew Scholle 21.5%
Ethan Childs 21.1%
Sebastian Danciu 20.5%
Mary Jo Childs 20.0%
Jeff Teutsch 19.9%
Larry Pedersen 19.8%
Joseph Steinlage 19.3%
David Levine 19.2%
Emese Orosz 18.8%
David Cady 18.7%
Molly Kemp 18.3%
Zachary Barker 17.9%
Lori Huberman 17.7%
Ian Smith 17.6%
Zachary Lyons 17.6%
Ed Hicks 17.4%
Emma Waddington 16.2%
Zachary March 16.0%
Eric Smith 15.7%
Jason Trump 15.5%
Keith Andersen 14.9%
Thayer Raines 14.9%
Alexander Bergstrom 14.7%
Nathan Ohrwaschel 14.3%
Bob Fink 14.2%
Dayle Lavine 14.0%
Tyler Borden 13.8%
Jackie Novkova 13.7%
Richard Y Ebright 13.4%
Barb Campbell 13.0%
George Walker 12.0%
Corey Thatcher 12.0%
Meg Parson 11.9%
Robbie Anderson 11.3%
Paul Caston 11.0%
Claire Dell 10.9%
Barbara Dominie 10.4%
Rob Wilkison 10.4%
Richard Ebright 10.0%
Advertisement  
Sep 24, 2009 12:39 AM # 
Keith Andersen:
I'm a bit curious to find out how this is calculated. Mind to indulge us on what this percentage indicates?
Sep 24, 2009 4:38 AM # 
O-scores:
With great pleasure

Here is the all men rankings of USMAOC cluster

it is calculated using same algorithm as used by USOF to calculate yearly ratings with three corrections (well..., differences)

- Everybody is put into the same data pool ( blue - to - white)
- normalization is done to make average runner score equal to 50
- All events are taken into account, not only A- meets


F.E. Keith Andersen is at 12th place with 17 recorded races during last 12 months, scoring on average (power) 99 and ranking score is 106 ( which is 4 best plus half of the rest, see USOF web page for algorithm description )
Power of 99 means K.A is twice as fast as average orienteer and it will take him 150 minutes to run the same course race which Patric Goeres (power ~150) runs in 100 minutes

Running Sprint with 16:41, 2.5 minutes faster then expected time, evaluates K.A. as ~114 points which is 14% above the expected value of ~99. This means he was faster than himself on average in this particular race.

As older slow races go out of the scope and new faster continue to be delivered, overall score and power will grow.
At the end of 2008 K.A. was 13th in a club with power 94

And if you are still not too tired look
here

Absolute score calculation allows to compare different color of the races and in this case shows that Ethan Childs (power 101) who normally runs about the same speed as K.A. was running Green Sprint and ran much faster, which allowed him to pass Clem M and catch Eric Kemp.

Enjoy
Sep 24, 2009 12:43 PM # 
j-man:
Any way you can get a server cluster to host this? :)

There is a lot of data there and it is slooow. But interesting.
Sep 24, 2009 12:55 PM # 
Charlie:
Not sure how these things happen, but when I follow through to the USOF rolling rankings, I find that I am both 8th and 10th in M60+.
Sep 24, 2009 1:47 PM # 
O-scores:
@j-man
it is at the cluster and server speed s not a problem

Problem is that php creates too large html file and transfers it then to user.
I didn't figure out how to make structure useful while creating reasonably small chunks of data on demand.
any php/java volunteers/advisors would help.... I will eagerly open the code and the database.
@ Charlie:
I assume you are talking about my webpage?
technically this might happen if you compare yoursellf with different groups of people or select classic only, classic and sprint or sprint only. Provided yours and others abilities vary in this disciplines place will be different in different versions of a rating.
having said this I found you being twice mentioned as charles and as charlie, whic hwas treated as different people, fixing..
Sep 24, 2009 2:30 PM # 
j-man:
I was beginning to suspect it was the PHP and HTML. I've had the same problem with large HTML tables generated with PHP.

And IE is bad at rendering these. I witched to FireFox and it was more tolerable, just improving the client-side experience.
Sep 24, 2009 4:06 PM # 
ken:
krechet, you need gzip output compression to fix your html download-size problem. almost all servers and browsers support this, and most large websites use it. it makes a very big difference. I have not dealt with it in php, but 2 seconds on google gives:

http://www.webcodingtech.com/php/gzip-compression....
Sep 24, 2009 6:40 PM # 
O-scores:
Thanks Ken
Implemented

This discussion thread is closed.