okansas.blogspot.com
Occassional thoughts about orienteering


Tuesday, July 31, 2012

Watching "Medaljens Pris"

 

A bunch of the Swedish orienteers I follow on Twitter have been watching a Swedish sports documentary called "Medaljens Pris."  I figured it was worth take a look, so I fired up a web browser and watched it on SVT Play.

The documentary is about elite track and field athletes and the price they are willing to pay - mostly in terms of injuries - to try to reach their best.  The athletes come across as sympathetic, reasonable and nuts.

A lot of TV documentaries take a clear stand.  People are either good or bad.  Smart or dumb. Strong or weak.  In other words, make people a lot less complicated than they really are.  But "Medaljens Pris" doesn't really do that.*  And that's a strength of the documentary.

A few quick thoughts:

1.  When I was watching the temperature outside was 105F (41 C) and watching a hill interval session in the forest with some snow on the ground was very nice (around 9 minutes into the documentary).

2.  Watching video of athletes getting hurt is not fun at all.

3.  If an athlete can figure out how to train without getting injured they'll have a huge advantage.  That balance of short-term versus long-term is tricky to find.

4. I can understand taking a pain medication to deal with the discomfort of an injury, but taking pain medication to hide an injury so that you can train is nuts.  But I can also understand it.

5.  Training for orienteering - running through the forest with a map in your hand - is a whole lot more fun than training to run on a track.

* The coach they interview comes pretty close to being portrayed as one dimensional and that one dimension being bad.

Back to okansas.blogspot.com.

posted by Michael | 8:14 PM

0 comments




Sunday, July 29, 2012

Some MTB orienteering mapping

 

I spent some time today working on a mountain bike orienteering map.  I was inspired by the Swedish Orienteering Federation's step-by-step guide to converting an orienteering map to a MTB orienteering map (the PDF is worth a read if you can manage Swedish).

I began with the existing orienteering map. As I ran the trails this morning, I made notes about how to map them.  If I understand the MTB O' mapping standards correctly, you draw trails based on width (> 1.5 meters or < 1.5 meters) and riding difficulty/speed.  It seems straightforward.  Single track trails are < 1.5 meters and the occasional double track is > 1.5 meters wide.

Judging the speed is a bit difficult for me because I'm a lousy bike handler.  I tend to slow down more than a good mountain biker would.

The map I was working on today is the overlook area at Clinton Lake.  The trails are generally nice, with a few rough and rocky areas.  There are some hills.  There are some paved trails and roads.  I think it'll make a reasonable area for a local event.

My other MTB mapping project is the river trail area which is largely built on a trail network of one-way trails.  After I wrote about mapping those trails, I got an email with a new symbol that works pretty well.  Take a look below:

Back to okansas.blogspot.com.

posted by Michael | 3:37 PM

0 comments




Sunday, July 22, 2012

Mapping "one-way" trails

 

I've been working on a mountain bike orienteering map.  The area is small and flat with mainly single track trails that run along the river.  The trail system is essentially a big loop of a one-way trail.  This raises a mapping problem that I'm trying to solve.  How can I best show a one-way trail?

My current solution is below.  The purple arrows indicate the direction you can ride.  I'm not really satisfied but I haven't come up with a better solution (yet).



I've also considered showing the direction of travel arrows with bright green.  I was worried that purple would get confused with the course and an orienteer wouldn't even notice it.  On the other hand, the same thing could happen with the green.

The east-west trail in the map above is 2-way travel.  I'm thinking that I may need to have something to indicate that on the map.  One thought was to highlight the 2-way trails with a series of purple dots paralleling the trail to indicate that 2-way travel was allowed.

I'm thinking the best solution might be to field test a couple of options and ask the orienteers what they thought.

Back to okansas.blogspot.com.

posted by Michael | 5:25 PM

3 comments




Friday, July 20, 2012

US WOC relay team prediction

 

My super-simple forecast for the US relay teams at tomorrow's WOC relays:

Men 27th
Women 19th

The forecast method is very simple.  I looked at the results of the WOC relays since 2003 and gave a slightly higher weight to the last three years.

Back to okansas.blogspot.com.

posted by Michael | 7:30 PM

1 comments




Thursday, July 19, 2012

Looking ahead to the WOC relay

 


I've been thinking about SPROUT-like projections for the WOC relay races.  I haven't made any projections, but I did take a look at USA v Canada.  Canada doesn't have a women's team this year, but both USA and Canada have teams and have announced their runners.  There is always a friendly rivalry between the two nations at the WOC.  So it is fun to look ahead.

The graph below makes it pretty clear that the USA is the underdog.  The graphic shows world ranking points since 2010.  The first three runners (Boris, Eric and Giacomo) are this year's US team.  The graphic shows you the world ranking points for each of the runners for 2010-2012.  It includes results from this year's WOC.




While the US is the underdog, you can imagine a tight race.  Imagine the US runners perform above their medians and the Canadians perform below and you can see a tight race.

I'm not sure of the long term head-to-head results of USA v Canada, but in my younger days I had an even record in WOC relays against Canada.

Some details about the graphic...The horizontal lines indicate the median.  The boxes show the range from the 1st to 3rd quartile.  The width of the boxes indicates the number of ranking points included (e.g. Giacomo has a narrow box to indicate relatively few world ranking scores).  The height of the boxes indicates the range of points.  You can see that Eric and Robbie have relatively consistent scores (i.e. short boxes), while Giacomo and Will have greater ranges.  The single small circles show the more extreme outliers.













Back to okansas.blogspot.com.

posted by Michael | 8:31 PM

0 comments




Monday, July 16, 2012

SPROUT overview

 

Trying to predict US team results at the WOC was an experiment.  In part, I wanted to see if a really simple model would work.  In part, I'm always interested in ways of thinking about how a team is doing and trying to understand if a race is "good" or "bad."

SPROUT predicts the place of each US runner in each WOC qualifying race.  After 17 results, I think SPROUT worked pretty well:

4 times the prediction and the actual results matched
3 times the prediction and the actual results were 1 place apart
3 times the prediction and the actual results were 2 places apart
3 times the prediction and the actual results were 3 places apart

So, of 13 of 17 results were within 3 places.  The median difference was 2 places.

The worst prediction was off by 10 places.  Ali Croker's sprint qualifying result was much better than SPROUT expected.

The graph shows the difference between SPROUT and actual results.  You can see that the difference clusters around 2, with two outliers.


SPROUT is incredibly simple.  All it takes are a ranked list of the runners on the team (I used WRE points to put the runners in order) and a "normal  US place" and a number to add to or subtract from the normal US place.  I think a small tweak - looking at the actual values of the ranking list, not just the order - might have improved SPROUT a little bit.

Back to okansas.blogspot.com.

posted by Michael | 7:39 PM

4 comments




Sunday, July 15, 2012

Another test of SPROUT predictions

 

Another day of WOC qualifications means another test for SPROUT.

Today the US Team had one runner qualify for a final.  The places of all six runners were: 13, 17, 22, 27, 30 and 33.  If you look at the group as a whole, SPROUT predicted 13, 16, 22, 24, 32 and 32.

Compare those two lists:

13 17 22 27 30 33 (today's results)
13 16 22 24 32 32 (SPROUT's prediction)

This makes SPROUT look a little better than it is.  SPROUT had Ali C qualifying and Samantha S just missing.  In fact, it was the other way around.

Back to okansas.blogspot.com.

posted by Michael | 10:37 AM

0 comments




Saturday, July 14, 2012

SPROUT worked well today

 

After looking at the results from today's WOC sprint champs, it looks like SPROUT had a very good day.

SPROUT had projections for 5 US runners.

SPROUT got one of the 5 exactly right (i.e. predicted the place that the runner finished).  SPROUT got two of the runners within one place.

SPROUT got Ali Crocker's place wrong, but it did have her qualifying for the final.  SPROUT had only one US runner qualifying for the final.

SPROUT completely missed one runner - missing by a bunch of places (and a bunch of time).

I'm surprised how well the simple model worked.  But, it has only been tested once.  I'm interested to see how it works for the middle and long qualifications races.  

Back to okansas.blogspot.com.

posted by Michael | 11:30 AM

0 comments




Friday, July 13, 2012

Looking ahead to tomorrow

 

I was thinking about a simple way to forecast how the U.S. team will do at this year's WOC.  I came up with the Stupid Projection for Results of Orienteering by the U.S. Team (aka SPROUT).  Here's how it works.

Start with a scale of results (i.e. places in qualifying races).  I'm just making up these places to illustrate the system:

Very good  1-9
Good         10-19
Normal       20
Bad            21-29
Very bad    30+

I'll start by assuming everyone on the team will have a "normal" result.

Then, I'll move each runner up or down on the scale based on two things.

First, I'll rank the runners (I could a ranking system like the IOF world rankings or OUSA rankings or even something else, like Team Trials scores).  Second, I'll move the top 2 (of 5) runners up from "normal" to "good" and I'll move the bottom 2 from "normal" to "bad."

Second, I'll give each runner a chance to move up or down based on any recent WOC history.  For example, lets say I've got a runner who has been moved up to "good" but they've consistently run in the top 10 in their previous WOC races.  Under my system, I'll move them up to "very good."  On the other hand, if a runner has been moved up to "good" and they're previous WOC results fit in the anything except the "very good" category, I'll leave them in "good."

Basically, SPROUT treats everyone as a "normal" US runner and then tweaks them a bit.

The S in SPROUT is for "stupid."  That's because the system is purely mechanical.  It doesn't understand that Ross Smith has a broken elbow and won't run.

The system also tends to struggle with disastrous results.  It isn't going to project a DQ, even though a DQ is a possible result.

It also isn't going to project results that are much better than a given runner has done in the past.

I've actually got sources for the different levels (from "very good" to "normal" to "very bad"), the rankings of the runners within the team, and the history of each runner.  I'm not going to post the results, but I'll be interested to see how SPROUT performs tomorrow.


Back to okansas.blogspot.com.

posted by Michael | 7:57 PM

0 comments




Monday, July 09, 2012

And even better

 

I thought the peer nations results from yesterday's JWOC sprint were good, but today's were even better.  Emily Kemp from Canada took the bronze - Canada's first ever JWOC/WOC medal.  North America's first, too.  And, as far as I can remember, the peer nations' first ever.

The peer nations are USA, Belgium, Canada, Ireland, Japan, New Zealand and Portugal.

Belgium should get mention for a gold from this year's European Youth Orienteering Champ.  Tristan Blomen - who finished 12th in the JWOC sprint - won the EYOC sprint in France earlier this year.

Kemp's medal was the highlight of the day among the peer nations.  But, it is worth taking a quick look at the top 20 and catching that Matt Ogden (NZL) and Nick Hann (NZL) finished in the top 20.

I should say something about New Zealand as a "peer nation."  New Zealand fits into my peer nations scheme.  But, New Zealand consistently has the best results of the group.

Back to okansas.blogspot.com.

posted by Michael | 7:46 PM

2 comments




Sunday, July 08, 2012

Some nice results from the peer nations today

 

A quick look at today's JWOC sprint results showed some impressive results from among my list of "peer nations."  Peer nations are a group of seven nations that have some similarities and that I've used as a way of looking at performance.  The group is USA, Belgium, Canada, Ireland, Japan, New Zealand and Portugal.

Among the men in the top 20 at today's spring you'll find 3 of the peer nations represented:

Joao Figuerido from Portugal finished 7th
Matt Ogden from New Zealand was 10th
Tristan Bloemen from Belgium was 12th


Among the women, none of the peer nations had a top 20.

If you're really interested, you can read how I came up with my list of peer nations.

Back to okansas.blogspot.com.

posted by Michael | 1:38 PM

2 comments




Wednesday, July 04, 2012

Looking at 2011 WOC sprint finals results

 


I took a look at the 2011 WOC sprint finals results, looking at the year each top 20 finisher first had a top 20 finish in a WOC final.  The darker shades indicate more a more recent first top-20.  The darkest blue are people who finished in the top 20 for the first time in 2011 or 2010.



You'll notice that quite a few people reached a top 20 in a WOC final for their first time at the sprint. Compared to the 2011 WOC middle and long finals, the sprints saw a lot of "new" names.  

Back to okansas.blogspot.com.

posted by Michael | 1:13 PM

0 comments


March 2002April 2002May 2002June 2002July 2002August 2002September 2002October 2002November 2002December 2002January 2003February 2003March 2003April 2003May 2003June 2003July 2003August 2003September 2003October 2003November 2003December 2003January 2004February 2004March 2004April 2004May 2004June 2004July 2004August 2004September 2004October 2004November 2004December 2004January 2005February 2005March 2005April 2005May 2005June 2005July 2005August 2005September 2005October 2005November 2005December 2005January 2006February 2006March 2006April 2006May 2006June 2006July 2006August 2006September 2006October 2006November 2006December 2006January 2007February 2007March 2007April 2007May 2007June 2007July 2007August 2007September 2007October 2007November 2007December 2007January 2008February 2008March 2008April 2008May 2008June 2008July 2008August 2008September 2008October 2008November 2008December 2008January 2009February 2009March 2009April 2009May 2009June 2009July 2009August 2009September 2009October 2009November 2009December 2009January 2010February 2010March 2010April 2010May 2010June 2010July 2010August 2010September 2010October 2010November 2010December 2010January 2011February 2011March 2011April 2011May 2011June 2011July 2011August 2011September 2011October 2011November 2011December 2011January 2012February 2012March 2012April 2012May 2012June 2012July 2012August 2012September 2012October 2012November 2012December 2012January 2013March 2013April 2013May 2013July 2013September 2013
archives
links