Discussion: O-scores analysis of Flying Pig
in: Flying Pig XIV - U S Interscholastic Championship (Apr 9–11, 2010 - Batavia, OH, US)
Apr 12, 2010 11:29 PM
#
Thanks to cedarcreek for the A-meet data. Analysis is below and here
Official results are at the OCIN web site
Here is my analysis.
Absolute All standings
Patrick Goeres (M-21+), Coureurs de Bois *148.8*(1)
Jon Torrance (M35+), OOC *137.6*(2)
Eric Bone (M35+), COC *135.4*(3)
......
Alison Campbell (F-20), DVOA *95.4*(34)
Absolute Women standings
Alison Campbell (F-20), DVOA *95.4*(1)Corinne Porter (F-21+), DVOA *85.5*(2)
Maiya Anderson (F35+), RMOC *80.1*(3)
By course analysis is here
As always (!) and (!!) means 7.5% and 15% above the estimate for this runner, and (?), (??) correspondingly below.
Additional mark {new} was added recently. It shows runners whose name mentioned first time this year and therefore their score is not reliable due to lack of data.
Terrain was on average @ 1000 GV/km (it took 10 minutes for 100 point runner to get through each nominal km of the race)
Absolute standings analysis:
Interesting is that Jaxon Rickel again was ver close to the top in absolute rankings - overall 6th place running Green.
Greg Ahlswede running Red is the top Red runner and 7th overall
Joe Barrett (M-14), PTOC *100.3*(27) running Orange and some Green is top Orange runner and 27th overall
Alison Campbell (F-20), DVOA running Brown and some red is the top Brown runner and 34th overall
Zachary Lyons (M-14), ROC *94.4*(36) running Yellow and some Orange is top Yellow runner and 36th overall. Notice that Zach is very stable and didn't lost any point running Orange and not Yellow. This is very promising fact
Top White runner is Lucas Lyons (M-10), ROC *63.8*(128) and 128th overall
Women standings analysis:
Alison Campbell (F-20), DVOA *95.4*(1) is the fastest by far, running combination of Brown and Red. Red performance was worse than Brown, but this can not be attributed to the length of the course, as it was about the same difficulty (GV=4200) as both browns
Top Red runner is Corinne Porter (F-21+), DVOA *85.5*(2) and 2nd overall, almost 10 points behind Alison
Top Green runner is Elena Logvina (F40+), GHO *79.7*(4)
and 4th overall
Top Orange runner is Katie Williams (F-16), OCIN *70.7*(10) 10th overall
Top yellow runner is Nikki Singley (F-16), DVOA *65.6*(16), 16th overall
Nikki tried Orange during the event.
Top white runner Julia Doubson (F-10), BAOC *56.4*(29) was 29th overall
Here is updated Women A-series rankings
Here is updated Men A-series Rankings
Enjoy.
Official results are at the OCIN web site
Here is my analysis.
Absolute All standings
Patrick Goeres (M-21+), Coureurs de Bois *148.8*(1)
Jon Torrance (M35+), OOC *137.6*(2)
Eric Bone (M35+), COC *135.4*(3)
......
Alison Campbell (F-20), DVOA *95.4*(34)
Absolute Women standings
Alison Campbell (F-20), DVOA *95.4*(1)Corinne Porter (F-21+), DVOA *85.5*(2)
Maiya Anderson (F35+), RMOC *80.1*(3)
By course analysis is here
As always (!) and (!!) means 7.5% and 15% above the estimate for this runner, and (?), (??) correspondingly below.
Additional mark {new} was added recently. It shows runners whose name mentioned first time this year and therefore their score is not reliable due to lack of data.
Terrain was on average @ 1000 GV/km (it took 10 minutes for 100 point runner to get through each nominal km of the race)
Absolute standings analysis:
Interesting is that Jaxon Rickel again was ver close to the top in absolute rankings - overall 6th place running Green.
Greg Ahlswede running Red is the top Red runner and 7th overall
Joe Barrett (M-14), PTOC *100.3*(27) running Orange and some Green is top Orange runner and 27th overall
Alison Campbell (F-20), DVOA running Brown and some red is the top Brown runner and 34th overall
Zachary Lyons (M-14), ROC *94.4*(36) running Yellow and some Orange is top Yellow runner and 36th overall. Notice that Zach is very stable and didn't lost any point running Orange and not Yellow. This is very promising fact
Top White runner is Lucas Lyons (M-10), ROC *63.8*(128) and 128th overall
Women standings analysis:
Alison Campbell (F-20), DVOA *95.4*(1) is the fastest by far, running combination of Brown and Red. Red performance was worse than Brown, but this can not be attributed to the length of the course, as it was about the same difficulty (GV=4200) as both browns
Top Red runner is Corinne Porter (F-21+), DVOA *85.5*(2) and 2nd overall, almost 10 points behind Alison
Top Green runner is Elena Logvina (F40+), GHO *79.7*(4)
and 4th overall
Top Orange runner is Katie Williams (F-16), OCIN *70.7*(10) 10th overall
Top yellow runner is Nikki Singley (F-16), DVOA *65.6*(16), 16th overall
Nikki tried Orange during the event.
Top white runner Julia Doubson (F-10), BAOC *56.4*(29) was 29th overall
Here is updated Women A-series rankings
Here is updated Men A-series Rankings
Enjoy.
Apr 13, 2010 1:55 PM
#
Small comment to the data organization:
There were 32 course offered 10 first day and 11 +11 two following days. In addition to standard WYOBrGRB, there were off-color courses for IS champs. Effectivelly YOBrG were split on two.
Standard colors are named as X and off-colors as Y
There were 32 course offered 10 first day and 11 +11 two following days. In addition to standard WYOBrGRB, there were off-color courses for IS champs. Effectivelly YOBrG were split on two.
Standard colors are named as X and off-colors as Y
Apr 14, 2010 12:51 AM
#
Why aren't people penalized for MSP, etc.? It looks like they are effectively treated as a sporting withdrawal would be treated.
Apr 14, 2010 3:28 AM
#
@ bishop22: Why aren't people penalized for MSP, etc.?
Because the goal is to evaluate the speed. Ranking is by-product...
My goal is to be addition to your chronometer. Stop watch does not punish you for missed point, just tells you the time, right? Same here. I can not use MSP and DNF as they would be misrepresenting of how much time it takes to finish the race - so I do not use them.
Of course one (and bunch of other "purists") is right and number of MSP or DNF means course was difficult and it should be taken into account.
The answer is: there is no mathematically consistent and sound and elegant way of doing it. Any adjustment technique can be easily criticized to death.
So I DO NOT ADJUST, skip or add any result. (Well, to be polite, I also skip those who request "do not count", whatever reason they might have, before or after the race. If somebody is offended I do not feel any stupid database is so important to be the reason for that...)
If the number of runners is large, skipping MSP doesn't affect the result significantly, if the number of people not so big it does affect the result in a predictable way - you see your score and know that it would be somewhat higher if those DNF and MSP finished...
Enjoy and have fun!
P.S. For you math junkies: each person running infinite time affect your score as 1/n_valid_persons. So if 20 valid runners were there and one MSP, your score would be 5% more is he would be assigned infinite time.
If MSP were assigned time twice average, your score would increase by 1/2n = 2.5%. Feel free to contact if you need more detailed explanation.
Because the goal is to evaluate the speed. Ranking is by-product...
My goal is to be addition to your chronometer. Stop watch does not punish you for missed point, just tells you the time, right? Same here. I can not use MSP and DNF as they would be misrepresenting of how much time it takes to finish the race - so I do not use them.
Of course one (and bunch of other "purists") is right and number of MSP or DNF means course was difficult and it should be taken into account.
The answer is: there is no mathematically consistent and sound and elegant way of doing it. Any adjustment technique can be easily criticized to death.
So I DO NOT ADJUST, skip or add any result. (Well, to be polite, I also skip those who request "do not count", whatever reason they might have, before or after the race. If somebody is offended I do not feel any stupid database is so important to be the reason for that...)
If the number of runners is large, skipping MSP doesn't affect the result significantly, if the number of people not so big it does affect the result in a predictable way - you see your score and know that it would be somewhat higher if those DNF and MSP finished...
Enjoy and have fun!
P.S. For you math junkies: each person running infinite time affect your score as 1/n_valid_persons. So if 20 valid runners were there and one MSP, your score would be 5% more is he would be assigned infinite time.
If MSP were assigned time twice average, your score would increase by 1/2n = 2.5%. Feel free to contact if you need more detailed explanation.
This discussion thread is closed.