Discussion: Splits for score events
in: Orienteering; General
Dec 10, 2012 2:52 AM
#
undy:
What information do other people think would be good on splits for score events ?
The traditional control no, control code, leg time, total time is probably not so useful.
Possibilities (pick 3 - 5 to make things fit):
control number, points scored, cumulative points, cumulative points with late penalty applied, points/min for that leg, points per minutes to that point, leg time, total time
Example - https://dl.dropbox.com/u/46164711/snap046.jpg
The traditional control no, control code, leg time, total time is probably not so useful.
Possibilities (pick 3 - 5 to make things fit):
control number, points scored, cumulative points, cumulative points with late penalty applied, points/min for that leg, points per minutes to that point, leg time, total time
Example - https://dl.dropbox.com/u/46164711/snap046.jpg
Dec 10, 2012 4:36 AM
#
for individual splits printed at finish: control number, cumulative points (with penalty applied after time limit reached), leg time, total time - the points for each control is usually embedded in the control number anyway.
Also "total points" and "penalty" on header with elapsed time and points score.
one you haven't mentioned, but only applicable for post-event publication, is "relative position on leg" eg 4/17. The leg could be defined by the direction travelled, or using either direction between 2 controls. Most of the usual splits analysis tools are meaningless for score events - this would give an indication of performance on the leg.
Also "total points" and "penalty" on header with elapsed time and points score.
one you haven't mentioned, but only applicable for post-event publication, is "relative position on leg" eg 4/17. The leg could be defined by the direction travelled, or using either direction between 2 controls. Most of the usual splits analysis tools are meaningless for score events - this would give an indication of performance on the leg.
Dec 10, 2012 4:42 AM
#
We haven't done points per minute, as these can be a bit misleading (say the 100-pointer was close to a 20-pointer but there was nothing else nearby?), but perhaps it's a valid number to include. As of now we do this.
Dec 10, 2012 6:43 AM
#
undy:
Wow - impressive.
I'm thinking about a splits print-out whacked out from a receipt printer that is 3 inches wide, so that would be a bit ambitious, but it's interesting. It would require more course info than our local guys ever create.
I think that where there are a couple of controls together you can figure that out when looking at points/minute (although I considered a rolling average over the last n controls...).
@rockman - for most rogaines the score isn't embedded in the control number, so I think I have to keep that.
I agree that the penalty should probably be at the top (on the example its at the bottom & I truncated it).
Cumulative with penalty is probably a fun idea.
I'm thinking about a splits print-out whacked out from a receipt printer that is 3 inches wide, so that would be a bit ambitious, but it's interesting. It would require more course info than our local guys ever create.
I think that where there are a couple of controls together you can figure that out when looking at points/minute (although I considered a rolling average over the last n controls...).
@rockman - for most rogaines the score isn't embedded in the control number, so I think I have to keep that.
I agree that the penalty should probably be at the top (on the example its at the bottom & I truncated it).
Cumulative with penalty is probably a fun idea.
Dec 10, 2012 9:12 AM
#
Hey Tundra/Desert, how do you calculate the distance and climb for each leg?
And did the top team really only do 40km in 24hrs?
And did the top team really only do 40km in 24hrs?
Dec 10, 2012 3:19 PM
#
Sadly we don't quite have onsite splits capability. All you get onsite is a printout from Ór (if it's working) or SI-Config. All this info is post-event-derived.
Checkpoint coordinates are from Condes's CourseData.xml which is converted to CSV to which I then manually add the elevation of each CP. This CSV is then fed to our script. It doesn't take that long to figure out the elevations from just looking at the contours, perhaps close to an hour for a course with 60 CPs.
The distances are straight-line distances and the climb is elevation deltas. Delivering anything more sophisticated would be beyond our capability.
Yes the top team only did 40 km straight-line, 69 km actual distance... the place is somewhat steep, and they also went for some CPs they ran out of time locating, which isn't reflected in the straight-line official total.
Checkpoint coordinates are from Condes's CourseData.xml which is converted to CSV to which I then manually add the elevation of each CP. This CSV is then fed to our script. It doesn't take that long to figure out the elevations from just looking at the contours, perhaps close to an hour for a course with 60 CPs.
The distances are straight-line distances and the climb is elevation deltas. Delivering anything more sophisticated would be beyond our capability.
Yes the top team only did 40 km straight-line, 69 km actual distance... the place is somewhat steep, and they also went for some CPs they ran out of time locating, which isn't reflected in the straight-line official total.
Dec 10, 2012 4:59 PM
#
One additional possibility is time behind on that leg compared to anyone else who has already downloaded and run the same leg. If space is at a premium, I've seen Poles use underlining... lots of underlining is big mistake, none means fastest on leg.
Dec 11, 2012 12:43 AM
#
The splits printout from Ór for a score course would be my starting point. I can scan or photograph one if you would like to see it. It's a few fields and then four columns (going from memory):
12(34) 4:54 4:54 30
something like:
control? (code) cumulative time split time points for that control.
The fields at the top have something like points earned, penalty/bonus points, and final/total points.
The problem with 4th fastest of 20 teams is that it changes as the people download. I wouldn't worry about that for a printout at download.
If you're writing the software, you might do some sort of average speed and then legs with speeds faster or slower than nominal.
12(34) 4:54 4:54 30
something like:
control? (code) cumulative time split time points for that control.
The fields at the top have something like points earned, penalty/bonus points, and final/total points.
The problem with 4th fastest of 20 teams is that it changes as the people download. I wouldn't worry about that for a printout at download.
If you're writing the software, you might do some sort of average speed and then legs with speeds faster or slower than nominal.
Dec 11, 2012 1:35 AM
#
undy:
Thanks, I have Ór, and the splits printout is pretty good you are right. I'm looking at MEOS, which does a great job with time lost on line courses, but where the score event splits printing could probably be enhanced.
Dec 11, 2012 10:29 AM
#
For local one-man relays where the number of possible forking combinations far outnumber actual competitors I've found that having a split time per actual leg travelled is quite useful.
It seems like the same could be used for Score-O, possibly ignoring direction as another poster suggested.
If you want to be really fancy then you'll print out a gliding (exponential?) average of points gathered per minute. You need to average in order to avoid huge spikes around high-value solitary controls.
It seems like the same could be used for Score-O, possibly ignoring direction as another poster suggested.
If you want to be really fancy then you'll print out a gliding (exponential?) average of points gathered per minute. You need to average in order to avoid huge spikes around high-value solitary controls.
This discussion thread is closed.