Maybe you're thinking that I'm saying ALL our measurements were accurate to within inches. No. Absolutely not. Only the tape.
Is that it?
Maybe you're thinking that I'm saying ALL our measurements were accurate to within inches. No. Absolutely not. Only the tape.
Is that it?
DiscoGary wrote:
Well I don't angry. I get accurate.
This is what we did. All using the SPR method.
Miles : Method
2.9758 : Steel tape - gold standard
2.9717 : Bicycle with a Jones counter calibrated on the grass
2.9727 : Garmin Edge 305 GPS on a bike ~9 mph
2.9808 : Garmin Edge 305 GPS walking ~3 mph
2.9502 : Walking Wheel - expensive uncalibrated, probably similar to the one you used
2.9568 : Walking Wheel - inexpensive Home Depot model uncalibrated
I was wondering if you have calculated the differences in feet between the above measurements? Such as only 16 FEET off the measurement that you feel is most accurate in comparison to using a basic personal use GPS unit while biking at 9mph.
The short course calculation factor is used by USATF because they see the potential error in the measuring process of the bike with the Jones counter. The same reason that there is more potential error going slowly pushing a wheel, which is something you have failed to mention, or have taken into consideration. You seem to have this concept backwards. The slippage you mention would be purely user/process error, and should be discovered while calibrating against the steel tape on the same grass surface. Pushing a wheel while driving a golf cart may create this phenomenon, but any good surveyor looking for accuracy would not use this method and can certainly create adequate pressure to have a quality wheel remain accurate in multiple surface/grass types as one would find on a cross country course. Your USATF Certified course measuring buddy should know that his system is designed for roads and not recommended for soft, undulating surfaces found in most cross country courses because of the inconsistency of ground densities.
I guess what I find interesting is I was able to feel very comfortable with the course length that had been criticized for years by measuring based on the rules that were in place at that time. You mentioned using 12' in a previous post as an acceptable course width, which is far from what is actually the case, yet you somehow feel you have replicated the actual course. When I measured the actual set up State Meet course using the SPR method that same day; I ended up with a distance the is substantially shorter than what you are promoting.
Historian wrote:
DiscoGary wrote:Well I don't angry. I get accurate.
This is what we did. All using the SPR method.
Miles : Method
2.9758 : Steel tape - gold standard
2.9717 : Bicycle with a Jones counter calibrated on the grass
2.9727 : Garmin Edge 305 GPS on a bike ~9 mph
2.9808 : Garmin Edge 305 GPS walking ~3 mph
2.9502 : Walking Wheel - expensive uncalibrated, probably similar to the one you used
2.9568 : Walking Wheel - inexpensive Home Depot model uncalibrated
I was wondering if you have calculated the differences in feet between the above measurements? Such as only 16 FEET off the measurement that you feel is most accurate in comparison to using a basic personal use GPS unit while biking at 9mph.
The short course calculation factor is used by USATF because they see the potential error in the measuring process of the bike with the Jones counter. The same reason that there is more potential error going slowly pushing a wheel, which is something you have failed to mention, or have taken into consideration. You seem to have this concept backwards. The slippage you mention would be purely user/process error, and should be discovered while calibrating against the steel tape on the same grass surface. Pushing a wheel while driving a golf cart may create this phenomenon, but any good surveyor looking for accuracy would not use this method and can certainly create adequate pressure to have a quality wheel remain accurate in multiple surface/grass types as one would find on a cross country course. Your USATF Certified course measuring buddy should know that his system is designed for roads and not recommended for soft, undulating surfaces found in most cross country courses because of the inconsistency of ground densities.
I guess what I find interesting is I was able to feel very comfortable with the course length that had been criticized for years by measuring based on the rules that were in place at that time. You mentioned using 12' in a previous post as an acceptable course width, which is far from what is actually the case, yet you somehow feel you have replicated the actual course. When I measured the actual set up State Meet course using the SPR method that same day; I ended up with a distance the is substantially shorter than what you are promoting.
Yes, the GPS riding a bike at 9MPH was close to the real distance, but walking at 3MPH showed an error in the other direction. The point is that the GPS on an open course like this was a decent measurement, but not as good as the tape. Also, I cranked up my sampling frequency to once per second. If you did not do that then your measurements would be more likely to come up short.
The short course issue has nothing to do with what we're talking about.
Of course our USATF guy knew that the bike on grass would be less accurate than a bike on pavement. That's why he calibrated his measurement on grass, which was good, but it still came up short. Again, I don't know what you're getting at here.
You say that my wheel measurements were user process errors. Again, if they are then they are the same errors that everyone has made for the last 44 years. Did you calibrate your wheel measurements. No. So we didn't either because we were trying to show everyone what a typical wheel measurement would give you. If you think you have some magic way of walking with the wheel to avoid slipping, then you need to go check it yourself. Attempting to calibrate the wheel would have been a closer approximation, but still would have been dependent on the slippage variations due to surface variations.
With respect to the 12' path width. The path width has NOTHING to do with the actual course path. They are completely independent concepts. On any given course you can assume any given path width. We calculated that running down the center of a 12' path width on a 2.9758 mile SPR course with the number of turns on the course would give you a measurement of ~3.000 miles. Do you understand that?
I have no doubt that you measured a much shorter distance if you used a walking wheel, and if your GPS was set to a slow sampling rate, but you know what? You can't remember what your distance was, you don't know what your sampling rate was, and you didn't calibrate your wheel, so you got nothing. You are bringing absolutely nothing to the game other than the same old "I measured it and it came up short" stuff I heard when I first asked the question.
You've asked all the right questions and a whole bunch of irrelevant questions, but you've added nothing.
Your kind of response was exactly why we went there with some discipline and did it right.
The only thing left to do now is to go back and make sure they didn't change anything since 2008. That's the real unknown here, and if you want to point that out as a weakness, then you'd be right.
Let's try this. Here's why we took each measurement:
1. Steel tape SPR - the gold standard. This would be the official distance.
2. Two walking wheels - to show how all the legacy measurements were done, and what the error would likely be.
3. Two GPS passes, slower and faster - to find out how accurate the GPS was and to find out if it depended on speed.
4. The bike with a Jones counter - to see if a bike could be calibrated on grass to get an accurate measurement.
I hope that clears up some questions.
So much to debate with here, but too little time.
The Trimble GPS unit I used creates data points precise enough to allow mapping of the width of a standard city sidewalk while walking along it's edges on any out and back test. What it misses is the undulation of the actual running surface, just like a steel tape may do when stretched at 50-70 Newtons, and just like a measuring wheel will ADD distance based on these slight undulations, which are rarely consistent across the cross width.
On a dewy morning, take your bike and ride as straight as you can across a field, say about 100 yards (like a football field). Do this at your 3 mph speed. Now do the same thing at your 9 mph speed. Lastly, walk the same stretch pushing your wheel. Now observe the tracks in the dew. Which line is the smoothest/straightest? Which line should be the shortest regardless of what a measuring devise says? That is why USATF has a short course calculation factor, and why if done improperly wheeled devices should create short course because of this user error. Slippage while walking is total user error. The 16' differential from your GPS measurement and steel tape measurement is excellent and could easily be the result of user error, especially when piecing together 50m steel tape segments and trying to maintain straight tangents.
Your theory on the 12' path throws your work out the window. You lost big tangents because you are using too narrow of a course in comparison to how the State Meet course is actually set up on meet day. How do YOU not understand that?
Real data of measuring numerous courses over several decades have taught me when I find a course short, most kids run fast. When I find a course that is long, most kids run slower.
You must have failed again to read the part where I mentioned that using the old method of course measurement, when I actually measured it with THAT method, I was shocked on how close to the suggested 3 miles the course was.
Historian wrote:
...
Your theory on the 12' path throws your work out the window. You lost big tangents because you are using too narrow of a course in comparison to how the State Meet course is actually set up on meet day. How do YOU not understand that?
...
You seem like a decent guy, so I will try to find some common ground with you on this.
Let's go one at a time and start with the 12' path width business.
All the measurements I showed you done on the SPR path. Why should everything we did get thrown out the window?
We calculated that to get to 3.000 miles on this course using the middle-of-the-path method, the path would have to be about 12'. I actually measured it with the GPS on the bike (~9mph), trying to stay 6' out on all turns, and that measurement came out to 2.9950 miles, which if you add a little because the SPR 9mph GPS came up short, comes out to almost exactly 3.000 miles.
DiscoGary wrote:
All the measurements I showed you were done on the SPR path.
Damn it. I keep leaving out small words.
DiscoGary wrote:
Historian wrote:...
Your theory on the 12' path throws your work out the window. You lost big tangents because you are using too narrow of a course in comparison to how the State Meet course is actually set up on meet day. How do YOU not understand that?
...
You seem like a decent guy, so I will try to find some common ground with you on this.
Let's go one at a time and start with the 12' path width business.
All the measurements I showed you done on the SPR path. Why should everything we did get thrown out the window?
We calculated that to get to 3.000 miles on this course using the middle-of-the-path method, the path would have to be about 12'. I actually measured it with the GPS on the bike (~9mph), trying to stay 6' out on all turns, and that measurement came out to 2.9950 miles, which if you add a little because the SPR 9mph GPS came up short, comes out to almost exactly 3.000 miles.
So you recreated the course estimating it to be 12' wide to get your 3.00 miles while estimating the difference between your GPS reading and what you thought was accurate in comparison to prior steel tape measurements while measuring six feet out from where you thought the turns would be layed out? Then re-measured getting within 30 cm of where you thought the inside turns would be layed out?
Have you ever actually been to the state meet when the course is actually set up?
Look at this map and envision 12' lanes. Then envision 24' or 30' lanes. The course will measure out differently for each layout based on width.
Historian wrote:
So you recreated the course estimating it to be 12' wide to get your 3.00 miles while estimating the difference between your GPS reading and what you thought was accurate in comparison to prior steel tape measurements while measuring six feet out from where you thought the turns would be layed out? Then re-measured getting within 30 cm of where you thought the inside turns would be layed out?
Have you ever actually been to the state meet when the course is actually set up?
Maybe we've found the source of the disagreement.
We did NOT recreate the course assuming a 12' path. We assumed no path width at all because path width does not come into the equation when measuring the SPR.
We "recreated" the course by simply putting cones where the turn monuments are, and otherwise marking all other elements of the course so that it was clear where the path went.
We then made all those measurements using the SPR method, with 30cm turn clearance.
Then I road around in an attempt to verify our 12' wide path / 6' turn clearance with the GPS to see if it came near 3.000. Which it did.
A couple of the other guys attended the State meet regularly so they knew where the course was. I was at the meet the next fall, and can confirm that we measured the correct course as it was run.
To measure the tangents using the tape, they would have the tangent points marked with flags at each end of the straight. One guy would hold the tape at one tangent while the other guy moved back and forth until he was in direct line of sight from the first tangent through the second flag to the destination tangent. The second guy would then pull the tape tight, and put down a flag. First guy moves to the second flag, repeat. The flags on the 600 meter start to the first turn were dead straight. Their method was flawless.
All other measurements followed that trail of flags, so any lateral movement was minimized. Don't make me do trig to prove that 0.5 meter lateral movement adds close to nothing to a 50 meter stretch. If that kind of "user error" were happening to a huge degree then the wheel measurements would have come up longer than the tape, but they all came up short, so the only way that could happen is if the wheel slipped.
The same is true of undulations. The wheel should have measured longer, but didn't.
Exactly what is your method of measuring courses?
Looking at the map, which I'm sure isn't to scale, but a common mistake in course set up is a straight starting line. Teams on the left would obviously have a shorter path to the first turn.
Historian wrote:
http://www.ihsa.org/documents/ccb/sf_course.pdfLook at this map and envision 12' lanes. Then envision 24' or 30' lanes. The course will measure out differently for each layout based on width.
Not if you measure the SPR.
DiscoGary wrote:
Historian wrote:http://www.ihsa.org/documents/ccb/sf_course.pdfLook at this map and envision 12' lanes. Then envision 24' or 30' lanes. The course will measure out differently for each layout based on width.
Not if you measure the SPR.
You are only correct if ALL inside turns are permanently set and the ENTIRE course is widened/made to width OUT from the SPR. Having measured the actual set up course, looking at thousands of photos from the races over the years like the one below and looking at the times...that is definitely not the case.
http://photos.vipis.com/p839463956/h6ac6ec5a#h67ed15d4Historian wrote:
Looking at the map, which I'm sure isn't to scale, but a common mistake in course set up is a straight starting line. Teams on the left would obviously have a shorter path to the first turn.
I have a map like that which I made from a satellite image, so mine IS to scale, and it does show that the starting line is perpendicular to the line from the center monument to the first turn tangent, unlike what is shown in your diagram.
My map shows every tree on the course, and shows the locations of all the monuments, telephone poles, rocks, sewer covers, and large bugs. I told you you couldn't out-geek us. 8)
When I went to the meet I could see a slight bend to the starting line, with the middle monument directly on the curve. Was it a radius of 600 meters? I couldn't tell, but it's close enough.
Historian wrote:
DiscoGary wrote:Not if you measure the SPR.
You are only correct if ALL inside turns are permanently set and the ENTIRE course is widened/made to width OUT from the SPR. Having measured the actual set up course, looking at thousands of photos from the races over the years like the one below and looking at the times...that is definitely not the case.
http://photos.vipis.com/p839463956/h6ac6ec5a#h67ed15d4
Good. This IS the source of the disagreement.
The SPR is definitely set and determined by monuments and trees which are not moved from year to year. Are you saying the SPR is changing from year to year? Where?
What do you mean when you say the "ENTIRE course is widened/made to width OUT from the SPR"?
Just trying to show a photo with orange fencing that sets the width of the course in stretches and few athletes are running the tangents. I struggle to believe that you and your crew were able to replicate this exact layout and measure within 30cm of these inside turns.
I believe you mean to say that the old course is set with monuments, remember when you measured you were like 42m short of 3 miles using the SPR method. NFHS didn't change the measuring process until last year, so you were years ahead of the game!
If the ENTIRE course ISN"T adjusted properly off of the ENTIRE SPR line, some tangents will be taken away or extra distance will be added by improper placement of flags or fencing.
I'm willing to walk/jog/bicycle the course after next years State Meet and compare measurements and measurement devices. It's a great time to use measuring wheels because the grass/ground is pounded to a smooth firm green mat that minimizes "slippage"!!!
Can you post that map? Do you recall where on the starting line your crew began measuring for your project?
For a guy that isn't accepting of a device that was within 16 feet of the gold standard, "close enough" shouldn't be part of your vocabulary.
When measuring the Shortest Possible Route, the only thing that matters is the inside of the turns, which are marked with post holders set in the ground for the turns which are not defined by trees. There are only two of these.
We were not ahead of our time at all. If the current course was set up according to the old NFHS rules, and if the path width was assumed to be 12', and if their goal was to create a 3.000 mile course, then that's exactly what they did. I happen to know that the Bradley Engineering Dept measured and marked the current course, but I could not find anyone there who actually did the work.
So when IHSA reported that the course was 3.0 miles, they were right according to the rule book. And when everyone else said the course measured short, they were right too.
But the old NFHS rule was ridiculous (and I would like to think I had something to do with that rule change, along with the others who took part in our project) because someone could run the SPR which would be shorter than the official distance, thus violating the rule that a course should never be shorter than the official distance. To make matters worse, the variations depends on the number of turns, so there will be no consistency to the error across multiple courses.
I agree that the if the course changes from year to year, then no one can say what the length is with any certainty. In Illinois, they have put in a lot of effort to make sure that doesn't happen. Our evidence is that the people laying out Detweiller have always been diligent going all the way back to 1971.
We started our measurements from the middle monument on the starting line. BTW, for those who might go to the state meet looking for obelisks at the mile points, the "monuments" are little ~6" wide plaques sitting just at ground level. You have to know where they are, and then you still have to hunt for them.
I am with you on the remeasuring business, especially since the NFHS rule has changed and eight years will have passed since the last measurement and that's enough time for something to have changed. We need to set up another meeting at Detweiller and I hope to see you there.... because measuring with a tape takes two people! Two very motivated people. So make plans to be at the State meet next fall and be prepared to stay until Sunday. There will not be enough daylight after the meet to finish multiple passes on the course with tapes, wheels, and GPS devices.
And you are right, if we measure immediately after they race, then the grass will be beaten down, and wheel slippage will be minimized.
If I get motivated enough, I'll post the map online.
Again, the 12' assumption is not reality and the center of the course measurement would result in this course being extremely close to 3 miles based on the ACTUAL alley widths, not the 12' assumption you keep mentioning. So when measuring the SPR, you end up with a course that is in fact short of 3 miles. Based on the old NFHS measuring method, ALL courses SHOULD be short when measuring the SPR.
Please post the HFHS Rule that mentions the course can not be shorter than the official distance. The rule in fact reads totally the opposite, yet that is also never enforced.
Have the same people been laying out/setting up the State Meet course since 1971? Any growth or death of trees since then? I'm looking forward to seeing the course set up map.
Looking forward to helping out and meeting you next November!
You stat guys crack me up! Like any of this really matters, it's a historic course with an unbelievable atmosphere. Conditions very who really cares if its 40 ft short or long. With our states history of amazing talent this is all comical to me. I won the 81 title in 14:22. Never gave any thought to the time whatsoever. I wanted to win! I had tied Ryan Stoll earlier that year at Bloom Invite. Our state meet finish was the closest at the time in the history of the meet and that's why you see it broke down to the 10th of a second. People often ask me what time did I run. My answer is always the same, fast enough to win. I wasn't going to pull Ryan around for 2 3/4 miles and lose in 14:05. The next year Ryan ran 1:47 at Stanford and I ran 14:10 for 5000m on the track. Cheers fella's!
Jim Knudsen
1981 IHSA State XC Champion