I will supply the minimum information necessary, and include as much obfuscation as possible.

A couple weeks ago I measured a course with a new measurer to teach him the tricks of the trade. Early in the morning I gathered up my materials, hung my bike up on the rack on the back of my vehicle, and drove 80 miles to a location near his home. I met up with him there, and we measured out a 315-meter calibration course at that location. Just as we were finishing it started to rain, so we went back to our separate cars where we waited it out. About 20 minutes later the rain stopped so we got out and did our pre-cal rides. Then we gathered up our stuff, hung our bikes on our racks, and each drove the 40 miles back towards my house where the race course was located.

At the race course location we found the desired start/finish, and then rode from there together to measure the 5k course. The course had a turn-around that we would be adjusting, so we stopped at the google-earth-determined point, spun our bikes around, and continued with the measurement until we got back to the start/finish. The course ended up being a few meters long, so we drove back to the turn-around and mile marks to adjust with our bikes. After we were all done we put our bikes back on our racks, and he drove back to the original cal course to do his post cal, while I drove 40 miles home to do my post-cal on my home cal course.

After I finished my post cal I pulled out my laptop in order to enter all my data into a spreadsheet, and that's when I noticed something strange. As I was entering my data I noticed that my pre-cal started with 262000 and ended with 280747, with the rides averaging out to about 3750. The starting count of my course measurement was 222000 and it finished at 281660. The start count of my post-cal was 264000 and the finish was 279557, with an average of about 3560.

I stared at these numbers for a long time trying to figure out how they were possible. I tried all kinds of combinations of how I might have transposed digits when I wrote the starting counts down for one, or all, of the measurements, so that the start counts of the second and third measurements were not smaller than the finish counts of the previous measurements. I couldn't come up with anything. And then suddenly it hit me, and all was right again in my measuring world.

What had happened with my numbers? If you know the answer, don't post it here. Write it on the underside of the frame of a Faraday Porteur electric assist bicycle with 8-speed internal hub shift, Gates belt drive, and steam-bent bamboo fenders, and send it to 323 Griggs Street, Rochester(our fair city), MI.

Seriously, if you do know the answer please don't post it here. Let others puzzle over it for a while like I did. Email me or private message me with the answer, and then after a couple days I'll post the answer here. You can post questions or requests for clarifications here though. Or maybe the answer is obvious to everyone. It wasn't for me.
Original Post

I've gotten a couple questions about the length of my home calibration course. It is 300 meters.
But I'm not saying if this information is relevant to the puzzler answer or not.
Mike Sandford and Justin Kuo came up with the correct answer. Good job. After I did my pre-cal I put my bike on my rack and drove 40 miles to the race course. During that ride the front wheel started spinning backward. After I finished the course measurement I again put my bike on my rack and drove 40 miles to my home to do my post cal. And again, the front wheel of my bike was spinning backwards during this drive. That's how the starting count of my course measurement and post cal ended being smaller than the finish count of my previous measurements.

As far as I know this hasn't happened to me before, but it's possible that it has and I just didn't notice. Especially if the wheel was spinning forward there's a good chance I wouldn't have noticed the count numbers being odd. Has it happened to anyone else?
I think it likely that the wheel only spins backwards and not forwards, due to the asymmetric airflow.
It seems plausible that there is a greater down draft nearer the centre line of your car, due to the air curling down after passing over your car roof.
The half of the front wheel that is furtherest from the centre line, may have less down draft as the air curls in from the sides.
The net force would thus be a down force on the back half of the front wheel causing it to rotate backwards.
Interestingly if you turned the bike on the rack so the front wheel was on the other side of the car, the asymmetric flow still gives backwards rotation!
I think you would have to mount the bike upside down to get forward rotation - Ah... the endless possibilities for experimentation to test theory.....
I didn’t submit an answer, and here’s why:

1) I noticed right away the count increase during auto transport, but paid no attention to it as it did not concern the actual measurement. I have observed this quite a few times. On the interstate I see lots of folks with 2 to 4 bikes strapped on the back of their SUV. Often I'll see a bike wheel turning in the slipstream. Not at all unusual, so why should a measurer be affected? It has no effect on the measurement and can be safely ignored.

2) It looked like 5 precal rides were taken, as the rides averaged about 3750 as stated. Only 5 rides would account for this. No actual problem so far. Race course measurement seems OK so far.

3) On the postcal I was unable to get reasonable agreement between precal and postcal constant. I tried 300 m as a cal course length, and I tried various numbers of rides, but could not reconcile the numbers.

If the puzzle was about the count increase only, the answers are correct. However, the calibration data as stated looks wonky to me.

Can anyone help clear my head of this confusion?
Last edited by peteriegel
Pete,

The postcal numbers were a puzzler to me. So I asked. Mark explained it this way:

quote:
When I calibrate I always roll the counter forward to the next 1000. So the starting counts for my 4 rides were 264000, 268000, 272000, and 276000. I write down the first final number, say 267559, and then for the next 3 I just have to remember the last two digits. I can tell immediately if there is any kind of problem with my numbers, and it is easier to enter into my spreadsheet since the starting counts on my sheet for the second, third, and fourth rides are formulas that add 4000 to the previous starting number.

Thank you. -- Justin