Here is how I understand Neville’s proposed system to work:
1) Sufficient experiments are done to establish a predictable relationship between wheel size and tire pressure.
2) The bicycle is calibrated after a note of its internal pressure is made.
3) The course is measured.
4) The tire pressure is checked. If the pressure has risen, this is an indication that the tire is larger than at the beginning, and counts per km has declined. Thus the initial calibration may be used as official.
5) If pressure has declined, the new wheel size may be determined from (1) and adjustments made.
All this seems straightforward and scientifically correct. There are a few concerns that come to mind:
1) How much pressure/calibration information is needed to establish reasonable accuracy, and how is it to be maintained? If I sit around all winter, and go out in spring to measure, what do I use to support my use of pressure information? Last fall’s data? Do I need to re-establish a pressure/size baseline?
2) Neville refers to “pressure monitoring.” How is this done? Each time pressure is checked a little air is lost from the tire. How is this accounted for? Misgiving over this was the reason for my wondering whether the gauge was always connected to the tire during all phases of the measurement. If loss is minimal, error, as Neville says, would be negligible. But a slip of the hand can release more than a minimum.
3) How will varying weight on the front wheel be accommodated? I wear more in winter than in summer, and sometimes carry a backpack. Also, my own weight varies. As a result my wheel circumference, measured over 60 times over 5 years on different calibration courses, immediately after pumping to 125 psi, varied from 261.42 to 262.19 cm per revolution, or from 11060 to 11092 counts per kilometer. My tube has a Schrader valve. I find this pretty good agreement but not quite enough to use as reliable calibration information.
4) The purported purpose of using pressure as an adjusting factor seems principally to avoid a long trip to and from a remote calibration course. Neville seems to have a perception that it is difficult to lay out a calibration course. This used to be true when a half-mile or kilometer was required. With 300 meters or 1000 feet it is almost always possible to find enough straight road near the race course, and it takes less than a half hour to lay down the course, with one helper. I have laid out a calibration course by myself in an hour. Moreover, it need not be done on a cloudy day at 68F. It can be done under any condition. I always use a remote calibration course when a measurement is more than 30 miles from home.
5) How will the information be presented to a certifier in a way that is uniform, credible and understandable? As a certifier I’d prefer not to have to treat each measurement as a special case. A certain degree of uniformity in submissions makes the process easier. Also, the process must be understandable to the certifier. Picture Albert Einstein sending in a measurement complete with arcane mathematical symbols and telling the certifier that it’s logical and correct. After all, the guy’s Einstein – doesn’t this mean he’s right? Maybe, maybe not. Whoever he is, he’s got to convince me in terms I understand.
The whole justification for the proposed method originally rested on Neville’s erroneous premise that it is hard to set up an on-site calibration course. This is not the case. Is there another reason why this method should be considered aside from fun with science? Neville now cites “to improve the accuracy of certification.” Whether such improvement exists remains to be seen. Precal-measure-postcal is simple and direct. Eliminating the postcal in favor of pressure adjusting doesn’t seem to me to lead to accuracy.