View Single Post
  #2  
Old June 17th 07, 06:34 PM posted to rec.bicycles.racing
RicodJour
external usenet poster
 
Posts: 3,142
Default The Kunich Korrection: A required skill for r.b.r. navigation

On Jun 17, 12:35 pm, ronaldo_jeremiah
wrote:
A recent post by Tom Kunich (TK) clearly illustrates the need for
corrected estimates of post content by author. Specifically, TK
claimed to have been able to use a three-stage process to spin up a
53x15 gear to a cadence of 180 rpm (it is assumed that wheel size was
700c, though this is not specified). This equates to about 50 mph,
superior to the finest slightly-downhill sprint that Mario Cipollini
could produce in his heyday. For TK, this is par for the course (the
boast, not the actual achievement of claimed performance). After
illustrating the impossibility of the claim, r.b.r. veteran Carl
Sundquist showed fine skills, honed by voluminous TK experience, to
deduce that this realistically reduces to an actual performance of 120
rpm on the same gear, equivalent to a speed of 32 mph that one could
expect of a Cat 5 performance. (The alternate explanation, that TK
was using 469c wheels, seems much less likely, but cannot be entirely
ruled out). Though Carl did not label it as such, he has masterfully
applied the Kunich Korrection (KK). In this case:

KK = 120/180 = 67%.

This is a good starting value for quantitative TK posts, though it may
represent an upper bound. When given results are still implausible
using KK = 67%, KK's equivalent to 50%, 33%, or even 8% may be
appropriate. Further refinement of quantitative KK is necessary.
Thankfully, a vast backlog of data does exist.

The principles of KK can be extended to qualitative posts, though
judgment and experience are required. For example a TK post claiming

"I was a sharpshooter"

should be conservatively KK'ed to

"I have shot guns a few times"

or even liberally KK'ed to

"I have been in the same room as a gun, once."

Discussion:
Kunich Korrection (KK) is a powerful tool for extracting sense from
the hyperbole of TK posts. It assumes that there is always kernel of
truth in a TK post that can be estimated by careful discounting. KK =
67% is suggested as a good starting point for most quantitative posts,
though lower values may demonstrate their superiority as the technique
is refined. The principles of KK have been used by r.b.r. veterans
for years. The elucidation and naming of the technique, along with
the provision of worked examples, is hoped to benefit r.b.r.
journeymen and provide a shorthand that will streamline posts
exhibiting TK-associated deviations from rationality.


A yeoman-like workup, but without graphs, plotted data, footnotes and
a comprehensive bibliography you'll never get it published. Don't
worry about the bibliography - no one actually reads all of the books
they list and no one will check your sources.

R

Ads
 

Home - Home - Home - Home - Home