Numbers
count
But the lessons of
by Bigg Byrd
Originally published in the September 1989 issue of Cyclops USA.
The trouble with numbers is they are so definite and
quantitative. Most words have shades of meaning that can be pushed around or
bent to suit ones needs. Take, for example, a statement by a politician that
there will be “No new taxes.” Some
people seem to think that statement has a clear meaning. I look forward to
hearing the explanation in 1992.
Unless they are wrapped in a lot of ambiguous words,
numbers have the unfortunate property that they can be checked. In fact, it is
possible in certain circumstances to prove that a number is wrong. You seldom
have that kind of problem with words.
Words can be used to obscure numbers. For example,
USCF (United States Cycling Federation) treasurers were responsible for keeping
the books and paying the bills of the Federation until the Executive Director
took over in 1983, but some of them had not mastered arithmetic and had to rely
on their verbal skills to cover up that fact.
The treasurer at the beginning of this decade seldom
managed to balance his checkbook, but he handled that problem by keeping an enormous cash balance in the checking account and
shouting very loudly at anyone who asked why his financial reports didn’t make
sense. He lost a lot of earnings on the funds that should have been invested
instead of covering his arithmetic deficiencies, but at least he seldom bounced
checks.
Despite the intrinsic dangers of dealing with numbers,
Counting to 10
A USCF press release late in 1988 proudly announced
the results of the National Prestige Classic series and listed both men and
women riders who were in the “top 25” on points. However, careful examination
of these lists revealed that there were 26 names on each. There were ties on
points on both lists, which apparently confused the compiler – for example,
there were two men tied for 6th place, which meant the next person down was in
8th place, but he was shown in 7th, hence the rest of the list was off by 1.
No big deal, perhaps, but accompanying this press
release was another one announcing the 1989 U.S. National Teams, for which the
top 10 finishers in the National Prestige Classic competition were automatic
qualifiers. Sure enough, this team included the top 11 men from the NPC list. This
fact was promptly and discreetly pointed out to the coaching office through
channels, in recognition of the fact that having more than the planned number
of members on the team could cause budgetary problems. Not surprisingly, the
answer came back that no mistake had occurred – that the 11th rider on the list
was a “coach’s selection.”
That was a nice attempt at a cover up, but it didn’t
work. You see, the National Team list had asterisks by the names of each rider
who was an automatic qualifier and, sure enough, Mr. 11 had an asterisk. While
the coaching staff may be a little weak on counting, we hope that they learn to
manage their budgets efficiently so as to get the most possible out of the
National Teams, give or take a rider.
Counting noses
You might think that computers are better at counting than
people. That is true only if they are given proper instructions. It is not a
tough job for a competent programmer, but there is apparently no such person on
the staff of the U.S. Olympic Committee. The USCF database program for licensee
registration, which was written by one of their people in 1979, has been
producing statistical garbage ever since then, and nobody seems to be able to
fix it.
If you look at the tables of numbers of riders by
district and by age group in the back of the USCF Rule Book you might be led to
believe that we accurately know how many licensees there are in the Federation
at any given time, but the fact is that we do not – those numbers have been
faked for years because the computer outputs don’t make sense. For example, the
draft 1989 Rule Book showed that the total number of riders by district last
year was 31,717, while the total by age was 31,769. Why the difference of
52? The Rule Book editor gave up asking
that question long ago – he simply applied some creative arithmetic to the
district numbers to make them add up to 31,769 and let it go at that.
Les Earnest has confessed to having started this
cover-up in 1979. As editor of the Rule Book, he had been including
registration data obtained from Tim Nicholson, who had voluntarily put together
the first USCF computer registration system in 1974. Nicholson’s system had the
nice feature that it produced consistent results. In fact, one of the early
applications of that program was to detect a pattern of voting fraud in
When the first statistics were taken from the IBM
computer at the USOC, Earnest discovered that they were internally
inconsistent. The late Mary Cappy, who ran the
program at that time, said that she would get the USOC programmer to fix it for
the following year, but that somehow didn’t happen, so he fiddled the numbers
again. Earnest repeatedly wrote memos about this problem to various staff members
over the years, with no results. He pointed out that it would be worthwhile to
compare the number of licenses issued with the total licensing fees that had
been deposited, as a check on whether there were any financial “leaks” in that
part of the system.
In 1984 he finally talked Executive Director Dave Prouty into hiring an outside computer consultant to fix
the program. One was hired and paid substantial fees, but he somehow didn’t
manage to make it work correctly. In fact, he unexpectedly disappeared one day,
leaving no forwarding address.
The editor of the 1986 Rule Book tried to cover up the
trouble with the numbers by not showing the number of riders in each district,
but another consistency check was overlooked: the total number of men listed by
category, which is for men 18 through 34 years old, should equal the total
number of men listed in the age groups 18-19, 20-24, 25-30, and 30-34. In fact,
those numbers are rather far apart in the Rule Books for 1986, 1987, and 1988.
They match in the 1989 book because they were “cooked” again by Earnest.
This kind of thing would not be tolerated in a
business, of course, but nonprofits are supposed to be more tolerant when they
are receiving “free” computer services from a donor such as USOC. Undoubtedly
some clubs have been cheated out of votes as a result of this inconsistency –
or perhaps some have received undeserved votes. No doubt some district
representatives have lost riders and consequently lost income; or maybe they
gained – we will probably never know.
The best solution to the problem at this point appears
to be to get off the USOC’s IBM dinosaur, put the
registration data on a PC, and use one of the many database programs that are
available for those machines. In order to do that it will be necessary to hire
someone who knows a bit about computers. A sharp high school student should be
able to do it, but for reliability it will be preferable to hire a “pro.”
Counting postage
The USCF administration’s inability to count postage
has cost the Federation well over $100,000 during the last few years. Rule
Books had been sent to licensees via bulk mail up until 1984 because it was
inexpensive. In ’84, the administrative processing of licenses got severely
screwed up and, in order to partially cover up this problem, Executive Director
Dave Prouty decided to send the Rule Books via first
class mail so they would get there sooner. He apparently didn’t bother to check
into what it cost to do this.
This amazingly bad idea was institutionalized in
succeeding years, causing the Federation to spend in excess of $20,000 per year
more than was necessary. When Earnest got involved in editing the 1989 Rule
Book, he asked why it was not being bulk mailed. The answer from the staff was
that it was “too complicated” to use bulk mail.
After Earnest refuted that claim and pointed out that
large sums of money were going down this rat hole, it was agreed by Executive
Director Jerry Lace that the ’89 Rule Books would be bulk mailed. When Earnest
later inquired whether the back cover of the Rule Book was being printed with
the bulk mailing permit number to facilitate the process, he was told that this
idea had been dropped on advice of the USOC mailroom, which claimed that “the
savings would be insignificant.”
Earnest was certain that this advice was in error and
checked with the Postal Service. He learned that mailing first class would cost
at least $1.05 per book whereas they could be sent as bulk mail for about 9.5
cents each. Multiplying the 95 cent difference by the expected licensing volume
of 34,000 yielded an expected saving over $32,000 for one year. Thus, this
project got turned around once again.
With any luck, bulk mailing will again now be
institutionalized within the USCF administration, which will save rather large
sums each year. But don’t count on it. In fact, don’t count on anyone being
able to count – check the numbers yourself. Then get someone to check you!