# Calculated Traffic Counts in July 08

We have routinely maintained two counters at Point 1--the higher-traffic location below the campus gate--so that if one counter hiccups we have the other to back it up. For the period from 7/7/8 to 7/22/8, unrelated obstacles prevented a complete count.

During that period, a staff member recently trained on counter maintenance inadvertently switched the tubes on one of the machines, rendering its count meaningless. Coincindentally, and lamentably, the data captured on the other machine for the same period was overwritten by a different staff member in the process of offloading it to the computer.

Our solution was to project the daily totals as follows:

• Analyze the summer pattern for all known days
• Calculate the daily number of conference guests for the entire summer (even though many of them arrive in buses or shared automobiles)
• Add double the daily sport camper count (to deduce the number of trips for drop-off/pick-up, even though some are known to carpool)
• Plot the results on a chart in order to observe the traffic patterns
• Select as a baseline the highest-count week during which we had no camps or conferences (namely, August 11th to 17th)
• For each day of the missing weeks, add the number of visitors known to be present on campus to the corresponding day from the highest count week

The result is that for the two weeks affected, although the average daily visitor count was only 87 higher during that period, we have projected a traffic count averaging 383 daily trips higher--an ample "benefit of the doubt" margin. The chart below depicts the entire summer season. The blue bars represent the known values and the gray bars are the projected values.