HOW DID THE FINISH TIMES FOR THE 2017 MARATHON OF THE TREASURE COAST GET ADJUSTED TO PUBLISH FINISH RESULTS?

 

Here is the process used to adjust finish times for the 2017 Marathon of the Treasure Coast.  It was NOT a ‘computer program’ adjustment, and it was not a fixed time added to all who were affected by the error.

To recap:  During the first three miles of the 2017 Marathon of the Treasure Coast full marathon, half marathon and team relay marathon races held Sunday morning, March 5, 2017 in downtown Stuart, Florida, 159 runners were inadvertently directed off-course between the 1-mile and 2-mile marks, resulting in those runners covering less than the complete full marathon or half marathon race distance.

At that point, several complicating things occurred – first, the vast majority of the runners (those not among the first 159 runners) continued to run the correct course (and did not take the short cut those first 159 runners did), thus running the correct distance that their race comprised.  Second, some runners simply finished the race by following the course markings all the way to the finish, thus running 0.81 miles short of their correct race distance.  Third, some runners chose (either on their own or by encouragement from other runners or race officials) to ‘add’ distance during the race, by using their GPS-based watch (Garmin or something similar), so that when the runner finished, their time and distance approximated the complete distance of their race.  Fourth, some runners crossed the finish line, saw that their distance and time were short, and chose to continue running in and around the race site until their GPS watch showed the ‘proper’ race distance, and then this runner reported to the timing area and displayed this ‘complete’ finish time by showing timing personnel his or her watch.

PROBLEM:  How to make any sense of the finish times, finish order, and how to assign places for results and award purposes.  NOTE:  NO possible ‘solution’ or ‘correction’ will treat all runners ‘fairly’, or in the exact same way.  There is NO possible system that will completely and fully re-create the race as it would have unfolded had the course mis-direction error not occurred.  Some runners realized, during the race, that their race distance was incorrect, thus affecting that runner’s desire and ability to continue to run in the same manner and at the same pace as they might have had the error not occurred.  Some runners were unfazed, and continued to run as hard as they would have, not matter what had occurred.  A few runners even dropped out of the race once they realized the course error would adversely affect their finish time and place.  So NOTHING will completely and fully correct for those factors, and reconstruct the race as it should have been run.

THE SOLUTION USED:  The 2017 race course started with runners facing east, and the first 1.55 miles of the course were a loop through the Stuart neighborhoods that led runners back through and underneath the start/finish arch – and thus across the electronic timing mats – before the course proceeded into the downtown Stuart area, where the course error occurred.  Full marathon and team relay runners ran a course that crossed three more electronic timing points – at 7.34 miles, at 13.1 miles and at 17.7 miles – before coming back into the finish line area.

Ideally, each full and team relay marathoner would get split times for those three timing points, and those split times have been published in the online results.  Because the distance between each split point is accurately known (it was noted during the course measurement and certification process done before the race) and because the time between split points is accurately know, then a pace per mile for each segment of the race can be calculated.

Once all runners passed underneath the starting line arch and proceeded out onto the race course, the electronic timing mats were turned off.  This segregates all the runner starting times from any other times generated by this set of timing mats.  Although runners would soon pass over those same timing mats at 1.55 miles, any timing data generated by this passing of the timing mats is not used anywhere in results calculations, so the original plan was to keep the timing mats turned off while runners passed underneath the finish arch at 1.55 miles.  Timing personnel decided, on the spur of the moment, that more data is always better than less data, so the timing mats were turned back on before runners returned through the arch.

Again, once all runners had passed by the arch at 1.55 miles, the timing mats were turned off to close out that set of timing data, and then later they were turned back on in order to capture the runner finish times as runners came back and finished the race.

Originally, each full marathon runners or team relay team would have five timing data points – a start time, a 7.34-mile time, a 13.1-mile time, a 17.7-mile time, and a finish time.  Now, one more possible time was available – that of the 1.55-mile split point.  It is this extra timing point that provides the basis for the finish time adjustments.

The missing distance that some runners did not run due to the course mis-direction error was measured and determined to be 0.81 miles, meaning someone who was mis-directed ran 0.81 miles less than the complete 13.1-mile half marathon or 26.2-mile full marathon distance (technically, the half marathon distance is 21,097.5 meters and the full marathon distance is 42,195 meters, and both courses were measured to these distances).

The electronic timing data were used to generate split times for 1.55 miles, 7.34 miles (5.79-mile segment), and 13.1 miles (5.76-mile segment), and the pace per mile for each of these three segments was calculated, and those split times showed clearly which runners were mis-directed and which runners were not.  Here’s a visual of the split time data for the lead full marathon runners:

Photo

Notice that the split times for the middle segment are MUCH faster than either the first or last segment.  This means that those runners did not run the full 5.79-mile second segment, but instead they ran a segment of 4.98 miles.  If you do the math using that shortened split time for a 4.98-mile distance instead of 5.79 miles, then those split times fall very close to the first and third split segment split times.

Now notice what happens when we reach the 50th runner of the full marathon:

Photo

From that 50th place downward, all the segment split times are much closer to each other than those of the first 50 runners in the report.  This is how the determination was made as to who did and did not run the complete specified race course.

Runners were also queried when possible, and asked if he or she did or did not run the downtown Stuart boardwalk segment of the course – running the complete course as specified would have put one on the boardwalk, while those who were mis-directed would have missed the boardwalk segment.

A similar process was used for the half marathon times, and that process indicated that 108 runners were mis-directed off course.

Finally, the location in the ‘pack’ that crossed underneath the arch at 1.55 miles of the last ‘short-course’ full marathon runner and the last ‘short-course’ half marathon runner were identified, and it turns out that those two runners were very close together when each of them ran underneath the arch at 1.55 miles.

Having identified the runners who were affected by the course error was Step 1.  Step 2 was how to adjust the finish times for these affected runners.

Most runners run at a fairly even pace in the early miles of any race, and (for most – not all – runners) that pace only falls off (or speeds up) in the late miles of the race.  Having a 1.55-mile split time allows a pace per mile for those first 1.55 miles to be calculated – and that pace is shown in the photos above.  An assumption was made that all runners would continue to run at that same pace for at least the next mile or two, so all of those runners who were mis-directed can now have a time calculated for that runner’s missing 0.81-mile segment.  Each affected runner’s 1.55-mile pace per mile was multiplied by 0.81 miles, which gave a time in minutes and seconds, of what that runner would have run had that person run the full course segment.

This means if you were one of the affected runners, and if you ran a 7:00 per mile pace for the first 1.55 miles, then you would have run 7:00 x 0.81 = 5.6 minutes, or 5 minutes 36 seconds for the missing 0.81-mile segment.  Each runner ran a slightly different pace for the first 1.55 miles, so each runner’s amount of time added was different.  An 8:00 mile pace would have generated 6.48 minutes added, or 6:29 added.

It must be emphasized that this process does NOT treat every runner fairly – but NO process, including making no time adjustment, would do so.  This appears to be the most equitable way to make time adjustments that would ‘re-create’ the race as it would have been run, in such a way as to be able to better assign finish places and awards.  We remain open to any and all suggestions that would be ‘fairer’, but we don;t know of anything that would please everyone involved in the races.  This occurrence leaves everyone walking away unhappy, but possibly these adjusted times can make some runners less unhappy.

NOTE:  This process does NOT take into account any extra distance that runners may have added on their own – time was added to each runner who was identified as running the ‘short course’ using the process described above.  If a runner has any credible documentation showing extra distance run (a GPS track, or photo of watch showing time and distance run, etc) that evidence will be considered and may be used to further adjust that runner’s finish time (by using that credible finish time, or by using that data to adjust further the runner’s time in some way).  Evidence must be submitted to Mike Melton at mike@mcmelton.com, and will be considered on a case-by-case basis.