Over 5000 Disqualified From The Mexico City Marathon

An unprecedented number of runners have been disqualified from The Mexico City Marathon. Did they all cheat, or were some victims of timing errors?

After I posted an article regarding the accusations that 1000s cheated at the Mexico City Marathon, the race released the official results. According to the race website 5,806 runners were disqualified.

I did not pull the preliminary results immediately following the race, so my #s differ slightly from the #s the race published.

From the Mexico City Marathon website:

Total athletes who crossed the finish line 28,206

Total athletes who completed the full course (Finishers) 23,887

Disqualified 5,806

The # disqualified does not equal the difference between total athletes that crossed the finish line and those listed as finishers.

The data I had pulled show the following:

5,553 runners were removed from the results

1,570 runners were added to the results

Missed Splits – Full Field

You can see from this data that they disqualified practically every runner that missed 3 or more mats.

I did a quick analysis of the remaining runners – I checked for large negative splits, and pace variances following missing splits, and it appears that there are still well over 100 runners with questionable results.

There is more analysis to do in order to analyze for timing mat issues (more on that later).

Missed Splits – Boston Qualifiers

After the disqualifications, the # of Boston qualifiers went from 1296 to 495. There are still a handful of Boston qualifiers whose results are questionable.

I am also looking closely at the results to try to find an explanation for the 21.7 mile mat data. It doesn’t appear that any of the runners that only missed that mat gained any advantage over the portion of the course where they missed the split.

Some Official Results are Still Questionable

As shown in the data above, The Mexico City Marathon basically removed runners that missed 3 or more splits (with a few exceptions). Also, it is unclear on what happened with the runners that were added to the results. I did see some comments on the initial article that a runner reported that their finish time was not recorded initially. This may be a correction to an issue at the finish line.

Some of these results could potentially be justified. Most of the truly suspicious results come from runners that missed that didn’t register a time at the 10k timing mat (6.2 mi). Also, it is interesting that some of these runners have identical gun and chip times. In order for those times to be accurate, they would have had to start next to the elites. But, the runners that are showing 2 minute paces cannot have those times explained by faulty timing mats. Regardless of whether they intentionally cut this portion of the course, these official splits cannot be accurate.

In the original race results, over 4000 runners had identical chip and gun times. All but 500 of those runners were disqualified. It is physically impossible that 4000+ runners were all crossed the timing mat at the same instant. Typically this would indicate that no start time was registered for those participants.

Also, ‘Maria’ is still in the official results results.

Timing Issues or Cheating?

I still have a difficult time with accepting that 5800 runners cut the course (deliberately or otherwise). I, along with others whose help I have solicited, are continuing to analyze the data. We are looking deeper into the mats that were missed to see if there is a pattern.

It is possible that the timing system was not sufficient for the # of runners that were crossing the mats. If you have a non-redundant system, and more runners cross the mat than the system is equipped to handle, you can have runners whose data is not collected. For this, and similar circumstances, I am not necessarily concerned when a runner misses 1 or 2 mats, if they don’t show any pace improvement over that stretch of the course. However when a single runner misses multiple mats – it becomes more unlikely that they just happened to be victimized by an overwhelmed system every time.

The initial pace analysis showed a significant amount of results that could not be explained by timing error. More analysis is needed to try to determine how many of the 5800 runners actually cut the course (intentionally or unintentionally) and how many (if any) were victimized by a potentially overwhelmed timing system. By disqualifying the runners from the official results, the race is in essence saying that those runners cut the course.

To put this in perspective, I wrote about widespread cheating at Honolulu and Disney. I estimated those races had 400 and 200 course cutters respectively.

More updates will come as the data is analyzed further.

Please consider a contribution to help support the site. Contributions help to offset costs associated with running the site and help to enable me to compensate those that assist in data collection, etc.

Thanks to all of you that that support Marathon Investigation!