Thousands Accused of Cheating at The Mexico City Marathon


Shortly after The 2017 Mexico City Marathon, I received requests to review the results. A Facebook page popped up outing many alleged cheaters. I read comments that there were 1000’s of cheaters – runners that cut the course. I was skeptical at first.  However, when I looked at the results, and the paces, the data seems to support that the unprecedented amount of missed mats were widely attributable to runners cutting the course.

There were 8 intermediate timing mats on the course, roughly every 5k.

The Results

Total Runners                 27,873 100%
Missed 0 mats                 17,827 64.0%
Missed 1 mat                   3,502 12.6%
Missd 2 mats                   1,435 5.1%
Missed 3 mats                   1,381 5.0%
Missed 4 mats                   1,006 3.6%
Missed 5 mats                       766 2.7%
Missed 6 mats                       946 3.4%
Missed 7 mats                       719 2.6%
Missed 8 mats                       291 1.0%

What does this data show?

It shows that 36% of all runners missed at least 1 mat.

23.5% of all runners missed multiple mats.

Next I analysed which timing mats were missed – possibly there was an issue with a couple of timing mats that would explain this data.

Total Runners % of Field
missed  3.1 4978 18%
missed 6.2 4418 16%
missed 9.3 5659 20%
missed 13.1 4018 14%
missed 15.5 4323 16%
missed 18.6 2722 10%
missed 21.7 4179 15%
missed 24.8 1109 4%

This distribution is interesting. The highest # of runners missed the early mats. The final mat – just 1.4 miles short of the finish line was only missed by 4% of the runners. All other mats were missed my a minimum of 10% of all runners.

Course map

This Facebook page lists many examples of alleged course cutters. It shows pictures of runners taking taking the subway towards the finish.

Courtesy: CazaTramposos Maratón Cdmx 2017 – photo from 2016


They edited the runner’s photos. I have been unable to obtain the original.

The Facebook page has hundreds of photos. They make a strong case that each of the runners listed cut the course.

Boston Qualifiers

I shifted my attention to The Boston Qualifiers:

Total Runners 1296 100%
Missed 0 mats                       397 30.6%
Missed 1 mat                         89 6.9%
Missd 2 mats                         58 4.5%
Missed 3 mats                       171 13.2%
Missed 4 mats                       147 11.3%
Missed 5 mats                       132 10.2%
Missed 6 mats                       167 12.9%
Missed 7 mats                       106 8.2%
Missed 8 mats                         29 2.2%

69.4% (899 runners) of Boston Qualifiers missed at least 1 mat.
62.5% (810 runners) of Boston Qualifiers missed at least 2 mats.
58% (752 runners) of Boston Qualifiers missed 3 or more mats.
44.8%( 581 runners) of Boston Qualifiers missed 4 or more mats.
33.5% (434 runners) of Boston Qualifiers missed 5 or more mats
23.3% (302 runners) of Boston Qualifiers missed 6 or more mats.
10.4% (135 runners) of Boston Qualifiers missed 7 or 8 mats.
2.2% (29 runners) of Boston Qualifiers missed all 8 mats.

Which mats were missed by Boston Qualifiers?


missed  3.1 658 51%
missed 6.2 627 48%
missed 9.3 761 59%
missed 13.1 623 48%
missed 15.5 574 44%
missed 18.6 351 27%
missed 21.7 292 23%
missed 24.8 56 4%

This shows an even more distinct pattern than that of the entire field. If the data picked up by the mats is accurate,  many runners crossed the starting line, jumped off the course and jumped back in towards the end. The paces calculated by the timing splits seems to confirm this. For the most of the runners with multiple missed splits, the data shows unreasonable fast splits where data is missing.

A clear example of this is seen when looking at the split between 24.8 and 26.2. Most runners have a 24.8 mile split time. Comparing this split time with that for the full marathon, there are clear examples where something is wrong:


The section highlighted in yellow shows total mats registered (out of 8), the total pace for the full marathon (in minutes/mile) and the pace from 24.8 to 26.2. This list is sorted by the variance between the last split an overall pace. For example: the first runner shown finished the race in 3:09:45, but took 45:41 to traverse the last 1.4 miles. That same runner finished the 2016 Mexico City Marathon in over 6 hours.

Course cutting is not the only issue at Mexico City. A number of potential bib mules have been identified. In addition to the examples that have been posted on the Facebook group, I have identified others by looking for identical splits and checking the finish photos. Below is a clear example: It was emailed to me and also appears on the CazaTramposos Maratón Cdmx 2017 Facebook page:


You can see from this photo that the runner is wearing two bibs. The top one is registered to “Maria”. Maria is now qualified to run the 2018 Boston Marathon.

The 2018 Boston Marathon

Registration opened yesterday for the 2018 Boston Marathon. As I previously reported, the B.A.A. added this to the letter sent to all Boston Qualifying races:

If, after submitting results to the Boston Athletic Association, your event makes adjustments to results (including but not limited to disqualifying a runner or discovering a fraudulent performance), please contact the B.A.A. accordingly. The Boston Athletic Association relies upon the good faith of events to provide authentic, accurate results, and expects events to uphold standards of fair athletic competition. 

We also ask for your assistance should there be a question about results from your event. This may include an active review of the results, splits, photos or videos to determine the accuracy of a result(s) associated with a particular individual(s). 

I am conflicted in regards to what action I would like to see from the B.A.A. here. On one hand, the apparent cheating is so widespread, that I’d  advocate that the B.A.A. revoke the qualifying status of The 2017 Mexico City Marathon. On the other hand, despite the large amount of highly questionable results, there are 100’s of runners that likely legitimately ran the full 26.2 miles under their qualifying standard.

At the very least, I expect that every single registrant that uses the 2017 Mexico City Marathon be scrutinized by the B.A.A.

Once the 2018 Boston registrants are publicly available, I will individually scrutinize runners that I am able to determine used Mexico City to qualify. I will report what I find to the B.A.A. I will reach out to the B.A.A. in an effort to work cooperatively with them to scrutinize the results. There are potentially hundreds of runners that will attempt to use Mexico City to run the 2018 Boston Marathon that did not legitimately complete the course. That means that if these runners are allowed to run Boston in 2018, hundreds of deserving runners will be left out.


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!


One Time Contribution



  1. It is unfortunate the large-scale trap that is made in the marathon of my city; I have run it 11 times and I have never managed to qualify Boston even though my best time at the time has been 3:08:11; I did not even qualify when my superiors would pay me the whole trip if I got the goal; I know cases of individuals and clubs that use mules, cheat and / or have gone to Boston with numbers that do not belong to them, there are sale of them in clubs; I would not want to affect the ranking or tag of my marathon, in itself so complicated is to run it, but it is the only one in Mexico that is qualifying to Boston, or pay to run an international, not be able to enjoy this exit as a vacation and also have some misfortune that would not allow me to classify would be costly and not viable for me. I’m embarrassed by all of the above.

  2. “Once the 2018 Boston registrants are publicly available, I will individually scrutinize runners that I am able to determine used Mexico City to qualify. I will report what I find to the B.A.A. I will reach out to the B.A.A. in an effort to work cooperatively with them to scrutinize the results. There are potentially hundreds of runners that will attempt to use Mexico City to run the 2018 Boston Marathon that did not legitimately complete the course. That means that if these runners are allowed to run Boston in 2018, hundreds of deserving runners will be left out.”

    Except if they are initially accepted, it won’t matter if they are later disqualified – hundreds of deserving runners will already be left out, and I doubt the BAA will make an announcement in November that the cutoff has moved down 10 seconds.

    This really sucks.

    I hope the BAA takes notice of this and examines the data before admitting anyone from this marathon.

    • No joke. I’m registering on Friday (my time is well safe from whatever the cutoff ends up being), but I’m sure I could know one or two people who just miss the cutoff of a lot of these results stand.

    • as I wrote before; in other editions of Boston I have seen cases of trap to enter, such as: use of mules; reliefs; purchase of numbers; roadblocks. I have never been able to qualify for 3 minutes and it is not fair for someone without the minimum preparation and sacrifice to go for pure marketing and ego only because they have already bored the regular sessions and changed them for sports.

    • A friend of mine did not make the cut off in a previous year. Months later, she received an email stating there was room and she was in. She pressed for more information and was told they caught cheaters. With that, they might be initially accepted, but if removed, the legitimate BQ’ers who missed the cut off would be accepted. Not all is lost.

  3. I think it’s going to come down to, does the BAA have the means to scrutinize all of the runners using Mexico City as their qualifier? That’s potentially a LOT of data to be sifting through.

    BAA wouldn’t have to do much at all in the ideal world, as Mexico City should automatically disqualify anyone missing two/three or more timing mats. But given the short turnaround time for registration and review, I don’t see that happening. Ultimately (and sad to say for those who qualified legitimately at the race), I think a full revocation of Mexico City as a qualifier is the most expedient route.

    • This is fair and comprises directly the organizers; however in the statistics all that appear on the exit with 7:01 a.m. is because they did not appear to it and the chip by default gives them that timer. It is very easy to know who lied (with the exception of those who use mules or do relays, they require a more tedious investigation) with this and adding the illogical rhythms per mile; it should be the opposite, if I ask for a place in Boston (as I said I do not qualify because they ask me 3:05 in the official and I always go in second group with a delay of 1 – 5 minutes against the chip and my best mark is up of 3:08) and a failure in a brand is not reflected in my timer, I am the one that has to appeal to check that my times and rhythms are not trucados, for it has worked in a masterful way to the marathon APP to the mobiles. However as I stated before the label and ranking as well as certifications of my city should not receive such a severe blow because a horde of pseudo-marathoners = marketing, they only come for a self and they themselves evidenced presuming their achievement in social networks and when check the app you realize the truth (includes several foreigners). regrettable.

      Google Traductor para empresas:Translator ToolkitTraductor de sitios web

      Acerca de Google TraductorComunidadCelulares
      Acerca de GooglePrivacidad y condicionesAyudaEnviar comentarios

    • Maybe entrants into Boston should be required to have the equivalent of the bio passport, but based on past performances. How many people really run a BQ in their first marathon? Or first race for that matter. Maybe races need to partner with Athlinks and flag anomalous race results. Then as bogus entries are found, the next runner on the list has the option to pay a little extra to cover the costs of vetting suspect entries or the spot goes to the next runner on the list. It’s silly that all the data exists, but no races are interested in doing anything with it.

  4. “At the very least, I expect that every single registrant that uses the 2017 Mexico City Marathon be scrutinized by the B.A.A.”

    Have you alerted the BAA to this post?

    I’d be very interested to hear their comments on this.

  5. Are you telling me NO ONE saw this guy wearing two bibs and said something to a volunteer, race personnel, etc.?!?
    Already know the answer…no one cares. This is infuriating!

    When I volunteered at the Dallas Marathon for the finish line, we were required to verify they had a bib on, before handing them a medal.

  6. One element for qualifying to Boston is the combination of time+age+gender and in this analysis, only time was the data for the assumption that X runner has the register. But, no matter that information, today MCDMX 2017 is still unofficial results, so nobody can use this register to be accepted in BAA. Second, is known for everybody, that BAA will review the data you send and they will certified that the data is a valid register, so, don’t be panic so early, let them make what they are really prepared to do.

    • In the past, Boston relied solely on the qualifying marathon to validate their results. Unofficial results or not, this race is a qualifier for 2018 Boston.

      Bringing attention to this will hopefully help encourage Mexico City to clean up the results as well as alert Boston to request the extra validation.

      Boston does not individually validate all runners.

  7. WOW!! Looking at the FB page and crazy to see the pace times! One lady at the 40K had 3.32 pace after doing the 5K at a 8:35 pace!

    It would cost money for Boston to either hire Derek-his wet dream, or do verify each runner’s BQ. As with any race, they aren’t going to spend more money! It’s all about the money for races!!

    • this marathon runs at a height of more than 6900 feet of altitude also has differences of 328 feet especially in the last 8 miles that end in ascent; not for nothing the golden elite runners finish it with more than 10 minutes on the record; ie if you can start at a good pace and end very badly but it is obvious that the exposed cheaters did not pass through the checkpoints and have elite and unrealistic partial assumptions.

  8. Derek, how do you obtain the race results in a table format with intermediate times? Do you contact the race organizers or is there a way to scrape the data from the web? Or is this a trade secret that you don’t want to divulge?

  9. So proud to be a member of the Team Runners Plus MX. We were more than 100 runners, most all of us have finished the Marathon. Some of our members obtaining Boston goal.

  10. It’s not only a matter of a specific country. This millennial culture of pretending destroys everything. People cheating just to show in social media fake accomplishments, fake trophies, cars they do not own, fake vacations and fake happy lives. This is the culture of pretending in social media.

    • All those damn Millenials. Like Robert Young, Rosie Ruiz, Paul Ryan, Mike Rossi, Kip Litton, Julie Miller.
      Except for JS, these fakers are all baby-boomers.

    • Yeah, older generations had it easy, they could lie and people had to take it at face value. Millennial culture and social media culture are not mutually exclusive. If you’re saying “there is a rise in cheating due to social media culture”, I think you can make an argument. “Millennial culture of pretending destroys everything” is an invalid statement.

  11. I ran this full marathon, but there is a problem with the tracking system they used. It shows that I missed 2 Matts and the clock didnt stop when I crossed the finish line. And it shows that I didn’t finish…how can they fix this . And how many runners had the same problem?

  12. I´m from Mexico and this is a no brainer. The BAA should banned Mexico City 2017 Marathon as a qualifier for Boston 2018.

  13. Great work, as always.

    Just a note about the math, though. You mention that the first entry in the spreadsheet finished the last 1.4 miles in 45:41. They actually completed it in over an hour at a 45:41/mile pace.

    Just want it to be accurate.

    Good job outing the cheaters, Derek.

  14. It’s really funny, on that FB page they published a “finisher”‘s Facebook boast post, and next to it are her splits. See, she limped through the first 20k at a 8-10 minute per kilometer (14-16 per mile) jog/walk, then completed an elite-level 10k that took her half an hour, then continued the jog/walk (walk mostly) for the rest of the marathon.

  15. Cheating related to very long distance running is certainly shameful, but I’m not at all surprised that it happens on a regular basis, and I’ll tell you why.

    That portion of the public who is willing to voluntarily go to great lengths and time commitment to put their bodies through the great strain of a marathon-length run are not like the rest of the general public. These (mostly waif-thin) athletes push themselves to do extreme things with no apparent reward other than the congratulations received from others. If they were concerned solely with health benefits, I disagree that they would choose to run full marathon races of 26.2 miles. It has been well established that full marathons cause some additional damage to the body which is not inflicted by shorter (but still strenuous) races such as 5K or 10K. As many of you likely already know, the body cannot store enough energy for a hard 26.2 mile run, and compensates by partially breaking down some of its systems once the body’s energy store is depleted. I don’t mean the people who jog it. I mean those who run it hard, such as those who can run Boston, or wish (via cheating) to pretend they can.

    So here we’ve got a group of athletes so driven that they say “I want to run beyond normal levels until my body starts breaking down. That sounds like FUN!” Like I said, it’s not surprising to me that this group seems to also have a large group of cheaters in their midst.

    • I’m not sure if your goal is start an argument but I think your conclusion that you aren’t surprised by the number of cheaters based on a single full marathon is pretty much the definition of cherry picking data to support an argument. Maybe you can do a quick review of the contents of the site and see that out of fields of thousands there are normally 1 or 2 cheaters.

      This is a massive anomaly and not at all representative of the running community.

      • Most other races do not have the lax control that the Mexico City Marathon apparently had. And this is not a one-time aberattion. I do read the “contents of the site” and Derek himself said that “Honolulu and Disney” had “widespread cheating” and that he “estimated those races had 400 and 200 course cutters respectively”.

        When we start batting around numbers like hundreds and thousands of runners with invalid results, we’re no longer talking about “1 or 2 cheaters”, we’re now entering Tour de Marathon levels of ridiculous behavior.

  16. I’m supposed to believe that 1296 runners qualified for Boston at a,race being held at almost 7,000 feet in elevation?

    Sorry, I’m not buying it.

      • In the provided course map, you can see that the route has many jagged zigzags and switchbacks. It looks like some string after a cat played with it.

        I would assume that this peculiar-looking route introduced multiple opportunities for course cutting. Perhaps they also had an abnormally low number of course officials to keep an eye on things? That’s just a guess.

        • The advantage of this route, which recovered the Olympic Route Mexico 1968; is that the control points are at the edges remote from each other (in addition the closure of roads and colonies that serve as a cut, or do not save route or are dangerous as in the case of the exit to kilometer 5 in Tlatelolco, to cut would have which crosses by the well-known “Barrio Bravo de Tepito”), that is why many of those who get used to cutting routes (maintaining a balance that seems adequate in their rhythms) were exposed not passing through 2 or more records.

    • this is not the worst thing; Today I agree that the ranking of this marathon is very much to my regret; because the organizers are making up the results; the 40000 numbers were supposedly exhausted and in the official record – including the suspicions you mention – there are only 33000 records including the disqualified.

  17. Lord. Not only is the bib mule not hiding the fact that he’s wearing two bibs, he’s got the one labeled “Maria” on top. Talk about blatant.

  18. Don’t you find the amount of “cheaters” suspicious? We are talking about more than a third of the total participants. That will initially lead me to think that there was a problem in the bib tracking system. Maybe I just have too much faith in people hehe. I would expect SOME percentage of cheaters, but no 36%.

    • I touched on that in the article. I never said *All* missed mats were the result of cheaters. I even argued against the conclusion that all of the missing mats were the results of cheating on my FB page.

      The fact remains that whether from timing error or cheating, the validity of many Boston qualifiers is in question.

  19. I was a spectator at KM 34, and I’m not surprised by these numbers at all. The first cheaters came through not long after the elites. Until I left, around the time the 4:15 runners would come through, the vast majority of “runners” had not run the full 34 kilometers. And seriously, most could not have.

    It was awful for all the serious runners, as they had, at times, to run slalom through an increasingly dense field of joggers.

    While observing the run, I could not fathom what was happening. It certainly ruined it for me. Now I know why, but still cannot believe it.

  20. I’m not trying to give any cheater the benefit of the doubt, but with the volume of people apparently cutting the course at points where the course twists and turns so much, would it be possible that some runners simply followed the pack blindly? I’ve been on autopilot in races and haven’t paid attention to course markers. I simply went where the pack was heading.

    • I highly doubt it. The parts of the course that I have seen were all very well marked. And there were actually a lot of posts standing along it, both volunteers and police men and women. Plus, the runners that I saw (at km 34) were very diverse with many different running speeds.

  21. For context, what is the average number of missed mats on a course? Because these numbers are extreme, but it’d be interesting to compare with normal technical issues and a less egregious number of cheaters.

    • The most important curves and checkpoints were not only well signposted but also had containment fences; only a blatant trap explains the numbers; in addition there is a blue line of the whole route that is painted on the asphalt a few days before and every kilometer is marked with totems of several meters.

    • The marathon I work for had 1983 finishers this year, and we have 5 mat locations (not counting the start line), so that’s 9915 potential readings for our finishers (not counting DNF/DSQ runners, and not counting our relay teams which also get chip splits – Marathon finishers only). We had 53 missed readings out of the 9915 opportunities, approximately 1/2 of 1%. I believe that is about what the chip companies advertise for expected read rates. FWIW, we use B-Tags. We’ve been with B-Tag/D-Tag for about 10 years, the read rate has improved significantly over that time. Early on I’d estimate our read failure rate was closer to 3%.

      • Thanks for the info, Joe. I have never been a race director, so I was curious. That data definitely puts this in context.

  22. The first picture is from 2016 Marathon, not 2017. Second picture clearly violates this man’s privacy. You did not know his side of the story.

    Lastly, your post is clearly racists and against Mexicans, very in line with your current government. I hope you “personally” check all BQ, including those from US Marathons.

    • I’m not claiming that photo is from 2017. The caption names the Facebook page from which it is being attributed.

      There’s really no side to hear. He’s wearing two bibs. It was posted on social media.

      How is it racist? I’ve made no statements or judgements about any race or nationality. I’ve written about dozens of US races and runners. This is the first post time I’ve written about Mexico City.

      • The rights of the photograph are of ORPHOTO who distributed them free of charge; thanks to them I have a photo of me of good quality; in addition to signing the consent of the marathon, one of the clauses declares the public of the image when registering, that is to say, rights are lost on the image itself; the racism seems to me unjustified, since a good percentage of cheaters are of other nationalities and races, including whites, the character of this event is international, how then could we fall into myopic judgments about our race ?. You need to be more neutral and less susceptible.

    • He has literally investigated dozens of US races. While there is certainly racism out in the world, this it clearly NOT an example of it.

      Also, the picture doesn’t violate his privacy at all. He’s out in public.

  23. This is disgusting , what I have suspected has come to light! I will not associate my self with such a tainted marathon in the future. It is a shame, because Im Latino, and my roots are from here. When I drop the ball at work , the consequences can exclude me from other group participations, and lead up to termination. THIS MARATHON SHOULD BE EXCLUDED FROM BEING A BOSTON QUALIFIER. I will personally use my valuable time, and due diligence to expose those that I catch that have cheated.

Comments are closed.