I wanted to elaborate more on the project that was the topic of the Runner’s World article. You can go to the bottom of this post for links to 3 particular cases that I felt warranted further attention.
A group also was formed on Facebook to discuss racing and to investigate potential cheaters outside of a public forum. Last August I made the following post to the group: (typos included)
The group got together and was instrumental in coming up with the initial process and criteria of who to review. Boston issues bib #s based on qualifying time, so we were able to accurately estimate BQ time based on the bib # of the runner. We settled on marking for review everyone that ran Boston 20 minutes slower than their estimated qualifying time. One member came up with the scatter plot of bib #s vs. Boston time. You can see a clear trend, and see clearly where the charity bibs start. We plotted Boston finishing times vs. bib #.
 |
| Boston bib #s plotted vs. finish times. |
The runners in orange are those that we have flagged as having invalid results with 100% certainty. The runners in blue were either cleared or we just did not have enough evidence to put them in with the invalid results. The green dots are runners that were initially marked for review but have not yet been examined. The grey dots were not flagged initially. There were a few instances where other information and tips led to a review some of those runners.
At greater than 40% variance between Boston time qualifying time, about 11% of the runners were marked as having invalid results.
At 20% variance between Boston time and qualifying time, the rate of invalid results leveled off at around 2%.
In total:
1409 runners reviewed – 54 results deemed invalid (Additional runners were detected that were not validated prior to sharing the data with Runner’s World
– either at the qualifier, or at Boston (3.8% of runners reviewed in total deemed invalid)
- 12 cut course at qualifier
- 33 bib swaps – someone qualified and sold/gave bib to another runner
- 11 others – Did not qualify legitimately – either someone else ran the qualifier under their name or race results were modified.
With the 2016 marathon less than a week away, the BAA is focused on putting on a safe and successful event. I suspect that anyone not yet removed from the 2016 race will be allowed to run. I will follow up again after the race and give the BAA and qualifying races all the information they need to make decisions on these runners for the future.
Although I am not naming all the runners that were found publicly, there are some cases that I felt were noteworthy to report on – the articles are below.
2014 Marathon du Monte St Michel – Results Falsified to Give BQ Results?
- Multiple runners qualified for and ran Boston that do not appear in the photos or the finish line video.
Running Blogger’s Marathon Times Under Scrutiny
- Seemingly impossible splits backed with social media evidence shows evidence of course cutting.
- This runner has admitted to the evidence that was in the article and taken down his blog. As a result, I felt it appropriate to remove the article.
Husband and Wife Team Have a History of Cheating to Run Boston
- Multiple instances of course cutting, disqualifications, and a forged bib.
Thank you for your diligence in catching cheaters and alerting race officials.
I would also like to express my appreciation for your perseverance in this tedious endeavor, all-be-it, quite possibly, of utmost importance to those of us who have qualified or those who would give their right-eyebrow to be a qualifier, hah! 😉
Incidentally, while it may be water under the bridge at this stage, I personally know one cheater: Todd Oakes, who ran Boston in 2013 (Bib#: 8084, Time: 3:42:54) without having qualified. He needed a time of 3:15 to qualify, and, at the time of running Boston that year, his PR was 3:22:10; that PR was at the Wineglass Marathon on 9/30/2012.
Does your analysis factor in a qualifying marathon that might be flat course versus Boston which is reputed to be a tough, hilly course (I'll likely never know)? Or other things like weather or anything else that impacts a runner?
What about someone running Boston with a minor injury that pulls their time down?
You are my hero! I read the RW article today that used your data. Thank you for all your hard work. I'm so tired of these cheaters. No to figure out a way to catch them before they submit their times to run Boston.
Going into the project, we did not know where they qualified at. But, to your point, if I found that they qualified at St. George or a similar fast marathon, it was easier to clear them as having a legitimate result. There are many valid reasons someone would run slower than their BQ time. This was our method to narrow the initial pool to review.
To date, in my seven consecutive completions of Boston, I've done it once with a stress fracture; diagnosed only 3-weeks prior to race-day. Thankfully, with my Dr.'s blessing, I walked it for the most part (only ran the up-hill sections, b/c that would not re-injure my healing foot) in a 12-minute pace exactly. My time was somewhere around 5:15; no where near my qualifying time.
This year, I will be completing it for an 8th time, at a far slower pace than when I qualified in September. Why? Because I was hospitalized in February with acute Hepatitis A; had never heard of Hep. A before then. Anyway, it wiped me out! The recovery was painfully slow. In terms of running, I wasn't. It meant I missed the critical training I needed for Boston; I'm okay with it, so long as I can finish, I will be happy enough with that.
Variances could be due to injury or in the case of a runner that I know personally, pregnant. How do you take these things into account when you do your investigations. Your scatter graph just points out that some people had a really shitty race. That being said, if someone cheats then they should be booted and banned for life from the Boston Marathon.
You are correct. That was just how we narrowed the field to something more manageable. Most runners from that group were easily cleared. If we looked at the qualifying race and pictures showed it was the same runner and the splits looked good (if available) we would check them off. Unless there's something to raise additional questions in the qualifying race, we'd move on.
Also, the rate at which we caught cheaters declined as the variance dropped – so it indicated that the way we filtered was at least somewhat effective (there may be better ways – or additional criteria we can add to be more efficient – if we knew the qualifier race beforehand we could have accounted for course difficulty, etc.)
Cheaters should be handled with a zero tolerance policy Wendy!
As a fellow runner and an engineer, I find this data analysis quite fascinating. Keep up the good work!
I ran Boston 2015 after being in Tufts Medical Center ER with food poisoning less than 40 hours before the race. I was not going to miss my chance to run the race with my sister and yes, my time was not close to my qualifying. It was over 30 minutes more, but running on two bags of IV fluids on a cold rainy day was still not going to deter me from competing the race. Thanks for your investigations and hard work. I hope you share the above information with the Boston Marathon officials so any cheater will be denied entry in the future. I work really hard at my running as do most of the runners out there.
I ran Boston 2015 and actually beat my BQ time (just barely — by seven seconds), but because of all the fast qualifying times for 2016, I couldn't participate in 2016. I was stunned when I got the email informing me I wouldn't be visiting Boylston St. on Patriots Day. Even removing the runners who got in under false pretenses, I likely still would not have qualified. But I think of the runners who barely missed and were excluded as a result of someone cheating. Words can't express the anger. I will try again for 2017. If this site ever needs financial support to continue its mission, I hope you'll make that an option for supporters in addition to visiting Amazon through the site. Seems BAA wold have an interest in supporting this site as well. So any runner who's tempted to cheat to get into Boston might think twice. It's a public service for the running community and beyond. Keep up the great work.
Me and Marlon have a mutual friend. Me and this friend both run a sub 17:30 5K. This friend has run with Marlon and has told me that he is not capable of the BQ. The guy is a cheater,unfortunately. The 5k time he sent you is obviously a red flag. You'd have to be Ray Charles not to see that. I hope you follow up on this. Make sure this cheater gets what he deserves. He is still on my friends list on Facebook. I want to see how this plays out. His sheep will choose to ignore the evidence and continue to support him.
This year forth, the action takes place in Llanwrtyd Wells, host town for the unique horse vs. man marathon.
I just want to say good work exposing these frauds. Have you considered using the marathon time converter at http://www.findmymarathon.com to help find Boston finish times outside the 20% range? Not all marathons are equal. My running club uses this to standardize times between different marathons. You could convert their Boston times to their qualifying marathon time. I would think that you would be able catch a few that fell out of the range using just straight up time comparisons. Just a thought.
I actually have done just that. I was already in process of adjusting for the qualifier when I came across those rankings. I used those to adjust the predicted Boston finish time. Mainly it served to drop some of those with large gaps down or off the list that qualified at the fasted marathons.
Wanted to say good work Derek. Fantastic effort and a worthy cause.
I cannot understand those who criticise what you're doing. I can only presume they're also marathon cheaters and/or Mike Rossi 😉
Do you have plans to widen your search to runs other than Boston?
This comment has been removed by a blog administrator.
Comments are closed.