2017 Boston Marathon Results – First Look at The Data


As I bring in all the data and crunch the #s to prioritize the list of runners review, I thought I would share some findings.

The data below is based, in many on runners where I have identified the qualifying marathons for.


Average Finish Time for Those With a Qualifying Time (Elites Excluded)


Average Boston Qualifying Time (Elites Excluded)


(An 11.26% Differential. I calculated the 2016 variance as 10.4%)


Average Finish Time for Those With a Qualifying Time – MALE (Elites Excluded)


Average Boston Qualifying Time – MALE (Elites Excluded)


(A 12.41% Differential) 


Average Finish Time for Those With a Qualifying Time – FEMALE (Elites Excluded)


Average Boston Qualifying Time – FEMALE (Elites Excluded)


(A 9.91% Differential) 

Relative to their qualifying times, Women performed better in Boston in 2017.


Average Finish Time for Those That Used Boston 2016 as Their Qualifier for Boston 2017


Average Qualifying Time for Those That Used Boston 2016 as Their Qualifier for Boston 2017


(A 5.4% Differential) 

Runners that qualified for Boston 2017 using Boston 2016, on average ran slower than last year. Weather conditions were very similar to 2016. It was warmer at the start in 2017.


“Easiest Qualifiers”

Marathon # of Identified Entries Average Boston Time Average Qualifying Time Difference % Difference
2016 REVEL Mt Charleston 147 4:06:31 3:24:44 0:41:47 20%
2015 Indianapolis Marathon 20 4:03:32 3:26:24 0:37:08 18%
2015 St. George Marathon 284 4:00:56 3:24:54 0:36:02 18%
2015 Naperville Marathon 26 3:58:00 3:22:38 0:35:21 17%
2016 The Tunnel Lite Marathon 84 3:55:49 3:21:49 0:34:01 17%
2016 Albany Marathon (formerly the Snickers Marathon Energy Bar Marathon) 54 4:00:33 3:26:22 0:34:10 17%
2015 REVEL Canyon City Marathon 106 4:04:32 3:29:53 0:34:40 17%
2016 Light at the End of the Tunnel Marathon 91 4:01:34 3:27:46 0:33:48 16%
2016 Town of Celebration Marathon 32 4:07:20 3:32:45 0:34:34 16%
2016 Jack and Jill’s Downhill Marathon 41 3:59:30 3:26:25 0:33:05 16%
2015 Rocket City Marathon 20 3:51:39 3:19:43 0:31:56 16%
2015 Columbus Marathon 206 3:55:01 3:23:02 0:31:59 16%
2016 Mercedes Marathon 21 3:47:48 3:17:05 0:30:42 16%
2015 Mohawk-Hudson River Marathon 97 4:00:58 3:28:39 0:32:18 15%
2016 Louisiana Marathon 33 4:04:21 3:31:49 0:32:31 15%
2016 Glass City Marathon 80 3:54:26 3:23:22 0:31:03 15%
2015 Peak to Creek Marathon (formerly Ridge to Bridge Marathon) 31 3:57:35 3:26:07 0:31:28 15%
2016 Big Cottonwood Marathon & Half 198 3:55:53 3:24:46 0:31:08 15%
2016 Surf City Marathon (prev. Pacific Shoreline Marathon) 27 4:01:29 3:29:42 0:31:46 15%
2015 Grand Rapids Marathon 73 3:55:25 3:24:29 0:30:56 15%
2015 Victoria Marathon 45 4:06:53 3:34:35 0:32:17 15%
2015 Milwaukee Running Festival Marathon 21 3:52:20 3:22:01 0:30:20 15%
2016 Pocono Mountain Run For The Red Marathon 131 3:49:50 3:19:52 0:29:59 15%
2015 Lakefront Marathon 110 3:57:12 3:26:24 0:30:48 15%
2015 Melbourne Marathon 33 3:48:33 3:18:55 0:29:38 15%

For races where I’ve identified that 20 or more runners used that for their Boston qualifier, I looked at the variance in qualifying times vs. Boston times. The races above had the largest variances, which indicates that these qualifiers were the “easiest” to qualify for Boston ’17. The Tunnel Marathons and Jack and Jill are run on nearly identical courses, and comprise 3 of the top 10. Many of the races also have significant overall declines. I plan to do further analysis on this list for a future article.


I will have much more to come in the coming days. I am sure I will have incidents of cheating to report on. The data above is just s portion of what goes into the algorithm that I use to prioritize the result review. The biggest improvements to the algorithm will come from better understanding of the qualifier marathons. I will use all of the data I have to determine ‘predicted finish times’ and review based on variances against the predicted finish times. I will again reiterate that if you had a bad day, no need to justify your time to me.


Thank you to all of those that have already contributed to the site. Contributions go towards site expenses, and to compensate those that help gather the data. Please consider making a small contribution to support the site.


One Time Contribution



  1. A lot of things jump out as obvious. A this year was a scorcher so people are going to be going slower. B a lot more people take Boston as a victory lap rather then a serious attempt. C as far as some of the race differentials I know that Charleston is a fast screaming downhill race that had absolute perfect weather last year (I am doing it in 11 days). ST George and any of the Revel races like Big Cottonwood and Canyon City are also very downhill. I ran both St George and Big Cottonwood last year.

    • Yes, most of the races are obvious. Quantifying will allow me to push the runners who used these races down the list to review. In the past 2 years, I spent too much time with runners that were prioritized too highly that qualified at some of these races.

  2. SEND DEREK TO THE 2018 BOSTON MARATHON. All: Today I read an article that reported that Derek has never run the Boston Marathon. (http://www.sacbee.com/sports/article145022939.html) This, of course, is unacceptable! So, also today, I set-up a GoFundMe to send Derek to Boston 2018 as a charity entrant. https://www.gofundme.com/derek-murphy-to-the-boston-marathon. All proceeds will be used to get Derek entered, which means the funds will be used to support a 501c3, since that’s what the charity entrants’ funds are used for. Any remaining funds will be used to help get Derek there and home. I don’t know Derek except through his great work on this website. Seems fitting that he run the race he protects for runners and others who simply want to preserve the integrity of the Boston Marathon (and frankly many, many other marathons). So, donate to the website here, but also donate to help send the guy to Hopkinton next Patriots Day.

      • I was about to donate and then that exact thought crossed my mind. But if Derek wants to endorse this effort I will absolutely contribute!

    • The funds go to charity, the secondary benefit is that the donations (if enough are raised) also result in a complimentary entrance into Boston. He, of course, does not have to accept (as fun as it would be to force someone to run a marathon). But it’s a win/win, and mostly a win for the charities that participate in the BAA program.

      • I do appreciate the sentiment. I do hope to run Boston one day. Not sure that it would be 2018. I plan on running a fall marathon and going from there. Likely that would put me out of the running for being able to run with a charity bib for 2018. I don’t want to run Boston if I’m not sure that I can complete the distance. If and when I do decide to run Boston, anyone would be welcome to contribute to whatever charity I am signed up with.

    • It’s got to be Australia. The Publix Florida Marathon held in Melbourne Florida only had 226 finishers in 2015 so no way 33 would be BQs.

    • I expect it’s Australia too, although it’s interesting to see it on the list, given that the course has slight elevation (according to my Strava data, 201m of elevation vs 138m for New York). I’m looking forward to Derek’s analysis and insights!

  3. So, Derek: With all the data you’re compiling, are you going to enter the cutoff prediction fray later this year? Seems like a natural road to go down…

  4. Just curious, how exactly do you weed out the legitimate folks who had bad race days, were ill, were pacing a friend, etc? Sorry if you have addressed this previously. And kudos to you for all the great, hard work you do for the integrity of this sport!

  5. Another factor is that Boston entrants can be up to 1.5 years older than when they qualified. This doesn’t matter when you’re under 30 or 35, but after 40 it can be significant. It also wouldn’t have mattered for previous years’ analysis, but your algorithm’s sound so precise that this may be worth taking into account.

  6. Why do you think the Indianapolis marathon (monumental) is #2 on the list of easiest? It’s just flat, not downhill at all. It does consistently have pretty good weather, but I don’t see why that would make it “easier” than all those downhill marathons. That’s where I qualified the two times I qualified (totally bombed Boston last year and probably will be slow at Boston 2018 if I run because it will be 6 months post baby and I’ve been out a couple months with an injury). I’m sure I’ll be on your list of people to check as I expect I’ll be 30 minutes slower than my qualifier (on a good day).

    • Because weather is the number 1 way to make a marathon easier hands now even moreso than a downhill course. The largest organ on your body is your skin if its too hot or too cold your body has to expend energy to either warm you up or cool you off. That is energy you no longer have to run. I think a lot of people are going to be shocked at Charleston this year as many have booked it for a BQ attempt after 40% BQed last year. But last year was a one year race its first year that had absolute perfection as far as weather goes. Yes its a downhill course but it had perfect weather. Its going to be hot hot hot this year and people are going to suffer. I think you see the BQ rate fall by 15% this year if not more. Weather is the most critical element for an easier race.

  7. “Weather conditions [in 2017] were very similar to 2016.”

    To be completely fair, you should not only take into account the weather on the day of the marathon, but also the weather leading up to the marathon. Most of the New England-based runners I know performed very poorly relative to their potential even though it wasn’t scorching hot because it’s been a relatively cool Spring and getting acclimated to running in the warmth wasn’t an option.

    • TRUE! 2017 was nothing like 2016. I have run the last 7 Bostons (2011-2017) and this year was worse than 2012…I had to walk. Med tent to rehydrate. I’m from midwest and no training run was over 50’s except one. It was brutal. So many people stopped to help me! Last year I was about half hour faster! I barely trained for that one (I have a horseback riding accident issue that flares up) and trained well for this one. It was also a little cooler in Athlete’s Village last year compared to this one. This was the first of 17 marathons in the last 8 years when I didn’t know if I’d finish.

  8. I think the weather was a little worse than last year. It was slightly cooler last year and we actually had a headwind for a good part of the race that to me was a welcome cooling breeze as opposed to the hot tailwind we had this year. Granted I was pacing my dad last year so I’m not sure what I would have felt like had I actually been racing.

    This year was rough. I backed off after about the 7 mile mark when it became readily apparent that a sub 3 hour time was not only not in the cards, but continuing to pursue it would have been an exercise in late mile misery as it was for so many people that I passed in the second half. Very frustrating as my training was good and my workouts and shorter races leading up to this indicated I was in sub – 3 shape. But I’ve run enough of these to know that fighting the conditions is ill-advised and you need to accept what race day offers you. If not you end up walking a good portion of the second half or end up in the medical tent. Or both.

    I ended up running 11 minutes slower than my qualifier which all things considered wasn’t too bad. I’m hopeful that the 2:34 cushion that I have will be enough to get me in next year. It’s cutting it close, but I’m not sure I want to run another one between now and then. I do want to be back and hopefully see better conditions.

    • My story exactly how about 2 bqs this year already Boston 1:32 ,Disney 1 min . I think all Boston BQ on there course should be automatic as why should I get kicked out on a course that’s easier better conditions . 2 bqs this year and one at Boston and I might be out . Let a Boston BQ in automatic it’s there race and standards

  9. The spreads you are showing seem really wide. Have you considered looking for a correlation with marathons that have an out and back. My 2017 Boston time 3:29 is within 16 minutes of my 2015 Big Cottonwood time 3:13 where I PRed and 9 minutes of my 2015 Surf City time 3:20 and 4 minutes of my Qualifying time 3:25 from 2015 St George (3weeks after Cottonwood).

    Thank you for all your hard work… hang’m high!!

  10. I fully intend to just have an easy jog from Hopkins to Boston should I ever be so fortunate as to qualify

  11. I would agree with some of the comments. I qualified with a 3:06 at one of these marathons and the heat destroyed me at Boston. I had a 1:33 half and then completely bonked walking many of the hills. Last year too, my wife tore her labrum in her hip 3 weeks before Boston so we walked/ran to a 1:30 over her qualifying time.

    But, I do appreciate all the hard work here to keep the sport clean. Keep up the good work.

  12. I ran 3:39 :07 Disney in Jan. As half was cancelled and terrible course 1 percent qualify BQ there I did . I ran 3:38:28 in Boston this year as I’ll be 57 and 2 BQ in 3 months . Due to other races easier in country I might not get in . I feel Boston sets the standard and if you run there time at there own course should be automatic in the next year . Why should some lie downhill course keep me out . Run the BQ at Boston not some other easy coursec. .let a Boston BQ at Boston stick it’s your rave and I ran a BQ there .

  13. Is there an easy way to see a list of all the 2017 finishers? The search function on the BAA.org seems restrictive (e.g., limiting searches to the top 1000 finishers).

    • I have a sheet of all the finishers..but since I wasn’t the one that actually pulled the data, I’m not comfortable sharing. Also I don’t want people to contact everyone that missed a mat.

  14. Just curious: how many runners missed both the 30K and 35K timing mats this year? I emailed you about one that I am certain I saw take the train but looking at his splits and based on the mats that he did actually cross, it isn’t obvious that he cheated….Does it become more statistically obvious that folks who missed both mats likely cheated if there are only a small number of runners who missed both mats?

  15. Interesting to see Columbus 2015 on this list as there are a couple of first time marathoners that BQ’d at. Columbus 15 but did not have course photos between miles 13 and the finish or any short races that Support their marathon time.

    • Hmm, I BQ’d at Columbus 2015, I don’t recall have many photos from the 2nd half (I just checked my Marathonfotos from there and I can’t tell where most of them are (except for by the stadium). However, all my race data support my times and I have missed no mats.
      Looks like I did better than my cohort group from there as I only slowed by 9 minutes from Columbus vs the average slow down of 23 minutes, but I trained a lot harder this time.

      I just donated to support this wonderful site!

  16. For future data dump articles like this would you consider also displaying the standard deviation for the averages?

    It’s quite interesting seeing the difference between the average qualifier and average Boston times, but without knowing how “tight” that average is you can’t see if it’s because one of the races or because of a bad fit. I don’t mean to lecture you on statistics! It’s just that for somebody like me just looking at the averages they don’t actually mean anything by themselves.

  17. OMG it was so much hotter this year than last! At one point, it was in the low 8os inland. It dropped 20 degrees by the time we got in to Brookline where there was suddenly cloud cover and some cross-wind.

    I was DEFINITELY 35 minutes slower than normal… but I planned from the get go, to do this as a super easy, fun, catered long run. LOL Nevertheless, everyone was in the same boat. The air was hot and suffocating. It is nice to see how all suffered along.

  18. If I was running the event I would put some mats at the subway entrances to see if they get any hits. I bet the results would be interesting

Comments are closed.