Sunday, August 21, 2016

2015 Wineglass Marathon Analysis

Like Mohawk, this one came on the analysis radar when the BAA released a new set of top feeder races a couple months ago. I ran this race in 2012 and really enjoyed it! I definitely want to do it again at some point in the future (maybe 2017?). It's small, point-to-point, with a very gradual downhill profile. It goes through lots of small towns and while there are some longer stretches with no spectators, the communities on the course make up for those quiet portions. It is already sold out this year, so, this is one you would want to register for earlier rather than later.

It finishes in downtown Corning, NY which has a history in glass production. The medal is blown glass and very unique!

On to the data!

AG Group2016 Qualifiers2016 AG TotalPercentage2017 Qualifiers2017 AG TotalPercentage

So, Wineglass 2015 didn't go as well as 2014. Looking at the weather, it was a little warmer (44F at the start instead of 34F), but still very good temps. Either way, about a 3.5% rate of qualification decrease year over year.

<1 minute186.12%146.33%
1-2 minutes237.82%135.88%
2-3 minutes268.84%125.43%
3-4 minutes186.12%167.24%
4-5 minutes134.42%125.43%
5-10 minutes6823.13%5424.43%
10-20 minutes7124.15%5625.34%
20> minutes5719.39%4419.91%

Squeaker Pack 2016 vs 2017: 33.33% vs 30.32%

This race, interestingly, has a high percentage of less than 5 minutes margin qualifiers.

The Totals!

AG Group2016 Qualifiers20156 AG TotalPercentage2017 Qualifiers2017 AG TotalPercentage

So, the last update, the BQ rate for 2016 was 14.94%. After the 2014 Wineglass results, it remained unchanged. For 2017, the 2015 results bumped the rate down .01%. So about the same, but obviously, this will not make the prediction go up.

The breakdown of margins of qualification:

<1 minute18555.55%16065.63%
1-2 minutes18245.46%16085.64%
2-3 minutes18405.51%15365.39%
3-4 minutes16765.02%14255.00%
4-5 minutes15744.71%13154.61%
5-10 minutes700020.96%609221.37%
10-20 minutes938628.11%792727.80%
20> minutes824024.67%700324.56%

Squeakers 2016 vs 2017: 26.26% vs 26.27%

Pretty much the same.

The final bit of data... the number of qualifiers achieving at least 148 seconds of margin in the 2016 qualification year for the races we've analyzed is 28814

The buffer has gone up from the last update; there are now 302 fewer qualifiers in 2017 than that above number.

So, all things being equal (including runners' intentions to register and the field size set by the BAA), this implies that there will be no cut off.

Field Size

This leads me to a question for you all, spurred on by some runners in my running group and a reader's ("On Lam") comment about the BAA's plans as far as the field size of the 2017 race. After the 2013 Bombing (a race that did not fill up after the initial registration weeks and also the first year the standard were dropped 5:59 across the board), the BAA expanded the field size (there were also some folks who were not able to finish in 2013 and had their registrations deferred to 2014).

Over the last 6 years, the field size (this includes all types of registrants including charity runners):

2008 25K
2009 25K
2010 26K
2011 27K
2012 27K
2013 27K
2014 36K
2015 30K
2016 30K
2017 ???

I didn't look for data before 2008, but we can see that it has steadily climbed and never decreased since 2014. 2014 was a bit of an anomaly given (a) the extra deferments, (b) heightened level of interest post-attack.

My conjecture is that given how big the cut off was for 2016, it is unlikely they would choose to reduce the field size for 2017. I could see that if there is no cutoff this year, they may consider reducing the field size for 2018 (though, the precedent to reduce it isn't there, looking at the last 9 years). If I had to bet, I would say odds are better they leave the field size as-is for 2017.

Do you have an opinion/theory on what the BAA will do with field size for 2017? Let's hear it!


  1. This comment has been removed by the author.

  2. I have been wondering what the field size would be, as well. My guess is that it will stay the same for 2017 and that the BAA might start looking at new options for 2018. At least that is what I am hoping for!!! I only have a 2.37 buffer and keeping the old fingers crossed that it will be enough!!!

  3. My bet is field size will be the same, why would they decrease revenue and people coming into the city when it's worked in prior years? They will fill it, no matter what.

    Thanks for the updates, this helps keep me sane in the last 4 weeks until we know if we made it! Best of luck to all us Squeakers.

  4. I love reading this blog. I qualified in July with a 3:33 buffer, so I'm confident but nervous. This blog somehow helps me remain excited (things look good for squeekers) and realistic (there's no guarantee I get a bib). Thanks for all the work you're doing!

  5. Thank you for doing this. As a squeaker this year (2:50 buffer) I like my chances as long as they don't drop the field size. I get the impression they are comfortable at 30k, but I could be wrong.

  6. The only way I see a field reduction is: when registration closes for the <5:0, they close registration with a slightly smaller field but with no cutoff.

  7. I don't see why a "no cutoff" outcome in the initial registration period would lead them to lower the field size cap. The race will still fill-- registration will stay open, so all the qualifiers in the fall marathons will have the chance to register.
    It seems clear that the drop in qualifiers this year, so far, is due to unusually hot conditions at some of the major feeder races, not from lack of interest. I was at Duluth, it was pretty brutal. Myself and almost every runner I spoke with missed their target by 10-20min.
    They've probably determined that 30k offers a feasible balance betw revenue, cost, and capacity of the host towns, hard to see what causes that to change.

  8. Thanks for this blog. It has been helpful in prepping for my BQ attempt, today st Santa Rosa. Weather was perfect, but the carnage continues. The entire first corral took a wrong turn in Mile 1. I had to add 0.84mi to get back into the course. I still qualified but most others did not. Not sure what the race director can/will do, but it's looking bad. As you can imagine, hundreds of people were very angry at the finish.

    1. Oh man. No, cannot imagine! After all the training, traveling, managing to get to the line free of injury & illness. I've missed my target race due to a month-long chest infection. I can handle this and will have to focus on Boston 2018.

      To have a race course error like this occur? I would be devastated.

  9. Can anything be done in cases such as this ?

  10. I haven't yet heard any response from race officials. Here is more info:

    1. ^ Wow! that's ugly, hope the SR officials can come to a resolution for those runners. If not, this will spread like wildfire and wouldn't expect many BQ attempt runners to be joining this Santa Rosa's event for next year. Crossing fingers for those squeakers who needed this race to BQ.

    2. You know, even if the SRM officials develop a plan (as opposed to just hoping it goes away), would the BAA accept it? I managed to find a way to use the adversity to my advantage and ran my 26.2mi at 8 minutes fast, allowing my 27+ to still be a BQ by 4 mins, but could I be DQ'ed from Boston because I ran off-course by 0.84mi? Even though I re-entered at the same point? Probably not, but that's my paranoid fear.

  11. This is certainly going to make the numbers interesting for Santa Rosa. Boy, has 2016 not been a good year for races - it is isn't the weather, it's stuff like this.

    I'm really not sure what they can do. Without mats, how do you determine who ran longer and who didn't in order to give a time credit? Honor system? Err, then you might be giving people time who didn't actually run extra, and, as we know, plenty of runners will say they did and didn't and cheat (ahem, runner whose name that rhymes with Flossy is a recent example). Garmin data, maybe, but then how does the BAA feel about that data?

  12. Wow! That sucks! I thought the fact that the OC marathon was .2 miles longer (26.40) is where my garmin ended along with many other runners I asked, was bad. So sorry for this news on Santa Rosa.

    1. I have run 7 marathons in Toronto and Ottawa. All are big races. Every time my Garmin logged 200-400m more. I was told by other runners that it was normal as we zigzaged thru the course.

    2. Agreed. Also, the measurements are done so that they make the shortest possible path, and then they add 0.1% for wheel wobble of the measurement device.

  13. A few years ago this happened in San Antonio, and I knew from impacted friends, times were adjusted

  14. I'll update when/if I hear anything about what they'll choose to do.

  15. Thank you so much for doing this throughout the year. It was extraordinarily helpful. And if we "all" get into Boston, we need to buy you a beer or two when we're there. Re: Santa Rosa: Awful. If it can be verified, they should at the least reimburse race fees for anyone running with that pacer. I don't fault the pacer as much as the race organizers. For goodness sake, have course marshals or clear markings when turns are upcoming (Forgo some of their "profit" and put it back into an already flawed race?) Seriously, demoralizing and I hope that at the least some small accommodations are made for those who were impacted by this artificial screw-up.

    1. The race director replied back to me yesterday that they are communicating with BAA to see what can be done.
      For some reason I don't see them refunding 500 marathon fees (that's a guess at the number since it seemed to be the entire first corral), but I agree that something along those lines would be a good faith gesture.
      What is unsaid about this is that people tried to make up the time, so almost everyone bonked. Just a giant fail for the race.

    2. from The Press Democrat: "Sarabi (race director) said the marathon will do everything it can to make things right with aggrieved runners. He is offering anyone incorrectly led off course a free 2017 Santa Rosa Marathon entry, or a refund of their 2016 entry fee. Interested parties should email him at

      Sarabi added that race organizers are currently working with the Boston Athletic Association to sort out complications with those Boston Marathon qualifications. That might be easier said than done, though.

      Jack Fleming, director of communications for the Boston Athletic Association, explained that his organization doesn’t do detective work. It lets USA Track & Field sanction the qualifying marathons, and it relies on official results forwarded by those individual races."

      “We do not go out and have a representative at each of these races. There are probably 250 to 300 in the U.S. alone,” Fleming said. “We generally rely upon the race management itself.”

    3. This situation is crunched for time - registration opens for the speedsters very soon (9/12 - less than two weeks). IMO they need to have it figured out and updated by then.

      I may sit on the Santa Rosa results for a while - start working on the pulling previous years of the other two still left (Via and Erie) that way I am half way there when the results come out.

      It's really too bad this went down at Santa Rosa. The lesson here (and even why I am running CIM in December) is not to wait until the last minute to attempt to BQ (if at all possible). I can understand if it was a second or third attempt for the year, but I would never use this as my sole attempt to qualify. There just isn't enough time to re-up if something goes wrong.

  16. Wow. What a mess. It's still better than the Columbus marathon where their course is 26.55 miles due to the 64 turns that must be negotiated.

  17. Wow. What a mess. It's still better than the Columbus marathon where their course is 26.55 miles due to the 64 turns that must be negotiated.

  18. Boston actually took Santa Rosa off their list of top feeders (at least looking at it right now) Verrry interesting...

    1. That IS interesting! I have a big-ass spreadsheet of data, an I had copied the new list in there when I saw the update and Santa Rosa was definitely on it. Now the list only has 24...

      This is giving me pause as far as if I should even include the data in the analysis.

    2. I love what a small community this is. I know I've been posting here anonymously, but it's interesting to see the same names in various sites discussing this.
      As for your last comment, and I'm far from a statistician, I can tell you that I and surely many others hit their BQ last Sunday at SRM. I suspect it won't matter this year, since the cutoff may be 0:00 anyways, but there will be qualifiers from race. Other than the additional 0.8-0.9mi, it was FAST. Perfect weather, great terrain.
      All that assumes I won't be DQ'ed for having run off-course, my paranoid fear, with BAA just throwing up their hands in disgust at the whole SRM ordeal.

  19. On a different note. Does anyone know when we'll know if we got in (not the official letter) but is there anyway to tell before that what the cutoff is? Like, any clues or anything? And does it really take 5 weeks to get an acceptance/rejection letter?

    1. So, they give you pretty much no clues - last year they did say they had high registration rate but not a whole lot of detail. I was sitting on eggshells for 9 days. The news broke and about maybe 30 minutes later (? not sure exact time frame) I got the confirmation email from the BAA. Really, the official acceptance letter is a formality (unless they can't verify your time but that is very rare).

    2. Last year, registration closed at 5pm on Wednesday, Sept 23. The cutoff time was announced at 10am on Wednesday Sept 30.

  20. Just got an email from BAA about the upcoming registration dates...they have kept the field size at 30,000!!! Yipeee!!! Hoping and praying my 3:00 minute buffer gets me in this year!

    1. That makes it just about 24,000 time qualifiers. They took 24,032 for 2016. Your 3 minutes is a lock for 2017.

    2. Hoping my 2:37 will get me in!

  21. I'm at 2:06 under. This is killing me waiting!

  22. If we expect a low cutoff... it means that we over 2:00 are safe?

    1. A prediction is never a certainty. Last year the cutoff was 2:28 (and the prediction from the other gal that did this same analysis was 2:15 - pretty darn close - it all depends on who decides to register, i.e. demand). If I had more than 2:00 I would feel pretty confident at this point. I have 2:37 and logic tells me that I am sitting pretty good, but of course, I'm still a little nervous. This weekend, the last of the feeders happen so hoping to get that data in as quickly as possible (it's a lot of races to import and compile, so, I'll do my best).

  23. Thanks Chacha for all your hard work! And to think, I thought my 1:05 buffer was in no way going to make it, so I signed up and trained harder for another marathon to increase my buffer. Which I did. It's going to be interesting to see those that squeak in. Congratulations in advance to those going to BTown 2017! See you there.

  24. Well, just got word. BAA is not allowing any adjusted times for Santa Rosa. All previous squeakers are even safer.

    1. Ugh, well, it doesn't surprise me because it's really hard to prove who gets what. And with other BQs that aren't in question (where as far as we know, the time is accurate) possibly not getting in because of the cutoff, it becomes a tough argument to win that the BAA should take adjustments where it's not known exactly who was affected and by how much. And Santa Rosa is a relatively small group of runners across all Boston qualifier races.

      Ah well, and I just noticed the results finally posted to an aggregator (because the race results provider they use makes it really difficult for me to harvest the results in an automated fashion - I really don't feel like right-click, name file, save 50 times) so I can use an existing routine to pull them down. I should have Santa Rosa done probably tomorrow night.

  25. The Grand Rapids BQ2 is running this weekend, calling for thunderstorms, poor guys. Small race - about 250. Last year 60% qualified there. Wishing October would hurry and get here so we know!

    Thanks again Chacha!

  26. There are 2015 runners registered for Erie on Sunday. Last year they had a BQ rate of 33%. This year should be higher since the 500+ extra runners is probably from those who missed BQ at Boston. I would think the BQ rate should be higher. 40% BQ rate would be my guess for this year.

    1. Weather looks decent - high 71F but the day before says 85! I think last year may have been a little cooler but overall, seems to be about the same. Possibly a little windy. We'll see! I've got the data for the previous years of the other last minute ones already loaded up so I'm hoping that loading this year's will be quick!

  27. Chacha, thanks for doing this. I'm a squeaker from the Santa Rosa Marathon. I was far enough back not to be affected by the course detour but thinking I wasn't going to make it until I found your site. Thanks!

  28. I'm confused why there would not be a cut off this year as speculated above? I don't think charity runners were accounted for. If the field size is 30,000 will likely be 24,000 qualifiers and 6000 charity- doesn't that still leave several thousand not getting in if ~29,000 qualified. I'm a squeaker...thanks!

    1. Hi Ginger, so this is simple a dataset where we work off comparison. There are likely way more than 29K Boston qualifiers when you take into account London, Berlin, every other smaller town marathon, LA Marathon, etc, etc, etc.

      We are simply comparing year over year of key feeder races.