Jim2's Running Page

Boston Marathon Qualifying Statistics

Home
About Me
Favorite Links
Contact Me

1/21/09

From time to time, questions arise on running forums concerning qualifying for the Boston Marathon. Questions such as:

 

  • What percentage of U.S. marathon finishers produce a Boston Marathon qualifying (BQ) time?
  • Are the qualification standards easier for women than for men?
  • Are they easier for older age divisions than for younger divisions?
  • Which races are more likely to produce a Boston qualification (BQ) performance?

Responses to such questions often either reflect individual experiences or opinions, are based on the results of a specific race, or attempt to relate comparatively to other sources of available data, such as the differences between men’s and women’s world records and age graded tables and calculators.

 

In an attempt to shed some light on answers to some of these questions and bring some reality to the subject, as well as to satisfy my personal curiosity, I researched and compiled BQ data from almost 2/3 of the marathons run in the U.S. in 2006, 2007, and 2008 as listed on Marathonguide.com. This essay is intended to report and, to some extent, analyze that data.

 

The criteria used for selecting the races included in this project are:

 

  • The race had to be run on a USATF certified course, which is a requirement for BQ certification, which excluded 81 or approximately 23% of U.S. races … most of them either trail races or very small events.
  • The race had to be run in 2008, but not necessarily in 2006 and/or 2007. The Lewis and Clark (MO) Marathon was run in 2008, but was shortened to 10 miles due to flooding on the course. Only 2006 and 2007 data for this race is included in this project.
  • Races with 100 or more finishers in any one year were included. Races with fewer than 100 finishers in all three years were excluded.
  • The Men’s and Women’s 2008 Olympic Trials Marathon, which were run on 11/3/07 and 4/20/08, respectively, and in which all participants ran a BQ time, are not included.
  • Race results had to be tabulated and presented by gender in 5-year age divisions or individual ages had to be listed so that results could be manually compiled into appropriate BQ age/gender divisions.

Based on the above criteria, the data base includes a total of 227 marathons … 205 of 347 marathons run in 2006, 219 of 340 marathons run in 2007, and 226 of 357 races run in 2008.

 

Data Base

 

The 59-second “leeway” permitted by the Boston Athletic Association in each of the age-based BQ standards was factored into this study. For instance, the 40-44 male standard of 3:20 was considered to be extended to 3:20:59.

 

The data base includes the results of 1,172,229 U.S. marathon finishers in 2006-08, which is approximately 97% of the more than 1,210,000 million estimated by Marathonguide.com to have finished U.S. marathons during the three year period.

 

In a few cases, the number of finishers included in this project’s data for a particular race is less than the official number of finishers. These instances occurred primarily in races that had Clydesdale and Athena divisions with no ages provided and who were not included in the age divisions’ results. In these cases, it could not be determined how many, if any, met BQ standards. Since Clydesdale and Athena runners seldom BQ, exclusion of these participants tends to inflate the BQ rates somewhat, but the numbers are so small as to be negligible.

 

The data and resultant BQ rates are based on finishers’ ages on the day of the individual races, not what their ages would be on the day of future Boston Marathons. No consideration is possible for those cases in which a participant did not achieve a BQ time for his age on the day of the race, but would qualify for Boston by moving into a new age division within a qualification window of up to 18 months. The result of this factor is to understate the resultant BQ rates to a degree that could be significant ... perhaps by as much as 20% in the 5-year age divisions that range from age 35 to 79 and to a lesser extent in the under 34 and 80+ divisions, if the actual BQ performances were evenly distributed across the ages of those comprising each division.

 

Unless otherwise noted, data and results included in this report are 3-year averages with the exception of 24 races were not run all three years. In those cases, the data reflects 1-year or 2-year averages, as appropriate.

 

National Averages and Individual Race Summaries

 

It isn’t surprising that the marathon that produced the largest percentage of BQ finishers (37.5%) was Boston, itself. Boston’s BQ rate was so much greater than all but a couple of the other 226 marathons that, combined with the race’s size, it has a very noticeable effect on the overall BQ rate … 11.9% including Boston vs. 10.4% without Boston. On a gender basis, the overall BQ rates with/without Boston were 12.1/10.7% for men and 11.5/10.0% for women. Of course, individual races varied widely from Boston’s 37.5% highest rate to the lowest rate of 2.7% in the Los Angeles Marathon.

 

The commonly used rule of thumb that on average about 10% of marathon finishes result in a BQ time appears to be a reasonable one, especially if Boston finishers are excluded.

 

The 27 races that averaged a BQ rate greater than 20%  over the three year period are shown in Table 1.

 

Marathon

Finishers

# BQ

% BQ

Boston

20,653.7

7,749.3

37.5%

Bay State

858.7

292.7

34.1%

Mohawk-Hudson River

483.3

161.3

33.4%

Inland Trail

150.0

44.5

29.7%

Last Chance for Boston

129.3

36.0

27.8%

California International

4,571.7

1,216.3

26.6%

Tucson

1,054.0

275.7

26.2%

Newport

613.0

152.0

24.8%

Tallahassee

147.7

36.3

24.6%

Run for the Red Mountain

388.0

95.0

24.5%

Wineglass

526.7

128.3

24.4%

Steamtown

1,644.0

399.0

24.3%

Clarence Demar

254.3

60.3

23.7%

Skagit Flats

173.7

40.7

23.4%

Ridge To Bridge

97.3

22.3

22.9%

B&A Trail

212.7

48.7

22.9%

Foot Traffic Flat

114.0

26.0

22.8%

Ohio River Road RR Club

116.3

26.3

22.6%

St. George

4,894.3

1,091.3

22.3%

North Central Railroad Trail

366.3

81.7

22.3%

Grand Rapids

1,149.0

251.0

21.8%

Green Mountain

206.0

45.0

21.8%

Hyannis

345.0

74.7

21.6%

Rails to Trails

171.0

37.0

21.6%

City of Trees

189.7

39.7

20.9%

Johnstown

102.0

21.0

20.6%

Glass City

284.0

57.0

20.1%

 

 

Table 1. Races with a 3-year average BQ rate greater than 20%

 

For an extended summary of all 227 races, see the BQ Overview sorted by BQ rate, alphabetically by race name, or by race field size.

 

Gender and Age Division Demographics

 

Figure 1 illustrates 3-year average BQ rates by race size for each gender and combined. For the >10,000 race size category (NYC, Chicago, Honolulu, Boston, Marine Corps, Los Angeles, Rock N Roll San Diego, and Disney, in order of race size), BQ percentages are shown with and without Boston to illustrate the impact that the very high Boston BQ rate exerts on the averages. The significantly lower percentages for this category without Boston, compared to smaller race categories, are probably due to a combination of:

 

         The popularity of the other seven mega-marathons that comprise this category with charity participants.

         A large number of first time and “recreational” marathoners who flock to these events.

         Participants running them to “experience” them rather than race them.

         Difficulty in pacing a marathon in a very crowded race.

 

Figure 2 illustrates 3-year average BQ rates by age division for each gender and combined.

Figure6.jpg

Figure 1. Gender BQ Rates by Race Size Category

Figure4.jpg

Figure 2. Gender BQ Rates by Age Division

 

Gender Comparison

 

Despite the oft stated opinion that women’s BQ standards are “easier” than those of men, women had slightly lower BQ rates in all categories of race field sizes except for the two smallest of 500-1000 and 100-500 finishers (Figure 1).

 

Some have opined that women’s BQ standards, which are 30 minutes slower than men’s standards for all age divisions, are “easier” than those of men and that BQ rates that are similar do not reflect their relative degree of difficulty because the number of female charity and “recreational” marathoners is greater than that of men, which tends to dilute women BQ rates. A frequently heard opinion suggests that BQ standards should be patterned after the difference between men’s and women’s marathon world records, which typically differ by about 8-10%. (The current women’s world record of 2:15:25 is 8.4% slower than the men’s world record of 2:03:59.) Women’s BQ standards range from 15.8% slower than that of men for the 18-34 age division, to 12.5% slower than that of men for the 60-64 age division, and to 10% slower for the 80+ age division. Setting BQ standards to reflect the 8-10% difference between world records would lower the women’s open standard, relative to the 3:10 standard of men, to about 3:25-3:30 from the current 3:40 standard, for example.

 

The problem with using world records (or American records) as a basis for judging the relative degree of difficulty of BQ standards is that overall and age division record setters constitute a minute segment of the running population who are, by definition, exceptional. They don’t necessarily represent differences in physiological abilities or potential of a broader range of even advanced runners, much less the more average marathoner.

 

Another data point is the difference of 30-minutes between the average finishing time among all U.S. men and women marathon finishers ... 4:29:52 for men vs. 4:59:28 for women in 2007 according to Marathonguide/com. Again, however, the question arises of the mix of men and women at the back of the pack diluting women’s BQ rates more than that of men.

 

I suggest that a better frame of reference is what occurs at the front of the pack ... the relative performances of actual race and age division winners. Table 2 lists the average 2008 overall and age division winning times of men and women in the 226 marathons included in this project. Net (chip) times were used when available. Figure 3 illustrates the differential in average men’s and women’s winning times relative to the constant 30-minute differential in Boston's standards.

 

Winning Times

Division

Men

Women

Differential

O/A

2:35:30

3:03:42

0:28:12

< 34

2:38:11

3:09:04

0:30:53

35-39

2:48:36

3:18:51

0:30:15

40-44

2:53:48

3:23:55

0:30:07

45-49

2:59:14

3:30:47

0:31:33

50-54

3:11:30

3:46:33

0:35:04

55-59

3:23:13

4:13:14

0:50:01

60-64

3:39:42

4:35:21

0:55:39

65-69

4:07:42

5:04:36

0:56:54

70-74

4:42:19

5:39:27

0:57:08

75-79

5:19:18

6:13:38

0:54:20

80 >

5:48:01

6:35:18

0:47:18

 

Table 2. 2008 average winning times

Figure5.jpg

Figure 3. Differential Between Winning Times vs. Boston Standards

 

It is clear that the differentials between actual men’s and women’s overall and age division front runners in U.S. marathons are almost exactly 30 minutes up to age 50. Beyond age 50, the data depicts a falloff of women’s performances relative to men, which is also evident from BQ rates shown in Figure 2 and discussed below.

 

Age Division Considerations

 

Figure 2 reveals a couple of interesting characteristics of women vs. men marathoners. Figure 1 suggests that women consistently BQ at a slightly lower rate than men. However, Figures 2 and 3 indicate that their relative performances are a function of age. In the 35-39 and 40-44 age ranges, the percentages of women BQ performances are noticeably higher than those of men and their winning times are consistently about 30-minutes slower than those of men. However, that changes as age approaches 50. From there onwards, male BQ rates become considerably higher than those of women and the differential between winning times increases substantially.

 

I think the trends that these figures suggest can be explained by a look back at history. It has been 25-30 years since women began to show interest in participating in road racing and, particularly, marathons in any significant numbers. Over 24 years ago Joan Benoit Samuelson won the inaugural women’s Olympic Marathon in Los Angeles in a dramatic fashion that received wide publicity and fired the imagination of Americans, especially American girls and women. Twenty eight years ago the U.S. Government established policies and guidelines for applying and administering Title IX, which is intended to equalize opportunities for the sexes to participate in sports and other activities at the school level.

 

The prime formative and development years of women who are now in their 50’s and older preceded these historical developments. They have probably carried some old societal inhibitions with them into their later years, as well. As a consequence, they find themselves less able or willing to train and attempt to perform on the same level as their male counterparts in their middle-to-late ages.

 

In contrast, women who are now in their 30-40’s, which are prime years for marathoning, have benefited from the developments of the 1980’s and are better prepared and more willing to perform athletically on a level comparably to that of men ... indeed, to excel.

 

An alternative theory might be that biological changes that occur during women’s midlife years have a negative affect on their performance ability.

 

It would be interesting to take another look at this picture in another 20-30 years to see if women dominate the BQ rates across the board. It just might become necessary to raise the bar in women’s BQ standards so that men can maintain parity, if not supremacy. J

 

The plunge in women’s BQ rate in the 75-79 age division and subsequent surge in the 80+ category, as well as the drop in the men’s 80+ category, are probably anomalies resulting from very small numbers. For instance, in the last three years there were five BQ performances out of only 58 finishes in all 227 races in the women’s 80+ division.

 

Figure 2 does clearly suggest that the BQ standards for older age divisions are “easier” than those for the men’s and women’s open division of 18-34. I wondered if that might be influenced by younger runners in the open division whose physiological and athletic development might not be as advanced as their older counterparts in the division. Were they “dragging the division down”, so to speak? So, I collected 2008 open division data in sub-divisions of 18-24, 25-29 and 30-34 and found little difference among them. Table 3 presents those results.

 

Age Division

Men BQ %

Women BQ %

18-24

8.2%

8.4%

25-29

8.7%

8.5%

30-34

7.8%

8.3%

 

 

 

Table 3. 2006 BQ Percentages within the open age division

 

As Figure 2 indicates, with the exception of the 45-49 age division, the rate of increase of the percent of male BQ performances is fairly consistent throughout the age divisions until the oldest divisions of 70-74 and beyond where the anomalies of very small representation come into play. (The peak in the 45-49 division is probably a result of the 10-minute increase in the standard from the previous 40-44 division and only a 5-minute separation from the 50-54 division that follows it.) This clearly indicates that runners lose physiological ability with age at a lesser rate than the increase in BQ standards.

 

While this trend of the older male divisions appearing to be easier might seem to some to be a bit “unfair”, the numbers are comparatively small, especially from age 55 and beyond. For instance, the men’s 65-69 age division, which recorded the highest BQ rate, had less than 1/10 the number of BQ finishers as did each of the 35-39 age division. And it does compensate for over 100 years of much tougher standards for the older divisions compared to the younger divisions before the changes that occurred in 2003 when the standards for the older divisions were relaxed. For instance, the current men’s 60-64 standard of 4:00 was 3:40 prior to 2003 … and it was 3:30 prior to 1989. Most of the masters divisions were underrepresented at Boston for a very long time. It’s about time the pendulum swung the other way in favor of us old guys. J

 

Conclusions

 

I think it is pretty clear that the current BQ standards accomplish their intended purposes. They present a challenge to runners of both genders and all ages who aspire to running Boston while also producing demographic distribution in the Boston Marathon that is reasonably representative of the American running community.

 

The split between the average number of men and women finishers in the 2006-08 Boston Marathons was 60.4%/39.6% compared to Marathonguide.com’s estimate of 60.5%/39.5% and Running USA’s estimate of 60/40% for all American marathon finishers.

 

Age distribution is weighted somewhat toward the older runner … 55.6% of Boston finishers were masters runners compared to Running USA’s estimate of 46% of total American marathon finishers (and 46.2% in this 3-year project). If any fine tuning of Boston’s standards could be justified it would be to tighten the standards of masters men and relax them a bit for masters women.

 

Appendices

 

For anyone who is interested in the gender and age division specifics of individual races included in this project … or who just enjoy pouring over spreadsheets filled with numbers J ... the following views of the data base for individual races and all age divisions are available:

 

Men’s 3-year age division averages sorted alphabetically

 

Women’s 3-year age division averages sorted alphabetically

 

2008 winning times

 

Combined 2008 age division data sorted alphabetically by race name

 

Combined 2007 age division data sorted alphabetically by race name

 

Combined 2006 age division data sorted alphabetically by race name

 

Note: All of the lists contain the most commonly used race names of “Los Angeles” and “San Francisco” for those marathons instead of the “official” names of “City of Los Angeles Marathon” and “The San Francisco Marathon”. I did that because I kept forgetting to look for Los Angeles under “C” and San Francisco under “T” in Marathonguide.com’s alphabetical listing. J