Boston analysis

Also, I can't find it at the moment, but Ken also explains that his RTB analysis does not include the kind of statistical outliers you are worried about, malmo (2-sided results etc.).

Again, it doesn't say what the algorithm is. Brett, you seem to know a little bit about it, why don't you find out the details? thank you.

I asked Ken to come on and explain his methods after he finishes his analysis of this year's race. Hopefully later today.

Thanks, that would be very cool.

Brett in Tokyo wrote:

According to Ken Young of the ARRS' statistical analysis of the Boston elite results from 1973-2010, the men's race has averaged 0.4 sec/km, or a total of 17 sec, slow:

http://arrs.net/HP_BosMa.htm

In the 1994 tailwind year it was 1.7 sec/km fast, 1:12 fast, for men. In hot 2007 it was 5.8 sec/km, 4:05, slow.

Looking forward to his analysis of this year's race.

In 2007 it was cold with a strong headwind much of the way from the Nor'easter.

Every opinion can be respectable, but not the opinion that the advantage could be of more than 3 minutes. You start with the idea that Boston can penalize top runners of about 2:00. But we have many examples this is not true.

Where did Joan Benoit run 2:22, in a not windy day ?

Where did Gelindo Bordin beated the Italian Record after winning Olympic, running in 2:08:19 without wind ?

Where athletes as Benjamin Maiyo had their PB, in not very good conditions ?

I had athletes winning Boston 3 times (Elijah Lagat in 2000, Rita Jeptoo in 2006 and Selina Kosgei in 2009), and always their results was influenced by the tactic of the race, not from the course. In 2006 Rita ran the last 7,195m faster than all the first top men, because the race before was not very fast, and in any case she improved her PB.

What I want to say is that an hilly course can have different impact for different athletes, but in some case doesn't penalize the final time. Athletes as Geoffrey (very high agility) or Moses (very high eccentric strength) don't suffer this kind of courses.

Leonard Komon, for example, improved last year the WR of 15 km in a race called "Seven Hills", that of course is very hilly. If there are not long hills to climb, but short and hard, athletes with good strength can attack at high speed, and the next downhill can be run at high speed with good relaxation.

Women suffer more the conditions of Boston, because have less strength. So, personally I'm not sure that Boston can penalize of 2' the athletes.

Second point : Geoffrey ran last year two marathons in 2:04:55 and 2:05:10, and currently is very much stronger, expecially under the mental point of view. So, there is no reason he could not run, at the moment, in 2:04 in the same races of last year. Why somebody continues to think at his time like a 2:06 in any other race ?

I explain Moses was in a shape for 2:04. The only doubt was his reaction in the first marathon. Since he was amazing under this point of view, his result is totally normal, looking the resultances of his training.

I'm sure that the same athletes in a fast marathon with the same weather conditions (but without wind) could run under the WR of Gebre in Berlin. So, my personal evaluation of the advantage is no more than 40 seconds, because we know the real technical value of the athletes, and cant be surprised from their results. If we had athletes as Duncan Kibet able running 2:04:27, there is no reason because somebody can have doubts about the possibility for Geoffrey Mutai to run one minute faster. About Moses, I advice anybody to wait and to see the future. But never happened in my life rhat a my athlete could end his career with a PB in his first marathon (I think the best must come after 7-8 marathons, if the training is correct), so I think he can have good room of improvement, because I know what he did, but I know what he still didn't do, that in my training phylosophy is important for reaching the personal best.

Note: I posted this post without seeing Renato's post immediately above

I respect Renato tremendously and agree with him that the marathon is changing. This statement is so true, "THE MARATHON IS CHANGING BECAUSE IT IS NOT LONGER WHAT YOU MOVE UP TO."

However that has been the case for a few years now. Look at the tremendous London field. My question for Renato would be with the above statement, "How come in London no one ran 2:03?" With 0 wind then do you think London is slower than Boston?

If we had huge fields with Haile G in Berlin and they all went out on WR pace it could be interesting to see what happened. I don't view Haile G as the best marathoner in the planet so I acknowledge people can run in the 2:03s.

Having said that perhaps we are debating slightly different things. Renato may be comparing Moses' time to what he thinks Moses could run on a flat course. Statisticians will be comparing Moses time to what he could run on a normal Boston course. If Boston with no wind is slower than a flat course then that means the wind played a factor.

(If Boston with no wind actually is faster than a flat course then that is a different issue but still means the marks are aided).

With the 2:03:02 out there, people may attempt to run the marathon differently. It changes the sport in many ways.

However, on a course where 1 person EVER had run under 2:07 until Monday and all of a sudden 6 guys do with many significantly under the mark and there happens to be a 20mph wind behind the runners back and people like JK were predicting 2:03 before hand then I'm going to say the marks are aided in some way. Especially when after a 1:02 opening half marathon, they ran close to 14:00 for 5k, one of the fastest 5ks EVER run in a marathon up Heartbreak hill. We will have more on that tomorrow.

Weldon, people aren't suddenly running the marathon differently. 25 years ago Steve Jones took it out in 61:40 and he didn't have the benefit of a 20 mile an hour wind at his back.

Everyone needs to let go of the hyperbole. IT'S THE WIND, STUPID. You aren't suddenly going to see people running 2:02s and 2:01s. Let them break 2:05 first and then 2:04 and hope that if everything goes perfect one of them might run 2:03:30.

Malmo, Steve Jones had the same behavior of Geoffrey Mutai in Boston. With a small difference : was one minute slower in 10000m. This means that, if they have the same coefficient of sèpecific endurance, at the end there are more than 4 minutes of difference in their performance. And this is exactly what happened.

doubter wrote:

Plus - it's going to be awesome to see Hall crank out another 2:04 in Houston! And if he just improves a little between now and then I can only imagine the time he will be running!

This statement reflects ignorance, both with regard to the Houston course, and the potential tactics it will likely engender.

Thank God malmo is calling both Renato and Weldon stupid. I feel genuinely honored to be a called a delusional idiot by comparison.

People need to understand that human factors, variability in conditions and topography are not allowed in the "Malmo Model". Only single-factor explanations, enforced by post and thread deletions, will be tolerated!

Arithmetic and a cudgel will trump calculus and an open mind every time!

"IT'S THE WIND STUPID" is an American idiom, a twist on a political campaingn slogan, Einstein.

You not only have a closed mind, you are like a little kid with his eyes shut and his hands covering his ears, spinnning around and yelling, "I'm not listening. I'm not listening."

You don't have to invent complicated theories why Boston was so fast this year.

Renato Canova wrote:

Malmo, Steve Jones had the same behavior of Geoffrey Mutai in Boston. With a small difference : was one minute slower in 10000m. This means that, if they have the same coefficient of sèpecific endurance, at the end there are more than 4 minutes of difference in their performance. And this is exactly what happened.

Renato. You are looking at the mouse in the living room (OK a small puppy). I am looking at the elephant in the living room - the wind.

Hopefully my use of English idioms isn't confusing you? ;)

difference between Canova and Malmo

Canova = nice guy, cool dude

Malmo = cool dude but can sometimes be rude/arrogant

Renato

Thank you for your comments and I agree that the athletes in the marathon are changing tactics and limits dramatically. If you look back 10 years ago how many men in their prime moved up to the marathon that could aslo break 27:20 in a track 10000m you would have a short list. Now it's common to see numerous athletes in their prime capable of running 27:20 for 10000m on the track and several that could go sub 27 minutes.

While I think the wind did aid the runners alittle at Boston it WASN'T significant, the amazing 10000m and 1/2 marathon athletes moving to the marathon will continue to knock down boundaries and redefine the marathon.

Renato Canova wrote:

Every opinion can be respectable, but not the opinion that the advantage could be of more than 3 minutes. You start with the idea that Boston can penalize top runners of about 2:00. But we have many examples this is not true.

Where did Joan Benoit run 2:22, in a not windy day ?

I disagree Renato. 3-4 minutes is a good estimate. It put Mutai at 2:06-2:07 at Boston on a normal day and 2:04-2:05 at Berlin, Amsterdam, or Rotterdam.

Joanie best loop marathon was 2:26:00 in Sept of 1982, Nike OTC Eugene Oregon. Flat and fast course. The next Spring, at Boston, she ran 2:22:43, with a strong wind at her back.

Here are some Boston Marathon historical stats for you Renato.

1975 1) 2:09:55, 22) 2:19:58, 114) 2:30:00, 842) 2:59:15

1976 1) 2:20:19, 18) 2:30:31, 77) 2:38:55, 416) 3:00:01

1977 1) 2:14:46, 8) 2:19:48, 58) 2:29:56, 230) 2:41:44, 932) 3:02:24

1978 1) 2:10:13, 33) 2:20:03, 161, 2:30:04, 2430) 3:07:25

1979 1) 2:09:27, 53) 2:19:53, 131) 2:25:01, 3032) 3:00:24

1980 1) 2:12:11, 15) 2:19:46, 475) 2:42:24, 1232) 2:55:45

1981 1) 2:09:26, 32) 2:16:34, 326) 2:30:50, 527) 2:35:32

1982 1) 2:08:52, 23) 2:20:00, 156) 2:29:59, 2027) 3:00:00

1983 1) 2:09:00, 83) 2:19:51, 303) 2:29:41, 789) 2:40:00, 2588) 3:00:00

1984 1) 2:10:34, 25) 2:19:52, 194) 2:29:59, 2121) 2:58:35

1985 1) 2:14:05, 35) 2:31:15, 197) 2:42:10, 901) 3:00:00

1986 1) 2:07:51, 17) 2:19:19, 74) 2:30:33, 1227) 3:00:34

1987 1) 2:11:50, 19) 2:19:19, 71) 2:29:49, 2304) 3:08:46

1988 1) 2:08:43, 28) 2:18:17, 110) 2:30:05, 1558) 3:00:00

1983 Boston

1. G. Meyer, MA 2:09:00

2. R. Tabb, OR 2:09:31

3. B. Durden, GA 2:09:57

4. E. Mendoza, AZ 2:10:06

5. C. Bunyan, IL 2:10:54

6. D. Edge, Canada 2:11:03

7. M. Layman, WA 2:11:24

8. D. Schlesinger, MA 2:11:36

9. J. Wells, OR 2:11:42

10. B. Rodgers, MA 2:11:58

11. D. Hinz, MI 2:12:05

13. D. MacDonald, CA 2:12:49

14. B. Coates, PA 2:13:02

15. D. Gordon, OR 2:13:11

16. H. Schulz, CA 2:13:37

17. D. Rinde, CA 2:13:48

18. R. Sayre, OR 2:13:49

19. G. Vega, NY 2:14:01

20. K. McCarey, CA 2:14:09

21. T. Fleming, NJ 2:14:14

22. C. Law, NC 2:14:21

23. D. Matthews, GA 2:14:46

24. T. Shibutani, Japan 2:15:12

25. D. Patterson, PA 2:15:20

26. J. Dimick, VT 2:15:23

27. B. Allen, CO 2:15:36

28. E. Castellanou, Venezuela 2:15:40

29. L. Barthlow, MA 2:15:43

30. M. Mesler, MI 2:15:44

31. M. Pinocci, CA 2:15:50

32. P. Cummings, UT 2:16:05

33. J. Hope, CA 2:16:10

34. J. Anderson, OR 2:16:19

35. R. Serna, CA 2:16:26

36. J. Thomas, MA 2:16:28

37. F. Stonecipher, MO 2:16:35

38. A. Azocar, Venezuela 2:16:38

39. S. Molnar, PA 2:16:41

40. M. Patterson, PA 2:16:45

41. B. Hensley, CT 2:16:50

42. J. Metcalf, OK 2:17:02

43. R. Hagemann, TX 2:17:05

44. F. Torneden, KS 2:17:08

45. A. Cendejas, CA 2:17:15

46. P. Friedman, NY 2:17:17

47. D. Ryberg, NC 2:17:18

48. G. Fanelli, PA 2:17:29

49. A. Treffinger, PA 2:17:30

50. R. Fritzke, CA 2:17:31

...

56. G. Tuttle, CA 2:17:40

...

77. T. Maxwell, CA 2:19:13

...

83. T. Ratcliffe, MA 2:19:51

Renato Canova wrote:

Malmo, Steve Jones had the same behavior of Geoffrey Mutai in Boston. With a small difference : was one minute slower in 10000m. This means that, if they have the same coefficient of sèpecific endurance, at the end there are more than 4 minutes of difference in their performance. And this is exactly what happened.

Renato. The first 4 miles of Boston were straight down hill, then flat, plus a 20 mile an hour tailwind. That's almost 9 meters/second. Mutai's split was 1:01:57 at the halfway mark. Jones was 1:01:40 when he ran 2:07:13 in Chicago.

If Usain boit ran a 9.45 100m with a 9 m/s tailwind you'd be laughed into exile if you ever claimed it was a legitimate performance.

Last years perfomances were taken into the Worlds Best list, so why in case of such a perfomance there´s made an exception?!

Do oyu think it was a mistake for mutai to go there, if it´s not possible to record there?!

Kevin52 wrote:

In 2007 it was cold with a strong headwind much of the way from the Nor'easter.

Whoops, right you are. What was I thinking of, 2004? At any rate, according to Ken Young's calculations the conditions in 2007 with the Nor'easter were 5.8 sec/km slow, 4:05 altogether.

2004, the hot year, I think, was 6.3 sec/km slow, 4:26 altogether. Along with 1994, the other tailwind years malmo mentioned were as follows:

1983: 3.1 sec/km fast, 2:11 fast.

1975: 4.2 sec/km fast, 2:57 fast.

1976, another hot year, was 6.9 sec/km slow, 4:51 slow.

1981 looks like another tailwind year, 3.8 sec/km fast, a 2:40 advantage. Do you know if that was the case, malmo?

In Ken's absence it's worth pointing out that his threshold of rejection for an 'aided' performance is 5.0 sec/km fast, 3:11 total for a marathon.