Posted by: David Stewart | December 14, 2009

CIM by the numbers

Is the runner in your life a data geek? Read on.

I ran the California International Marathon (CIM) a week ago, and I have been trying to understand the wealth of data available to me.

My coach has yet to return from Australia where he competed in the Western Australia Ironman, so I have not had the chance to get his analysis of my race results. But as I picked at the details, there were some things worth learning.

  • CIM’s placement of timing mats on the course is messed up, particularly at the 20 mile mark. This makes for some bad surprises in online tracking.
  • I seem to be unable to walk and chew gum at the same time. More specifically, when I am sucking down a gel, my pace seems to get really slow.
  • My Garmin’s heart rate monitor is pretty much useless in the opening mile or two of a race.
  • I am a huge victim of addled brain syndrome in the last few miles of the marathon. My questionable math skills are totally useless when suffering through those last bits of the race.

Tracking errors

In 2009, a friend could track your progress live on the internet as you moved along the course, thanks to chip timing and well-placed sensor mats along the course.

Calculated Reported
6 0:47:17 0:46:57


The glaring bogy here is at the 20 mile mark. My Garmin reported a time which is four and a half minutes earlier than was reported on the website. This is the difference between a 7:55 minute per mile pace (comfortable for meeting the 8:00 minute pace) and an 8:09 average (which means disaster).

In fact, I had a friend who was tracking me live on the internet and sent me a direct message on twitter:

oh no – you are getting slower… I’ve been tracking all morning – you started so fast… run, run, run

This was because he saw my pace go from 7:50 to 7:55 to 8:09 at the 20 mile mark. How in the world was I able to make up so much time, and finish with a 7:58 pace?

Problem? CIM doesn’t put the 20 mile timing mat at all close to the 20 mile mark.

In fact, they have a big archway at “20 miles” which is decorated as a wall, and is hung with advertizing. This is to commemorate the fact that a lot of runners get to about 20 miles into the race and feel like their legs are made of concrete. It’s like running through syrup.

In this year’s race, the “wall” was a good 3 or 4 tenths of a mile past the actual 20 mile mark. Of course, this can make your internet friends question whether you are fading.

Sucking GU makes Dave slow

My coach wanted me to consume about 150 calories per hour, and wanted me to take my customary GU gel during my warm-up before the race, and then every our on the :50 minute mark.

But I notice that the miles I was sucking down the GU are some of the slowest in the race. In fact, the miles at :50, 1:50 and 2:50 are on average nearly 30 seconds per mile slower than the other miles of the race, if I throw out the first and last miles.

Average non-GU miles: 07:52
Average non-GU miles: 08:19

This was a pattern I noticed on some of my long runs too. I just don’t seem able to coordinate my gels and running. I’ll need to talk this over with my coach and see if I should just walk while I am GU-ing to keep my pace up.

You make my heart race

In order to avoid being in oxygen debt, my coach wanted me to do a well structure warm-up before the race and to avoid going too fast in the opening couple of miles.

This meant keeping my heart rate below 150 beats per minute in the opening miles.

But as I started the race, my heart rate comically soared into the 170s, no matter how slow I went. Was my heart really racing so fast?

I decided that I must be picking up heart monitor signals from other runners who were close packed with me. In fact, my Garmin reports that my heart rate maxed out at 179 beats per minute for the first two miles of the race, which is well above my anaerobic threshold of 166 BPM.

Of course, this also made it impossible to figure out if my heart rate was below 150 BPM, so I quickly gave up on that piece of advice and kicked into my goal race pace, which I hit in the third mile.

Marathons make you smarter … NOT!

I remember in the final two miles of the race feeling a little too good. I checked my watch when I hit mile 24 and it looked like I had 19 minutes to run the last two miles and still meet my goal of 3:30. Hey, I can slack off to a 10 minute per mile pace and still be OK!

Well, it’s a good thing that I only relaxed a little, to about an 8:10 pace. When I got to mile 26, my watch showed about 3:27:30.

Holy cow! I forgot about the last 385 yards!

Of course the official marathon distance is 26.2 miles, not 26. So I had about 150 seconds to run the last 385 yards.

Needless to say, I “sprinted” the best that I could, and ended up finishing with over a minute to spare. But this lack of math ability nearly cost me my qualifying time!

The rest of the numbers

As I mentioned before, CIM is a highly competitive race. When I looked at my results in comparison with the rest of the field, it’s sobering:

Top third of men ages 45 to 49
Top third of all men
Top fifth of all finishers (nearly)

Here are the rest of the miles. I’ll be interested in seeing what these numbers mean when I sit down with my coach.

Mile Avg Pace Avg HR Comments
1 08:09.09 169 Heart rate too high, must be proximity to other runners’ heart rate monitors
2 08:03.68 162
3 07:42.75 154 Smooth sailing
4 07:43.44 155
5 07:48.09 156
6 07:49.96 157
7 07:51.86 156
8 07:59.20 156 Some slowing / higher heart rate. Taking my first sports gel.
9 07:54.94 158
10 07:48.42 156 Smooth sailing
11 07:45.42 155
12 07:54.92 155
13 07:41.76 156
14 07:54.69 156
15 08:28.76 156 Significantly slower. Long uphill, headwind and taking gel
16 08:05.38 155 Getting back on track
17 08:01.61 156
18 07:53.97 156 Smooth sailing again, but heart rate creeping up
19 07:53.95 157
20 07:58.36 156
21 07:58.47 157
22 08:27.89 155 Significantly slower, taking a gel again. Is this a pattern?
23 08:00.03 158
24 08:08.99 157 Took a caffienated gel to make last miles easier
25 08:08.99 159 Watch reads 03:11. It looks like I have almost 20 minutes to run two miles and can slow down.
26 08:10.64 160
26.2 07:59.00 165 Last two tenths I realize I need to sprint to get done in time.


  1. Thanks for posting all the data. This is extremely interesting. Looks like you didn’t run negative splits which surprised me… and the 20 mile thing – that explains a lot.

    I love the HR data – even though you slowed down on miles 25&26 your HR actually went UP!

    BTW: the bad readings at the beginning have little to do with other runners – that connection is authenticated with a random key (seriously). It almost always is lack of moisture on the contacts – you weren’t sweating, yet.

  2. Hmmm – the bad HR readings due to lack of sweat. It sounds feasible. It was pretty stinkin’ cold to start out, so I suppose it’s possible. I’ll have to experiment.

    The rising heartrate with the slowdown is typical for me when the wheels are starting to fall off. It’s a darn good thing I didn’t need to run any more that day.

  3. […] results timing at the 20 mile mark is really misleading. As I wrote elsewhere, the intermediate timing mat at the 20 mile mark is set up quite a bit after the actual 20 mile […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: