IM Run Assessment

Running has become my strongest of the three triathlon disciplines. I started running after moving to Chicago for law school and seeing the excitement surrounding the 2004 Chicago marathon. I decided I was going to do the race in 2005. Up to that point, I didn't run much. I'd never run longer than 5 miles and probably had never run more than 12 miles in a week. I had run in only one race, a tiny 5k in my hometown. In the five years I've been running, my times have continuously improved to the point that I'd consider myself a respectable runner. Here's a race history of the marathons I've done with first and second half splits in parenthesis:

2005 Chicago marathon - 3:27 (1:41/1:46)
2007 IMCdA - 3:28 (I can't find splits, but I recall something like 1:30/2:00)
2007 Chicago marathon - 3:03 (1:23/1:40)
2008 IMLP - 3:17 (1:29/1:48)
2008 Chicago marathon - 2:48 (1:18/1:30)
2009 IMLP - 3:19 (1:34/1:44)
2009 IMLou - 2:54 (1:21/1:32)

The most obvious trend is I run a fast first half and a slow second half. My nearest to even pacing out of my six most recent marathons is losing 10 minutes in the second half, meaning I've slowed by at least 40 seconds per mile from the first half to the second half. Now, I'm not of the opinion that negative splitting (i.e., running the second half faster than the first half) is the way to go, especially for an IM marathon. I think running the second half about 5 minutes slower than the first half would be a sign of good pacing that still allows me to push myself close to my limits. Such a split would mean losing about 20 seconds per mile from the first half to the second half.

My Louisville marathon is a great example bad pacing. Here are my splits:
FIRST RUN SEGMENT 3.38 mi. (19:17) 5:42/mile
SECOND RUN SEGMENT 8.25 mi. (30:26) 6:14/mile
THIRD RUN SEGMENT 13.1 mi. (31:37) 6:31/mile
FOURTH RUN SEGMENT 15.33 mi. (15:16) 6:50/mile
FIFTH RUN SEGMENT 20.19 mi. (33:55) 6:58/mile
SIXTH RUN SEGMENT 25.04 mi. (34:59) 7:12/mile
FINAL RUN SEGMENT 26.2 mi. (8:33) 7:22/mile
TOTAL RUN 26.2 mi. (2:54:03) 6:38/mile

I started out running a 17:15 first 5k, which would be a new 5k PR. Starting an IM marathon with a 5k PR is not good pacing! (In my defense it's been a long time since I've done a 5k and I could probably run a fair bit faster than 17:15 now.) I continued to run a 1:21 half marathon, which is about 4 minutes off my stand alone half marathon PR.

Why did I start so fast? First, my starting pace felt easy. I primarily pace myself based on feel. Under most conditions I think that's the best way to pace one's self because feel takes all factors (heat, hydration, course profile, etc.) into account. One drawback of running by feel is that adrenaline and the excitement of starting an IM run are difficult to account for. Those feelings are unique to racing and as a result it's difficult to get enough experience taking adrenaline and excitement into consideration when running by feel. Running by feel, after all, is possible as a result of accumulating experience doing runs at different paces and under different conditions.

However, I know that adrenaline will affect my pace in an IM. I know to be very cautious the first feel miles of an IM, as being cautious at this time allows me to wait until the adrenaline dies off 3 or 4 miles into the run before switching to running purely based on perceived exertion. So the second way I pace myself during the first few miles of an IM run is to look at my HR monitor frequently. I try to cap my HR at 160 bpm for the initial three miles.

Here is my HR data from Louisville for miles 3-26.2:


I didn't start collecting data until about mile 3 because I always screw up working my GPS unit during races when my mind is thinking about other things. Even though I didn't start recording data until mile 3, I could still view my HR and did so every couple hundred feet during the first three miles. Despite the bad initial pacing, my HR was at or below 160 bpm every time I looked at it during the first 3 miles. While running well under 6:00/mile at 160 bpm seems impossible for me, I have photographic proof of me inspecting my HR coming off the bridge about 1 mile into the run. Since I ran too fast even obeying my HR cap, I need to alter my pacing system. I've got some ideas below for preventing running that fast in the future.

Regarding the data I do have, I began pacing myself by solely based on feel without looking at my HR around the time I started recording data. My pacing based on feel looks pretty good, as my HR remains at 160 +/- 5 bpm until after mile 15 (mile 12 on the graph). I'm not sure what happened at mile 15, but I had trouble sustaining my HR. There's a noticeable drop to 155 +/- 5 bpm until about mile 24 (mile 21 on the graph) at which point the adrenaline of being near the finish likely kicked in. At IMLP I also had a depressed HR, and there I was confident it was because I wasn't getting much fuel in during the run. Here at Louisville I'm not so sure. Maybe I lack enough endurance for the effort I'd put in up until that time. That would make sense given my too fast start to the run. Maybe with better pacing in the future I'll be able to maintain a consistent HR throughout the run. Still, the decreased HR is basically a mystery to me.

Since I assert that my HR didn't go above my 160 bpm cap during the first three miles and I still ran too fast, I need to improve my pacing. One way to do this in the future will be to set a pace cap. Assuming my training goes well, I feel great off the bike, and the conditions aren't too harsh (a lot of assumptions for an IM) a 6:15/mile pace cap for the first three miles would be appropriate. Holding this pace for 13.1 miles would give me a 1:22 first half of the marathon, basically what I ran at Louisville but paced more evenly. Hopefully I'd be able to stay within my goal of only losing 20 sec/mile for the second half for a 1:26 second half and 2:48 total marathon time. Fast, yes, but hopefully I can cut off 6 minutes next year with improved fitness and better pacing.

Since I know the importance and difficulty of good pacing early on in an IM run, why didn't I use my actual speed to pace myself this year? After all, I run with GPS and thus my actual speed is readily available. There are a few reasons. One reason is that determining a pace cap is difficult. The pace cap could be too ambitious, or it could potentially be limiting. Now that I've run a 2:54, a 6:15/mile cap would still give me a chance at being the fastest runner at many Ironmans, so it is very unlikely to be limiting. Also, that cap is slower than I started at Louisville, so I don't think it's too ambitious if my race goes according to plan during the swim and bike (again, a huge assumption in an IM).

Another reason I didn't pace myself based on my actual speed - and this was the most important reason for me during IMLou - is that I simply did not want to know my actual speed. I was worried it would be slow and I'd be discouraged. I was not thrilled with my race up to starting the run (a horrible 1:12 swim and an okay 5:07 bike). Seeing slow run times would have been extremely discouraging and might have killed my drive and determination for the race. Ignorance is bliss, I reasoned.

In summary, here's my plan for my next IM -
Miles 0-3: pace by perceived exertion, but cap effort based on HR (~160 bpm) and pace (6:15/mile if I feel great)
Miles 3-26.2: pace by perceived exertion