Zeke has a nice post up on marathon pacing. I was going to leave a comment there, but why not get a post up for myself? Like a lot of things in life, it's all so simple in theory, but difficult to get right in practice. Perusing any marathon results will pretty quickly show you that if you slow down by less than a minute in the second half you're keeping rare company and should be relatively pleased with your execution of the race. Even splits and negative splits are special, special things.
As I say, all so simple in theory. The first step in theory and practice is to know that the marathon is a cruel thing, that relatively minor mistakes you can recover from in shorter races (even up to 30km races) can have really big effects on your marathon; that if you run race enough marathons, some of them will be bad. If you are well trained and "just" run a marathon it's a long training run, you'll get "pleasantly tired" (to quote Arthur Lydiard) and you'll be out doing speedwork soon there after. Put it on the line for sub 3:10, 3:00, 2:48:47, 2:37:20, 2:30:00, 2:22:00 or whatever you're going for, and you'll come a cropper one of these days. Just the way it is. Just the law of averages. Though I know that is no consolation on the day it happens.
Hitting the wall is absolutely no fun, but most of the time you should finish. That would be my rule two for setting a goal. "Always finish" is a solid rule for any length of racing, since once you've pulled out once it becomes so much easier to do it again the next time it gets hard (a DNF for acute injuries is totally acceptable, but how often does that happen?). But I have a sneaking suspicion that the effect is worse for marathoning, simply because there's just so many more moments in the race when you might want to drop out.
At this point, there's a huge gap between what you'll end up with and what your ultimate goal was, so my rule three is to have a succession of intermediate goals in between the "dream day when I find I'm fitter than I knew" and the death shuffle from 18 miles. Personally, I now set that first goal beyond just finishing as a Boston qualifier, since a Boston qualifier is still 5 minutes over my personal worst, but avoiding a personal worst time is probably a good part of any cascade of marathon time goals. Your specific aims will vary, obviously and if you get to this point of searching for a reasonable goal to keep going you'll probably be changing them mile by mile ... That's been my experience.
In between finishing, avoiding personal worsts, and the actual realistic goal for a race there may be many minutes, so rule three point five is to find other intermediate goals, so if you went for 2:45 and ended up with 2:50:05 you can still find something to appreciate in your performance.
This all gets us to the pointy end of establishing a realistic goal. I go with three rules of thumb here
On the last point, I think it's pretty well established that lactate threshold and aerobic threshold, and pretty much every other relevant physiological variable, will do just that, vary, from day to day. You can do a lot to reduce this variability, and increase the predictability of what you can do with appropriate time trials and other workouts, tapering, etc, etc ... But you can't predict the inevitable random variation. If you go in assuming you can run 6 minutes over 2 x 1/2 marathon pace, and that day the best you could have done if you'd paced perfectly was 8 minutes over ... there's four minutes you're giving up in the second half. There's your 1:24/1:29 splits.
Of course, the disaster that's going to occur is not necessarily obvious at half-way. It will still feel easy, because running 2 minutes over half-marathon race pace (for a half) feels surprisingly easy, 3-5 minutes over, that takes real, real restraint when you're feeling good.
Until this year I never quite understood how people might go into a marathon really uncertain of what they might end up running. Now I do. Back in April and May this year I was confident I was going to take a decent shot at the low 2:40s in Chicago. Running 62 for 10 miles in early April feeling like "this is marathon pace," does that for you, especially when you back it up a few weeks later with a 41 minute 7 mile tempo run. Going rapidly backwards for five weeks as your iron levels fall erases that confidence. Getting on the iron supplements, and doing a good month of base influenced by the Speed with Endurance schedules gave me some of that confidence back. Two weeks into the 12 week specific build for Chicago I was metaphorically patting myself on the back for getting over the iron problem, and doing two weeks of workouts that I think indicated 2:46-2:48 could be realistic. Then I strained my glute tripping over a rock and lost several workouts in the next month. I can't speak highly enough of active release therapy that with 2 weeks to go I think 2:50 is a realistic goal. This implies trying to hit the half-way at Chicago in 1:25.
If I really am in shape to run quicker then I'll be able to negative split. That's the part of me that believes the great spring and mid-summer training must still be there in some way. Negative splits still hurt, but they hurt a lot better than running 1:22/1:35 ever will. And if, after this up-and-down six months I'm really in 2:53 shape, 1:25/1:28 is not a huge explosion. 18 days to go.
Posted by eroberts at October 4, 2006 03:14 PM | TrackBackNice post. One bit of advice I have seen that I agree with in regard to dropping out of marathons is with regard to running a better race a relatively short time later. Because it can take a long time to recover from a full marathon, and because it takes a long time to train up for one properly, it can be useful to drop out before the twenty mile mark if it just isn't your day, with the idea that several weeks later, it may be.
I wouldn't recommend it to everyone, but to the experienced runner with very specific goals. I know elites do this from time to time. Obviously they are protecting their incomes and livelihoods. Still, it can be a good strategy in some non-injury situations.
Good luck at Chicago!
Posted by: Eric at October 5, 2006 05:56 AMVery nice post. I particularly like the 3 rules of thumb.
Is that active release therapy link correct? I didn't see anything there that pertained to ART.
"Negative splits still hurt, but they hurt a lot better..."
Mmmm, a better kind of hurt. I can't wait.
Posted by: zeke at October 5, 2006 07:54 AMThanks for the comments.
Eric, you're right that sometimes dropping out and trying again a short while later can be the right thing to do. Given my up-and-down year and the feeling things might be turning round I've actually thought about that option if Chicago is going poorly for whatever reason at 20 miles. So, totally in agreement that sometimes it could be the thing to do. However, my sense is that the elites that do this are the just-below-really-world-class elites who rely more on race earnings for their income than someone like Meb or Deena who clearly don't have to race frequently to pay their utilities. Since recreational hacks also don't have their payday riding on frequent race days it may be better to copy what the truly elite do, not the people who are doing it for a payday.
Zeke, yeah, the ART link was to a map/photo of where I strained the muscle. Oops.
The nearly even and negative split marathons I've done have hurt, but it's the hurt of a long tempo run with the satisfaction of taming the challenge of a marathon. That's way, way better than the death shuffle to the finish.
Posted by: Evan at October 5, 2006 08:24 AMI really enjoyed this post and I hope you have a great race in Chicago. I'm torn on the negative split theory, and probably the only way to really make an informed decision as to whether it leads to a great race is doing it myself sometime. I've always been in Pfitzinger's camp with regard to optimum pacing strategy. He seems really into running a very even effort for the full distance, but that condtitions, cardiac drift and fatigue generally lead to a one to two minute positive split as the pressure mounts in the second half. He thinks if you negative split too much, you probably didin't run the first half fast enough.
Oh, I almost forgot the obligatory "this won't work for everyone". For Downeast Andrew, a negative split worked wonders. He figured this out through trial and error and planned for it in his training runs, and it paid off big time.
Posted by: Mike at October 6, 2006 07:50 AM