Wow, first the PC thing and now this, you're on a roll. I smell a new expose' column!

There are too many accepted practices floating around in endurance sport that are thought to be the conventional wisdom and the "right thing to do", yet all the evidence suggests otherwise.

Hello,

When you look at WR ie athletes who seem to be the best then you see neg or even splits at every distance up to marathon most of the time. Granted their may be a differnence between elite and normal but i doubt it. My question is how many athletes have you advised and who neg split? One thing with the neg split pacing strategy is you can't miss by much. If your aiming for 3:30 and do the first half in 1:45 you are either going to hit 3:30 or die. If you die you weren't in shape for that time. If you go out in 1:40 you are setting yourself up for a slower second half and hoping it won't be too slow.



Styrrell

"I posted this on the "pacing for 3:45 thread", but given that most of you guys are too studly to read a thread about running at that pace"



Ouch!!! I was following that thread rather closely!!

The elites can negative split or even split. They are on the 100+ mile per week plan and have the fitness to pull it off. They are also operating in the 2:10 zone. Much different than the 2:50 - 5 hour zone that your weekend warrior is playing with. 10 min fade is the weekend warrior/average joe reality. Check the splits from most marathons if you don't believe it.

By the way, to answer your question over 15 runners and I have got all of them to Boston, something that they never dreamed of initially.

I'm headed to IMF in a few months, and even before reading your thread, I had decided on trying to positive split the marathon by 8-10 minutes. I've tried the negative split approach, and it never seems to work for me. Going slightly harder when you're fresher seems logical, but not balls out as you recommended. Just two weeks ago I ran a respectable 13.1 at the end of a 70.3 with splits of roughly 41 minutes, and 45 minutes. Worked well that day; hope it does again on November 5th.

---

I wanna go fast!

How did your effort go at that race? Were you steady effort but decreasing speed? Or was it increasing effort and increasing speed.

I have recently become interested in the idea of revealed preference or manifest truth. In your example the truth of negative splitting is manifest in the results of the population.

A confounding factor would be to decide if trying to positive split is the best idea. Right now everyone tries to negative split or even split, but they end up where they end up. Could it be that "trying" to negative split gives the best results?

If I recall correctly, my splits at Ironman LP this year were 1:47+1:57 for a 3:44. Its not like i even tried to make this happen. Its just the way that things worked out. I just ran as quick as possible in the first half to get me to 13.1 miles with gas in the tank to "race" the second 13.1 miles and even though my perceived exertion went through the roof, my time was 10 min slower.

Here is something to absorb and ponder:

http://www.sportstats.ca/res2002/usarun.htm

The reality of the fade...for everyone at Ironman.

I going to go ahead and call bullsh!t on this thread.

Physiologically the fastest way to run a marathon, or any of distance from 3K up is with even or slightly negative splits. Examine how any of the marathon world records of the last 20 years were run. Ronaldo DeCostas 2:06:05 is probaly the best example - he closed in 61 min after a 64 minute first half. Heck, even examine how the hundreth fastest time on any given year was run and you'll find negative splits.

The logic of your positive split arguement fails on several levels.

First:

"Let's just put it this way. Coaches the world around say you should negative split or even split. Everyone is trying to do his. Look at EVERY marathon and look at what the masses are doing:

THEY ARE POSITIVE SPLITTING"

That the majority of marathoners fail to execute on race day is not proof that positive splitting is in anyway effective in producing fast marathon times. The average marathon finishing time for a major marathon in 2004 was like 4:30 - do you really want to take running advice from someone who runs 4:30?

"If everyone is trying to negative split yet 98% are positive splitting, what do you think this is saying ? It says that the human body is designed to positive split at these distances. The data/real world stats are there to prove it. Anyone who tells you otherwise is full of shit."

Maybe not 98%, but the vast magority of people running marathons are vastly undertrained for actually racing - not just running - 26.2 miles. Hell, Runners(Joggers) World is pushing a 20 miles per week training plan these day - 20 miles per week!!! Show me a runner who has consistantly averaged 70+ mile per week during a competative season in previous years and I'll show you someone ready to START training for a marathon. Negative splits are much harder to run mentally, and require adequate preperation and realistic planning. Training 40mpw to break 3:00 and then failing to negative split after
you went out in 1:30 doesn't mean negative splits don't work - it means you didn't train hard enough to reach your goal.

"Personally I recommend to plan a 10 min fade. This reduces your overall race time and it is amazing how easy it is to run the first half 10 min faster than the second, than running the first half 5 min slower and then having to go into overdrive to run the second half 5 min quicker than the positive split strategy."

Of course it is easier to slow down 10 minutes + in the second half of a marathon - its called dying. For someone who runs 3:00 going 1:20, 1:40 - well they should have stopped at the half as 1:20 for a half is a better performance than 3:00 for the full marathon (1:20 equates to 2:52 on the Purdy Tables and 2:47 on the Hungarian tables). My coach had me do something similar once in college to make a point. He had me open a 10K in 15:15 - I finished in 33:01 a much worse performace than a stand alone 15:15. Two weeks later I opened in 15:57 and closed in 15:30 yielding an 40sec PR of 31:27.

In summary - Put in the training and negative split. Don't negative split because you held back the first half, negative split because your fit enough to run faster the second half after hitting a fast first half.

There are too many accepted practices floating around in endurance sport that are thought to be the conventional wisdom and the "right thing to do", yet all the evidence suggests otherwise.

don't confuse all those "I've never done it, but I will chime-in anyway" posters, Dev, it makes qualifying for Kona much easier when people are slaves to accepted wisdom!

Let's see:

• Coaching is the best way to success
• The only way to get a proper bike fit is to fly somewhere and talk to a guru because geometry is beyond most people
• It is all about the bike
• You can't possibly become a good cyclist without a Computrainer
• Success only comes from >20 hr weeks
• Without supplimental vitamins and suppliments you'd be undernourished
• Mixing protein into your carbo makes a difference in carbo value
• Racing should only be done a few times a year, never near another race, otherwise you will hurt yourself
• Learning how to race is not as good prep for racing as just training
• Swimming 4 hours a week (plus another 3 hours of travel time to the pool) is a good use of your training hours
• Tapers in just about any training book are short enough to get you fresh
• Resting more than 3 days will drop your fitness off the charts (despite what is said about it taking 6-10 days for training to turn into fitness)
• Orthotics are really good because otherwise, I'd be a cripple

Hi Dev - I'd like your thoughts on

1. At what distance the negative splitting goes out the window (you said for a mile it is a good plan, but for a marathon it is not - where do you draw the line?) 10K? Half marathon?

2. What do you consider "elite?" Is it really the elites - what about a 3 hour guy? 2:45? somethig else.

I'm not trying to rain on your parade, just want to hear what you think. Thanks

You make a good points, but what I needed to say (and said later), is that the 10 min fade applies to 98% of the population cause we don't have elite genetics to actually even or negative split a marathon. True, many are under prepared, but there are a boatload of dudes in the 2:45 to 3:30 range that are optimally prepared and still positive split.

Now if you don't believe the reality of positive splitting check the results from the top 25 at Boston this year on a very favourable day with no major headwind. Very few negative split and this is at the very pointy end elite level of the sport:

Bib Name Age M/F City State Country Ctz *

---

7 Negussie, Hailu 25 M Showa ETH Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:56 0:31:33 0:47:18 1:03:05 1:06:31 1:18:31 1:34:01 1:49:20 2:04:32 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:02 2:11:45 2:11:45 1 1 1 3 Onsare, Wilson 28 M Nairobi KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:57 0:31:33 0:47:18 1:03:06 1:06:32 1:18:32 1:34:02 1:49:31 2:05:04 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:03 2:12:21 2:12:21 2 2 2 23 Cherono, Benson 20 M Baringo KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:57 0:31:33 0:47:19 1:03:05 1:06:32 1:18:32 1:34:02 1:49:33 2:05:20 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:04 2:12:48 2:12:48 3 3 3 10 Culpepper, Alan 32 M Lafayette CO USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:58 0:31:34 0:47:19 1:03:05 1:06:32 1:18:32 1:34:08 1:50:11 2:06:25 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:06 2:13:39 2:13:39 4 4 4 13 Cheruiyot, Robert Kipkoech 26 M Nandi District KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:59 0:31:32 0:47:16 1:03:06 1:06:32 1:18:31 1:34:02 1:49:21 2:05:53 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:08 2:14:30 2:14:30 5 5 5 1 Cherigat, Timothy 28 M Chepkorio KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:57 0:31:33 0:47:18 1:03:05 1:06:31 1:18:31 1:34:01 1:49:42 2:06:57 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:10 2:15:19 2:15:19 6 6 6 19 Kipchumba, Benjamin 29 M Marakwet KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:57 0:31:33 0:47:19 1:03:05 1:06:31 1:18:31 1:34:02 1:50:28 2:07:36 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:10 2:15:26 2:15:26 7 7 7 15 Letherby, Andrew 31 M Adelaide AUS Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:58 0:31:34 0:47:19 1:03:06 1:06:32 1:18:52 1:35:13 1:51:58 2:09:03 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:13 2:16:38 2:16:38 8 8 8 6 Ouaadi, Mohamed 36 M Marseilles FRA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:59 0:31:34 0:47:20 1:03:07 1:06:33 1:18:52 1:35:14 1:51:58 2:09:03 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:13 2:16:41 2:16:41 9 9 9 22 Gilmore, Peter 27 M San Mateo CA USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:16:06 0:31:54 0:47:54 1:04:11 1:07:41 1:20:24 1:36:45 1:53:32 2:10:00 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:15 2:17:32 2:17:32 10 10 10 16 Shay, Ryan 25 M Central Lake MI USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:57 0:31:34 0:47:19 1:03:06 1:06:32 1:19:10 1:35:57 1:53:14 2:10:48 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:17 2:18:17 2:18:17 11 11 11 5 Kimutai, Benjamin Kosgei 34 M Elgeyo Marakwet KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:59 0:31:33 0:47:20 1:03:21 1:06:53 1:19:51 1:37:10 1:54:08 2:10:57 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:17 2:18:22 2:18:22 12 12 12 11 Omwenga, Thomas 25 M Kisii-Nyanza KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:57 0:31:33 0:47:18 1:03:05 1:06:32 1:18:31 1:34:37 1:52:40 2:10:47 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:18 2:18:57 2:18:57 13 13 13 8 Loskutov, Pavel 35 M Valga EST Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:58 0:31:34 0:47:19 1:03:07 1:06:32 1:18:53 1:35:39 1:53:31 2:11:17 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:18 2:19:04 2:19:04 14 14 14 30 Komen, Wilson K. 27 M Washington DC USA KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:58 0:31:33 - 1:03:05 1:06:31 1:18:53 1:35:36 1:53:10 2:11:05 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:20 2:19:41 2:19:41 16 16 15 14 Yae, Terefe 23 M Arsi ETH Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:57 0:31:33 0:47:19 1:03:05 1:06:31 1:18:53 1:36:11 1:54:13 2:12:16 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:22 2:20:42 2:20:42 18 18 16 21 Hyung, Jae-Young 34 M Albuquerque NM USA KOR Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:59 0:31:34 0:47:20 1:03:09 1:06:33 1:19:36 1:36:56 1:58:15 2:15:55 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:29 2:23:47 2:23:47 19 19 17 24 Lundstrom, Chris 29 M San Francisco CA USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:16:30 0:33:03 0:49:34 1:06:13 1:09:51 1:23:05 1:40:40 1:58:55 2:16:03 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:29 2:23:50 2:23:50 20 20 18 47 Pelletier, Matt 25 M Warwick RI USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:16:59 0:33:40 0:50:22 1:07:14 1:10:52 1:24:12 1:41:21 1:59:06 2:16:58 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:32 2:24:55 2:24:55 22 22 19 28 Rundell, Carl 37 M Birmingham MI USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:16:33 0:33:03 0:49:33 1:06:13 1:09:50 1:23:04 1:40:25 1:58:26 2:16:44 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:32 2:24:59 2:24:59 23 23 20 F1 Ndereba, Catherine 32 F Nairobi KEN Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:17:37 0:35:04 0:52:27 1:09:46 1:13:31 1:26:50 1:43:47 2:00:52 2:17:38 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:33 2:25:13 2:25:13 24 1 1 62 Post, Eric M. 26 M Centreville VA USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:17:07 0:34:21 0:51:51 1:09:14 1:12:58 1:26:35 1:43:58 2:01:16 2:18:00 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:33 2:25:22 2:25:22 25 24 21 40 Wardian, Michael R. 31 M Arlington VA USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:16:55 0:33:40 0:50:22 1:07:24 1:11:08 1:24:41 1:42:20 2:00:10 2:18:00 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:34 2:25:43 2:25:43 26 25 22 27 Frey, Jacob L. 23 M Rochester Hills MI USA Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:16:19 0:32:49 0:49:22 1:05:56 1:09:36 1:22:37 1:39:37 1:57:34 2:17:03 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:35 2:26:28 2:26:28 27 26 23 12 Amaha, Gidey 27 M Tigray ETH Checkpoints 5k 10k 15k 20k Half 25k 30k 35k 40k 0:15:58 0:31:34 0:47:19 1:03:05 1:06:31 1:19:06 1:36:10 1:54:14 2:14:48 Finish Pace Projected Time Official Time Net Time Overall Gender Division 0:05:36 2:26:56 2:26:56 28 27 24

Negative splits are not a good idea for a mile or anything shorter. For a mile first lap second fastest, second lap third fastest, third lap slowest (and most difficult), fourth lap fastest. First half 1 to 2 seconds slower than second half overall. The mile has been run literally 100's of millions of time - fastest times always have come this way.

Thats the great thing about record books, the same experiment is repeated hundreds of times and the most sucessful solution is put in the record book.

You have to be aware that the first half of Boston is downhill, and the second half is most decidely not.

I will agree with you that for most people negative splits are not possible, but only because they are undertrained or failed to execute on race day (improper pacing, nutrition, unfavorable conditions, blisters, etc...). Genetics have nothing to do with it. Everyone if perfectly trained would run there fastest physologically possible time with a negative split - it would be slower than the elite's time, but it would be thier fastest time. Its very hard to run negative splits (ie. to run your fastest), and most (maybe even 98%) who attempt will fail - doesn't mean its not the right way to do it.

Each second per mile too fast in the first half will make you 3 seconds per mile too slow or more in the second.

So Im doing the chicago marathon in a couple of weeks. I hoping for 3:50. You would suggest that I do the 1st half in 1:50 and the second half in 2:00? The half pace would be 8:23 and the second half would be 9:09, that seems to be a big difference in effort.

---

-----------------------:)
SUPPORT OPERATION REBOUND:
http://www.operationreboundcalifornia.kintera.org/ejs3

Kestrel Syndicate
Macca Fan Club

It does seem like a big difference in effort but a much more realistic option, at least to me, than running 1:55 and 1:54.

first off, you have the best name of all posters.

secondly, i agree with you. If you are trained properly for the distance then absolutely negative splits should be possible in an open marathon (not IM).

For me (an dperhaps most) it takes a few miles to warm up in a marathon and i bet 9/10 people don't start the 10min fade until mile 20. This is most likely a product of physical preparation, mental toughness and nutrition.

How many people hav e gone back to running a straight marathon after IM training volume? Once the race seems "mentally" small and you have dialled in the nutrition element, the physical preparation to get you strong enough to maintain an even pace is a breeze. I am a big believer putting in multitudes of bike hours contributes to this ina big way - superior aerobic fitness with less risk of injury. I am not saying it won't hurt in the final 10k but is certainly should be possible to maintain the same pace and finish quicker than had you blown out at too fast a starting pace....

---

-----------------------------------------------
www.true-motion.com Triathlete Casual Wear since 2007
(Twitter/FB)

I'm on the bullshit bandwagon with you Eddy. The only reason 98% of people positive split is because they are undertrained like you said or because they have no idea how to pace themselves to success. Telling your average 3:45 marathoner to go out and shoot for a PR by positive splitting is just begging for failure and/or bonking. Even if they do manage to hit a new personal best I guarantee they could of gone faster with a more balanced pacing plan.

I really can't get into this "98% of people are undertrained" argument. I train very hard. I read all the shit on what to do and what not to. I can meet or exceed all the recommendations on pacing when it comes to cycling. I am new to swimming but I feel I have been able to follow along with what the "coaches" say and see constant improvement. But when I try to negative split and fail and then see all these other athletes trying to do the same and failing its saying something to me. Maybe there is a place between even splits and a positive split of 10 min but, for me at least, a negative split seems counter intuitive.

"Telling your average 3:45 marathoner to go out and shoot for a PR by positive splitting is just begging for failure and/or bonking."

...or injury. The biomechanical differences of running at different paces must be considered. Not sure who posted it above, but the example of incorporating a "10-minute fade" resulted in a 40+second per mile differential. Those are different ballgames in regards to stride lenth/cadence and contact time. Not sure I would reccommend this approach. i agree with Eddy, do the training and make realistic pacing stratigies based on that training.

Ok guys. Let's buy into the arguement that 98% are undertrained and can't execute. Let's also buy into the concept that the perfect plan is even or negative split.

Now the last time I checked:

1. There are no frictionless surfaces, although they exist in theory in physics texts
2. There are no zero impedence conductors although they exist in theory in electomagnetics texts
3. There is no perfect free market economy although they exist in theory in some economics texts

So even if the even paced plan might theoretically exist and might be the best way to have the PERFECT race, might it not be better to work with reality and come up with a pacing plan that works in the confines of the real world, where as you put it 98% of us are undertrained and can't execute . I am proud to fall into this camp wiith a 1:25-1:34 a few years ago at Boston...first time anyone accused me of being undertrained. I've also done 10 other sub 3 marathons each using this strategy.

What I am saying is that this is real solution for real people doing marathons, not an idealistic template drawn from the profile of the ultra elite marathon racer.

Granted that the 10 min fade is the outside edge. Most of us can hope for somewhere in the 5-10 min range. Give it a go and see what happens. Its mentally a lot easier knowing you have a nice solid "buffer" to your pace time than trying to whip things off at the 11th hour by negative pacing.

This is a good point.

Thats also why only 2% (a fudged number so take it with a grain of salt) manage negative splits. Its very hard to do. A 5 to 10 minute fade from the first half to second half of a marathon is not a bad execution - many have and will continue to do far worse. I think the problem I have with your original idea is PURPOSELY trying for positive splits. It might be realistic to expect to positive split, and it is the most likely outcome for lots and lots of people, but it should not be a goal. Perhaps set an upper range goal, a likely goal, and a lower range goal - example 2:55 , 3:02, and 3:10 - go for the middle range goal pace at the half - 1:31 in this case - and then throw down whats left for the second half. Best case (if the training is there and the gods smile on you) you negative split and hit 2:55, worst case slow down 10 minutes and hit 3:10. Going out in 1:25 and hoping to hold on for a 1:30-1:35 will far more likely than not result in a realllllllllly slow last 10K and a 3:30.

More to the point, its not even about first and second halves - the marathon is a fast 20 mile warm-up followed by the hardest 10K of your life if paced properly - or the slowest if paced improperly.

More to the point, its not even about first and second halves - the marathon is a fast 20 mile warm-up followed by the hardest 10K of your life if paced properly - or the slowest if paced improperly.

This is actually the real crux of marathoning. In the 10 times that I have gone sub 3, my strategy has been to cruise at 4 - 4:05min/k pace for the first half so that it feels easy. Then ratchet back a bit to around 4:12 for the next 12 K or as long as possible and then hang on at sub 5 min per K for the final 5-10K.

I did do one marathon where I just experimented by really pushing the first half and running 1:19 expecting to cave and do the second half in 1:40+. I surprised myself and ran 1:31 for a 2:50 for my second fastest marathon ever. Alas, this was back in 1991, when I had the spring in my legs that youth brings ! Now I am old and cripple :-)

You may be right, but two additional things to consider.

1) Boston is not a great example. If you ran an even effort throughout, you'd probably positive split just because the first half is easier than the second half. The downhills in the first half set most people up for trouble when they reach Heartbreak Hill.

2) For lots of people, having the fastest possible time is not the only consideration. There is also the suffer factor. If you go out too slow, what's the worst that will happen -- you'll cross the finish line knowing you left a few minutes out on the course and could have gone faster. If you go out too fast -- well, as most of us know, that can mean some serious unpleasantness. It's so easy for runners to get caught up in the excitement and go out too fast that, for my money, making a conscious decision to positive split is asking for trouble. You'll probably positive split anyhow, but at least those last 6 miles won't be quite as bad a death march.