123

Negative splits for the marathon - theory or reality? (Read 931 times)

DoppleBock


    I have always tried to ride the thin red line.

     

    The only time I swung for the fense was 2010 Milwaukee marathon - There was a group shooting for a sub 3.  I was pushed to max to try and catch them. I caught then just after the 1/2 way point.  Rolled with them to 20, then passed all but one that left me in the dust.  I was over the red line since mile #3.  I hit 3/4 mile to go and my body shutting down - 1st time I could not force my will on it. I was running 7:30 pace then 8:00 the last bit.  Ran 2:59:30

     

    It is sad that one of my best run marathons - gutsy, where it was "My Day" was when I was in that poor of shape.  I had no right breaking 3:10.  But we had a light tail wind and I wanted a sub 3 bad.

     

    Yup, it's a crap shoot. Do you want to minimize your chances for major suffering, or would you rather take the 1 in 10 shot at a breakthrough knowing that there's a 90% chance you miss and suffer like an animal over the last few miles?

     

    I've always swung for the fences, which is why my ratio of positive splits to even/negative is 12 to 1. It's not a game for the faint of heart.

    http://a-big-horse.blogspot.com/ 

    2013 Goals ~ Mar < 3:00, 5M < 29, 10k < 35  

     


    I'm back!

      Yup, it's a crap shoot. Do you want to minimize your chances for major suffering, or would you rather take the 1 in 10 shot at a breakthrough knowing that there's a 90% chance you miss and suffer like an animal over the last few miles?

       

      I've always swung for the fences, which is why my ratio of positive splits to even/negative is 12 to 1. It's not a game for the faint of heart.

       

      Yes, I am too chicken to race that way, thus my high ratio of negative/positive splits. But I don't think I have left a lot on the course in my best efforts. Both Boston sub 3s were faster than fast, flat tune-up races predicted. I try really, really hard to know the pace I should be able to run, then run it.

        I agree with the second sentence (and the parts of the post that I didn't quote for that matter) but how is any of that putting the cart before the horse?

         

        In races we've spent months training for it only makes sense to pace ourselves based on what we think we could do on a good day, knowing full well it doesn't always (or even usually) work out. To me neither one is the cart or the horse--to run your best possible marathon you need to run the right pace AND you need to be having a good day.

         

        Cart: even/negative pacing

         

        Horse: have a good day

         

        I don't think any of the posters on this thread really said this, but I think the message that gets across a lot of the time is something like: if you want to have a good day, then you should aim for an even or negative split (and that folks who run positive splits somehow screwed up their pacing.)

         

        I guess I just want to say that the truth of the matter is more often the converse: "If you want to run and even or negative split, then you should have a good day!" And that positive splits are not always a mistake in racing.

         

        Meh. When I type this out, it doesn't really sound earth-shattering, but I guess my point is that if your splits are one or two minutes positive vs/ one or two minutes negative, your pacing was probably fine and more likely the difference was whether you had a good day or not. The marathon being something of a crapshoot/fickle mistress/frustratingly frustrating to get exactly right.

         

        MTA: I have run one negative split marathon (my first--the Monkey). It was a huge negative split (1:30/1:21), and interestingly I don't think I left a ton of time on the table. My PR marathon was about a two minute positive split (it hurt a lot worse than that other one.)

        J2R


          I have to say I hadn't really considered the point about the elites racing, not time-trialling, when I made my post, although it's pretty obvious when I think about it.

           

          Yes, I agree that a minute or two either side probably constitutes even pacing. What is nonetheless surprising is that they're pretty well all on the positive side, albeit only by a couple of minutes in most cases. It's not 50/50 by any means.

           

          For me, I have to say that keeping a rock steady pace was a useful means to an end. I broke 3 hours for the first time on this marathon, by 30 seconds (I wasn't aiming for any faster), and I think I was helped by focusing on precisely what time I should be on at the next 2km marker. It enabled me to break the remainder of the run down into 2km chunks, which was psychologically helpful.

          DoppleBock


            BTW - Congrats on breaking 3.

             

            Sadly I am going to run a marathon slow 11/18 and it is still going to suck the whole way - Damn Monkey

            http://a-big-horse.blogspot.com/ 

            2013 Goals ~ Mar < 3:00, 5M < 29, 10k < 35  

             


            I'm back!

              I don't think any of the posters on this thread really said this, but I think the message that gets across a lot of the time is something like: if you want to have a good day, then you should aim for an even or negative split (and that folks who run positive splits somehow screwed up their pacing.)

               

              I guess I just want to say that the truth of the matter is more often the converse: "If you want to run and even or negative split, then you should have a good day!" And that positive splits are not always a mistake in racing.

               

              Well, I think it's pretty close to true -- if you want to maximize your odds of running a particular time, then it's not far wrong to say that you should aim for even pacing for that time.

               

              Folks who run positive splits either screwed up their pacing, misestimated their fitness, or swung for the fences, which is not necessarily screwing up.

               

              The complication is that we have varying utilities for finishing at various times. The objective is not always to maximize odds of running a particular time, with no concern for finishing time otherwise. In practice, I would argue that this still tends to lead to running even to negative splits when fitness (odds of being able to run a particular time) is known relatively accurately, and pacing is chosen rationally. Expected finish time increases much more rapidly as you pace faster than your true fitness (blow up) than it does as you pace slower (leave something on the course), so to optimize your expected finish time you should pace for slightly slower than your expected fitness. If your fitness estimate was accurate, this will result in a slight negative split, so arguably you have aimed for a negative split.

               

              At least, this is how I see it.

              spinach


                Yup, it's a crap shoot. Do you want to minimize your chances for major suffering, or would you rather take the 1 in 10 shot at a breakthrough knowing that there's a 90% chance you miss and suffer like an animal over the last few miles?

                 

                I've always swung for the fences, which is why my ratio of positive splits to even/negative is 12 to 1. It's not a game for the faint of heart.

                 

                I agree with this.  You never know if you're going to have a good day and so I always go out as if it were a good day, just in case it is.  I am quite happy that I do this because two weeks ago I was expecting to do a 3:10 or 3:15 in my last marathon but I went out strong as i always do and I was quite surprised when I ended up doing a 3:03.  If I held back to my expected pace at the start I probably would have run a 3:15. 

                 

                I have noticed I can go out the first mile or two fast and then settle into my more comfortable pace and so I try to steal 30 or 45 seconds at the start and most of my best marathons have had a pretty even split with the first half those 30 or 45 seconds faster than the second half. 


                HobbyJogger & HobbyRacer

                  I remember watching Mary Keitany swing for the fences about one year ago. It was exciting, even through the clear "blow-up" part, which started not long after the halfway point, I think -- and it was the kind of "blow-up" in which the best could catch & pass her, not the kind with walking or sitting down.

                  It's a 5k. It hurt like hell...then I tried to pick it up. The end.

                  SprinklesRunner


                  Whippet

                    Interesting.  I ran my first marathon with a 1 minute negative split at 4:07 (2:06/2:01) and my second marathon 5 weeks later with a 4-minute negative split to come in at 3:56 (2:00/1:56).  II read in Pfitz that its actually not optimal- I'm leaving too much on the course and need to go out faster.  How to get through the crowd is the piece of advice Pfitz is missing.  I don't think I could have pushed through the crowd any faster than I did during the 3:56 race.  

                    13.1: 1:45 | 26.2: 3:55


                    HobbyJogger & HobbyRacer

                      A smaller marathon would probably reduce the crowd problem?

                      It's a 5k. It hurt like hell...then I tried to pick it up. The end.

                        A smaller marathon would probably reduce the crowd problem?

                         

                        Or a big one with waves/corrals? 

                        "When a person trains once, nothing happens. When a person forces himself to do a thing a hundred or a thousand times, then he certainly has developed in more ways than physical. Is it raining? That doesn't matter. Am I tired? That doesn't matter, either. Then willpower will be no problem." 
                        Emil Zatopek


                        HobbyJogger & HobbyRacer

                          I'm not convinced corrals help at all - when I've seen them, they pull the corrals out, and everyone crowds forward, before the race starts, so then it's back to super-congestion.

                          It's a 5k. It hurt like hell...then I tried to pick it up. The end.

                            Last Sunday I ran the Frankfurt Marathon, and was happy to get the pacing just right, or so I thought - very slight negative splits, second half 25 seconds faster than the first. When I looked at the times of people who came in in a time close to me, I was struck by the fact that nearly everybody else had positive splits. OK, I thought, but surely the elites will be showing negative splits. But no - of the top 50 runners, only 3 had negative splits! And by elites, I do mean elites - the race was won in 2:06:08 by world record holder Patrick Makau, and even he had positive splits, albeit only very slight.

                             

                            Given that this was a flat course, on a day without strong headwinds, there seems no physical reason why the second half should have been slower than the first. This makes me wonder whether this whole idea of the optimal way of running a marathon, with negative splits, might be some kind of 'in an ideal world' thing, and not in fact what happens in reality even for the elites? Of course, there is the theoretical possibility that the midway point was slightly mismeasured, but that seems unlikely - after all, it's Germany we're talking about hereSmile.

                             

                            So how often are negative splits in fact run in a marathon? Or are they only in the theory books normally?

                            Physiologically, even pace is the best (=most efficient) way to run distances.  You may be talking about ONE race but, if you look at more wide variety of situations, you'll probably find more PRs, not just for elite but just about anybody, had been set when they ran either even pace or slightly negative split I'm sure.  At least, that's proven to be the case as a theory.

                             

                            That said, it is extremely difficult to "prove it" in a practical sense.  If someone ran, say, 3:45 in, not negative split but say significantly "positive" split and set a PR by significant amount.  Well, how can we say that this guy's 3:45 is close to his/her max?  How can we be sure to say that he/she wouldn't have run even better PR is he/she ran negative split?  As someone else had pointed out also; those elites usually race and they have to be up there if they want to grab the opportunity to win or place high.  Just a few weeks ago, I was talking to Don Kardong who finished 4th in 1976 Olympic marathon.  He came back from behind to PR and finish "unexpectedly (he was a total dark horse)" well.  He admitted that, had he tried to place and medal, it may not have happened.  "Those guys would have to be up there and race," he said.  He ran it without much great hope and he started out conservatively and, who knows, that might have been the reason why he did so well.

                             

                            I'm not a statistician and I don't know the exact numbers; I don't know what kind of split Makau did when he set the world record; but we COULD even say that he didn't run negative split at Frankfurt and ran nearly 2.5 minutes slower than his PR.


                            Feeling the growl again

                              Well, it's all anecdotes but I'm convinced in the value of an even/slightly negative split...and the difficult of producing that.  I mean, setting out to do that is no easy task because that means knowing exactly what you are capable of on a given day.  No easy task.

                               

                              I've run two sub-2:30s.  My best was the first, a 2:29.  From memory it was a slight (perhaps 20sec) negative split.  Not that the effort was even...for reasons I won't go into I went out too fast the first 3 miles, and gave back almost an even amount of time the last 3 miles as a result.  But I was particularly strong miles 16-22 so it was slightly negative overall.

                               

                              My PR was about 40sec faster, but very positive.  There were many reasons why this was a crap race and probably ~6 minutes slower than my capability only a few weeks before so it's not worth analyzing....I showed up trying to salvage a 2:24 from what was supposed to be a 2:21 attempt and after 3 miles when the injury+overtraining was clear, just wanted to salvage a PR and that's what I did.

                              "If you want to be a bad a$s, then do what a bad a$s does.  There's your pep talk for today.  Go Run." -- Slo_Hand

                               

                                I ran two  marathons this year and each one was a PR (2:52:33 then a 2:49:54)

                                 

                                The first PR was in the spring and I ran about a 10 second negative split. I felt like it was a perfectly run race from an effort expenditure standpoint.

                                 

                                The second PR was about a 3 minute positive split. I believe I would have ran faster if I hadn't ran the first 1/2 as fast as I did. 

                                123