NGP vs AveP (Run VI)
From the IM Bike and Run TSS thread, nobody bit on the following so I thought I'd throw it in a separate thread:
What I don't have experience with is understanding how walking 30" every mile impacts that difference (which is essentially your run VI). However, I will readily admit that if walking 30" every mile impacts your VI as it did Dan's then it's an outright ridiculous strategy. Why? NGP is no different than NP. He's clearly wasting a ton of time if he's capable of running ~7:40, as his NGP indicates, yet ends up with a 7:59 AveP.
You see where I'm going with this?
---------------------------------------------------
I think a couple of people made some statements guessing that Dan's delta was due to walking the aid stations and this kind of opened my eyes. I'm curious why this hasn't raised others' eyebrows. I stand by my statement: If walking the aid stations produces a run VI similar to Dan's then I propose it's a horrible strategy.
The analysis for determining a good run VI is similar to the bike but there are some differences to take into consideration.
1. You target a sub-maximal NP (for that distance) on the bike but a maximal NGP (for that distance) on the run.
2. You're still moving forward, often at very high speeds, at zero power on the bike but you're at a dead stop at zero power on the run.
For the two reasons above, you want your AveP to be as close to your NGP as possible. Yes, you still want some variability on a hilly course since the same logic applies to running up a hill as it does climbing up a hill on the bike. IOW, you have to slow down on the hills since keeping pace constant would produce a run-equivalent power of something well over threshold. So you still have to pace accordingly in order to produce the highest possible NGP. However, walking or stopping on the run has a huge negative impact because we don't have the advantages of momentum and gravity like we do on the bike.
To sum it up, you are losing a TON of time by walking. Most importantly, we all know that you don't get back what you already put in on the bike so imagine how much worse it is on the run since zero power or the low power produced from walking has you at a dead stop, in the former case, or moving forward with no momentum working to your advantage, in the latter case.
Thanks, Chris
Comments
I think we need to clarify what it means to "walk the aid stations". If that means walking a few steps to make it easier to drink/eat what you need then it likely will not have that much effect. If it means walk 100 yards or whatever the area of the aid station is turning it into really a Galloway strategy it is something else entirely. IIRC P has talked about how he had to walk a while to bring his overall pace down. That never made much sense to me.
I have never been a fan of the Galloway strategy but people have certainly had decent result using it. Like about everything else in the sport, because it is so hard to study in isolation or with any decent controls, it is often hard to tell if the results are because of doing it that was or in spite of it. If I understand your post you are taking the data and suggesting that Galloway is dead wrong as the time lost is not nearly worth the energy saved. This also requires the assumption that NGP is a meaningful metric.
And, I frankly think that's a pretty big assumption to make. I'm not sure NPG has been around long enough or has enough data behind it yet to qualify it as a solid metric.
That said- you make a really interesting point Chris W. Certainly your VI if you walk the aid stations (or have a planned time interval for walks) is going to be something greater than 1.0. And while I disagree with your statement that the lower power of walking has you at a "dead stop" (clearly it doesn't, you are in fact still moving forward), I understand the point you are driving to. I need to look through some of my own data for runs where I ran steady vs runs where I intentionally took walk breaks to see if I can get a sense for how much of a change in VI that really results in. I'm a slower runner, so perhaps this is going to look a lot different than it would for a faster runner like Dan, but if I am walking with purpose (say at a 15 min/mil pace) and I run at a 10 min/mile pace when I'm running, I don't see that as necessarily having a huge VI swing.
I'm scratching my head- this is a very interesting topic! Hmmmmmmmmmm
Let's be clear, I didn't say that lower power of walking has you at a dead stop. I said, "Zero power has you at a dead stop." It's important to note that zero power on the bike, ie, coasting, basically never has you at a dead stop. Actually, more often, you're going about 37+mph at zero power. Still, the idea behind walking is that it produces low power which is intended to get back what you either put in, or, are going to put in. That's just categorically incorrect. You will never get back what you put in and that logic falls apart even more on the run because walking has such a huge impact on "time efficiency."
Now, I'm specifically addressing the comments about walking 30" each mile that some people made. Maybe they were exaggerating when they made those statements but 30" is HUGE!! I agree, I have a hard time believing someone is taking 30" to walk the aid stations but I gotta start somewhere.
Lastly, I think it's safe to say that NGP is meaningful. If it's not meaningful then I can tell you right now that rTSS is even less meaningful. I'm assuming I wouldn't need to explain to anyone why that is the case? I think the better question is: How accurate is it as a "true metabolic cost" metric? Again, it better be somewhat accurate or all of your PMC graphs (where running is included) are bogus.
Btw, this reminds me of a conversation I had with MarkyV. He said he doesn't believe in NP and only uses WKO+ for PMC. I had to laugh...
Thanks, Chris
Somewhat related, I have always had an issue "accepting" NGP. Maybe it is because I do not have much of an understanding as to how it is calculated. We can all agree that running say 7 minute miles up a hill is a lot harder than on the flats but not really that much but running down hill is not really much easier at all. Also, it seems as if you are taking speed and attempting to use in the same category in which we place power. How useful would it be to take speed and elevation data from a bike ride and then try to figure out a NGP for the ride and then attempt to quantify "how hard" of a ride it really was? Might work somewhat if you had really good elevation data and there were no wind and you knew cda and rr at all times.
The NGP needs to assume that speed and elevation changes are the only variables. That wind/road surface/temperature whatever does not matter. I know that it is a lot harder for me to run on dirt and grass than a road. Wind and weather certainly matter sometimes. Does it take HR into account if you have a strap on? As with any TSS type measurement also need to have the ftp value dead on.
Isn't that the fact that speed does not matter the whole thing that makes power great?
Chris - I contemplated responding when I saw the post in the other thread, but it seemed you had your mind made up, so I bit my tongue.
I'm not going to argue your theory, but just present my own personal experience. I started out racing with the assumption that walking any part of a (longer) race would automatically give me a slower overall time. Back in 2005, I was searching for ways to improve my IM performance, as I was continually on the wrong side of the bubble when it came to Kona qualifying. Along with many other ideas, a training partner mentioned the concept of "strategic walking", I thought, "why not" so I started walking about 20-30 seconds every mile during my long runs. My pace per mile immediately dropped from about 9:05 to about 8:49. With a lower avg HR. That opened my eyes, so I did it in my next IM, Wisconsin '05 (the hot one). I qualified for Kona there, and never looked back. I'm convinced that for me, walking strategically in an IM DOES give me a faster final time, both for the IM and the whole race, than I would get if I tried to run every step. Also, this little experiment covered two consecutive Boston stand-along marathons. In '05, running the whole way )while splashing gatorade on my shirt) garnered me a time of about 4:07; walking each aid station on '06 netted a 3:46. My 5K and 21 K times were no faster, and I actually did LESS run specific training in 06, so that also alerted me to the value, for me, of walking.
I don't really care WHY this is, but I do have a couple of ideas about how it may help. First, my HR drops very quickly (within 10-15 secs) from about 122-127 down to 105 or so. I think this allows momentary increased GI blood flow, allowing increased absorption of fluid and sugar during and immediately after walking. (I don't start walking until I grab whatever it is I'm going to drink - Coke, gatorade, water.) Maintaining a constant glucose infusion at this point in the day is critical. Second, the different physical motion in my legs while walking may provide a brief "recovery" phase, not unlike doing turns in a pool compared to open water swimming. And third, don't disparage the psychological value of getting a break to maintain the brain work needed to stay on top of one's pace.
If I didn't walk every mile for 20 sec, would I do 8:49 miles instead of the 9-9:15 I do? I doubt it, but the only way to find out is do another experiment, and I'm not willing to risk it.
Does anyone know for sure the impact that a walk break on a 100% flat course would actually have on both NGP and AveP? I would think (and could be wrong) that both NGP and AveP would be the exact same, because the elevation is not changing.
If I completely misunderstand this, someone just tell me to butt out and read along instead.
I believe the change in elevation is just one factor (probably the main one, but not the only one). From the TrainingPeaks website:
Bear in mind that that not only changes in grade are used to calculate the physiologically relevant pace of NGP. As with NP in cycling, an exponential weighting step is utilized that is based on the relationship between intensity and lactate accumulation. So, not only is the NGP faster than the raw reported pace as a result of a positive grade, in some circumstances, an effort that has a component above the “threshold” will result in a more metabolically costly effort than indicated by the raw pace.
I still can't find the exact calculation that is used for NPG (and thus rTSS). That would certainly help with this discussion. That's also why I'm interested in testing it out myself on a flat route. One 6 mile loop done with and without walk breaks to see what the VI change is and how that impacts NPG would help me get my head around this a little better.
-it's all in your head, it didn't actually help, placebo effect. I don't buy this one, but it's possible.
-it's not about metabolic efficiency at that point in the race. Perhaps, it becomes easier to absorb the calories being taken in at a lower HR, making the body better able to keep the energy supply up.
-the actual impact to 'VI' is not as much as you think. I'm interested to see Nemo's experiment, and understand the impact of walking more on NGP. It's possible that the effect of short walk breaks does not up the physiological VI, though i don't really have a plausible reason to suppose this.
-could be a thermoregulation effect, where slowing from pace for short periods helps dissipate heat, enabling continued work at a slightly higher rate than otherwise possible. I believe that Patrick has said that the benefit he has seen is the ability to run a higher pace when not walking, but feeling it less.
Chris G, speed on the bike is different than speed on the run. Combine elevation data with that, apply the exponential weighting factor appropriately and you can certainly get a very good metabolic cost metric. Sure, it might not be as accurate if you're running into a very stiff headwind but under many circumstances it gets you very close. Certainly closer than AveP.
Matt, as Nemo pointed out, you definitely have that incorrect. Sounding repetitive, they use an exponential weighting factor just like NP. Again, the idea is to provide you with the equivalent of NP for the run. Remember, rTSS uses NGP. This stuff is all tied to estimating the true metabolic cost of your ride or run to give you a better idea of actual training/racing stress.
My personal belief is that walking primarily provides a mental-only benefit. There are things to consider with HR and thermoregulation in certain individuals but I'd argue there are other (better) ways to address those issues that allow you to run the entire marathon.
Btw, maybe I'm missing something but I don't think you need to be a rocket scientist to figure out that walking is going to impact that VI. Why? Because the algorithm uses an exponential weighting factor and low power from walking yields a significant decrease over your AveP unlike cycling which yields a very minimal decrease over your AP, or, quite often, an increase over your AP (eg coasting at high speeds). Do you guys remember the stuff Ashburn wrote on "time efficiency" for the IM bike? Always be pushing... Hey, that logic applies even more to the run because you lose so much time by walking. So, always be moving...
Maybe one of the positives to take away from this thread is that we should start paying more attention to NGP and see for ourselves. I had some doubt about NP very early on but came to the conclusion quite quickly that is was much more reflective of metabolic stress than AP. Clearly we see some value in rTSS so we have to ask ourselves why. You can't believe in rTSS yet not believe in NGP.
Thanks, Chris
I think that I will pay much greater attention to NGP for a few weeks and give this all some thought.
** I also want to make note that this a bit of a dangerous topic. Sure seems like you can't go to easy the first part of an IM run leg and that no matter how easy you go early it get really really hard later on. It is a great spitballing discussion but want to make sure peeps are not going to ignore the walk break advice as CDA and end up walking a lot more than the aid stations as a result**
Good point. New Peeps- do not confuse this discussion with official EN protocol or advice.
Now, paying attention to NGP and VI has already produced an interesting result. First, I checked my last several runs and my VI is listed as N/A on every one (and those range from 20min to 1:30min runs). Hmmmmmmmm. So how can the NGP use VI if it's not calculated? FYI- I always fix the elevation even though I almost always run on entirely flat routes
Today I did a Race Sim (6 miles) where I ran 6 miles (totally flat stuff, net elevation gain/loss of 90 ft). I took 4 short (30sec) walk breaks and had to stop for about 10 seconds to cross a street because a car was coming. My NGP= 9:33, My Avg P= 9:34
I realize this is a pretty short test- but it doesn't seem that the walk breaks impacted NGP by much at all (and again, VI is listed as n/a). I pretty much just walked through the door and downloaded the data, so I haven't done any work to figure out why I don't seem to have a VI (frankly, I need another cup of coffee and more food!!!). But I wanted to share the first sample of data.
PS- I went back to look at my IMWI run data from '08. I had a different Garmin back then and with the hilly route I thought surely VI would be calculated. Nope, VI still says n/a. Hmmmmmmmmmmmmm
I'm still fixated on the question of the potential cost of walking 30 seconds every mile vs. running every step. I did some back of the envelope calculations (the only kind available to me since my last math course was 44 years ago as a college freshman.) I assumed 15-16 minutes/mile for the walking pace and 9-10min/mile as running pace. It turns out that I "lose" about 4.5 minutes or half a mile by walking instead of running for 26.2 miles, IF (that's the Big Question) it's true that I can run at the same pace without walk breaks that I run with walk breaks. That works out to about 10 seconds a mile.
So I have to design an experiment sometime in my training (don't want to risk a race) and see if I can indeed run 10 sec/mile faster while running with walk breaks, compared to running with no walk breaks. Or, more simply, if my overall pace walking + running is the same as my running pace (without walking). I firmly believe it is, or maybe even a bit faster, but it would be nice to prove it. It's hard to eliminate bias in the experimental protocol, though (can't really double blind the thing, although it can be randomised.)
I think some of the bias against walk/running may come from the fact that the fastest pro IMers don't do it, stand-alone pro marathoners don't do it, and Galloway evangelizes it as a way to get people off the couch and into running - it's viewed by faster runners as something for the lame, the halt, the slow - people who are not "real athletes".
As a side note: I went to college at Wesleyan University from 1966-70, the same time that Jeff Galloway, Ambi Burfoot, and Bill Rodgers were all there. Maybe that had a reinforcing effect when, many years later, after I had started run/walking on my own, I learned that Galloway was it's biggest proponent.
For that matter, maybe we can set this up as a Challenge. And even get a spread sheet jockey to devise a little Excel formula to help calculate things based on one's walking and running paces.
So (directed more towards everyone else)....
1. I'm not really trying to diss walking. What I'm trying to do is to understand how variable-paced running impacts your time efficiency. Walking the aid stations or something similar is an obvious strategy to target. I won't argue for a second that a run/walk strategy could probably be ideal for anyone who's doing their first IM or two.
2. If it sounds like I'm convinced about what I claim above then don't let me fool you. I'm far from convinced. You guys know me, I just happen to do some analysis, form an opinion and sound like I'm taking a hard stance because it tends to get more people engaged. Clearly it's working... ;-)
Thanks, Chris
As a basic analysis, it would be interesting to collect data from HIM and/or IM racing, for both NGP, Avg Pace, and perception of effort on the race, well as any strategies employed (run/walk, etc). My n=1 experience tells me that you just don't lose that much time walking an aid station, and the Gatorade goes down much easier. That being said, that opinion is far from scientific.
Chris, it would seem to me that the best analogy to prove your point (it has to be metabolically inefficient) is to imagine doing the opposite, ie. Doing a Z3 interval through every aid station. Surely, that would increase the metabolic cost significantly, without helping overall speed very much.
Mike
I wish I had something to add to this discussion, I got nothing other than my splits were pretty even if anyone wants the file msg me.
@Al per your request
Run walk spreadsheet
This sheet allows you to put in run and walks paces and the times you plan to spend runing and walking for each segment. It calculates the overall pace and marathon and 1/2 marathon times.
Matt
Thanks for clarifying Nemo and Chris. I was convinced I read somewhere that NGP only acconted for grade changes, but I couldn't find it again anyway to even determine if it was a reliable source.
I personally have done HIMs in 1:33, only to realize that I was running 1:28 pace, just walking more...and more....and more....at each aid station as the race went on. But then again it was all I could do. I have run pretty close to the same pace at the last 4-5 Ironman races, time savings have come from reducing the fade at the end (better fueling) and not slowing down more at aid stations (stay focused!) as the race goes on. I just can't imagine running a set pace that I can sustain all day in an IM, but that could be just that I am mentally soft by that point in time.
Given I have run a 2:59, I am confident that when fit (and not in a wheelchair!) I can run a 3:15 on IM race day, not my typical 3:25. But that would come from faster overall pacing, not by cutting the walking.
I have a lot of thoughts on this matter and will do my best to get to them before the end of the day. However, I will just drop down situation that has me thinking about how I train.
After last weekes tues interval, I went out to do my brick at the predetermined paces. 7:47/6:36. It is a 10 min out and back over rolling terrain. If you look at my Garmin the paces for the first half vs. the second are 7:42/and 6:42. This morning after reading the discussion, I went and checked the WKO file and my NGP for the two halves was 5:56/5:42. What it really means I am not sure, but I sure felt like I was working during the 20 minute run.
Todd
Ow. Pulled a brain muscle. Will go over all my data when home tonight. Interesting topic and have to agree that during marathon in IM it makes sense to "let up" a bit at aid stations to take in fuel. Even the 5-10' respite allows for a decrease in hr and a chance to "reset" on form which hopefully = better efficiency!
Not sure. Will know more later.
I believe the amount of time you lose is by walking the aid stations is a relative thing. It's not much for someone running a 4+hr marathon but it's potentially significant for someone like me. Like the difference between qualifying for Kona and not. So, this is one of those subjects that really applies to the sub-4hr marathon person, imho.
Wrt time efficiency, we had a rather lengthy thread on this subject on ST about 2 or so years ago. I also have some personal e-mail from him too so I'll post some of that. Funny enough, I just sent Ashburn an e-mail about this subject and he responded yesterday. We're still exchanging e-mails so I'll let you know his thoughts after I get back from Honu.
If I was to just take a devil's advocate position on my own thread, the big problem I'd have with all of this is the fact that the IM run has a large mental component to it. Much more so than the IM bike. Clearly there's no way any of us could quantify that mental component either. So, it's pretty hard to make this just an academic study because of that mental component.
Thanks, Chris
i didn't see this post before I left and there are a couple of comments in here that got me thinking a bit more...
No doubt, I realize I was trying to take a "what is the best scientific approach" to something that is at a point in the race where things become very mental. I totally agree, it's highly individual so I could easily argue that telling someone that walking (the aid stations) is ok on the IM run can be equally dangerous (purely from a mental perspective). I've seen this countless times: Someone makes a strategic decision to walk a section of the aid stations. As the race progresses then it becomes the entire aid station. And then it becomes the entire aid station plus another 20 steps. Etc. They basically tell themselves that since they were already walking periodically then how can a little more walking hurt?
In Patrick's case above, it was all he could do but I think quite often it's simply a mental breakdown by the individual. I've been in that situation myself where I felt it was all I could do but then I had Mitch Gold yell at me, "Keep moving!! Whatever you do, just keep moving..." and once I did that, along with getting more fuel in me, my race changed substantially. And this was a race where I qualified for Kona too. I was so close to just walking more and more as the race progressed vs qualifying for Kona. If Mitch hadn't been there yelling at me to keep moving, I most certainly would have kept increasing my walking time.
Honestly, I have seen or heard of that same situation occurring in the past almost as often as I've seen someone be disciplined and stick to their original "walk part of the aid station" plan. So, just something to think about.
In the end, there's no way we can really look at this purely from a scientific perspective. However, I still think there's some applicability to run VI (NGP/AveP).
Thanks, Chris
I am that 4+ hour marathon guy. Maybe 5.
There is no doubt that there is an enormous mental component involved in the ironman "run" leg and that it does make it very difficult to draw and scientific conclusions regarding how the best way to approach it. I do note however, that I am far more able to keep going and battle through mentally when I am in better shape physically.
You can be in great shape and fall apart mentally [maybe] but there is no way that you can fake it if you have not done the work and just bury yourself mentally and keep going. Unless the physical is ready to roll the mental component does not really matter that much.
Actually, it's not a diversion at all and here's why:
For the exact same reasons you tell yourself to execute a walk break strategy, you might be better off running the entire way. First off, I've heard several people claim before that you need to walk the aid stations in order to get properly fueled. That's simply not true. I've done both and neither approach has allowed me to fuel better.
Secondly, Joel, you just said that you need to walk for mental (focus) reasons yet you also claim that making the mistake of walking and walking more often as the race progresses is a mental focus breakdown. This is exactly my point. So, you're doing something, ie, walking, for (good) mental reasons that also just happens to be the result of a mental breakdown.
Why do the one thing that happens to be the cause of what you're trying to avoid in the first place (a mental focus breakdown)?
Thanks, Chris
I've kind of stayed out of this since the topic moved on to the concept of mental focus with regards to the IM marathon and away from the actual topic title.
If you want to calculate Run VI (Normalized Graded Pace divided by Avg Pace) the same way Bike VI is calculated, what I theorize you will get is the variability of the terrain, and not variability of effort (as others have pointed out.) While Garmin does a kind piss-poor job at times (even with elevation correction in WKO), you can test this (either for real or in your head) by contemplating this test protocol:
So, assuming that elevation correction actually gives you a flat course for your track (never does for me, but it should be close enough) what is your NGP for each of the two tests? You ran 3 miles in 30:00 running at 10:00 min/mi and you ran 3 miles in 30:00 run/walking at 9:34/16:00 min/mi. So your average pace for both runs will be 10:00. Convert both pace times (NGP and AvgPace) to seconds. Divide NGP by AvgPace (actually to keep things consistent, numerically, with the way Bike VI is calculated you probably want to divide AvgPace by NGP because AvgPace in seconds will likely be the numerically larger value, indicating that it was the physiologically lower value.)
I contend that you will end up with something very close to 1.00 if the altitude correction is close to reality (i.e. no altitude changes shown) which illustrates that NGP is supposed to provide an estimate of your pace relative to elevation, not relative to effort.
Now if you want to be thorough, you can go out and conduct Test #3: On a rolling course (more is better). Run 3 miles at 10:00 min/mi pace. Be very disciplined and stay right on 10:00 both going up and coming down the hills. Import into WKO+, run altitude correction, convert NGP into seconds and calculate the putative Run VI again. What's the result? I hypothesize that it will show a very "variable" run despite pace being constant.
Theories are great, data is better - any takers on actually doing it?