I really like your approach and thorough testing. However I think some of the data is so close that the statistical variance is way too low to draw any conclusions except for some obvious. I run both Nike VF, Asics MSP and Adios Pro 3 and yes, the VF and MSP are very fast and aggressive but I think the Pro 3 will come to be a benefit after 20K into a marathon. Keep the video's coming!
I agree. Data set is too small. Conditions variable. Fitness evolving. Etc etc…. But it is what it is. You can still make some informed judgements. The AP3 is a disappointment…. I’ve done a lot of long runs in them, but the AF3 is hands down the better option for me. I’m also so curious how the AP4 will fair… especially given the tech they I think is in it. I can’t wait to do those tests.
@@SagasuRunning what's your view on the aging of super critical foams. One insider in the industry told me that it's more the aging of the foam that makes a super shoe as these perform less than wear and tear
It’s definitely a thing. Aging of any foam is a thing, super foams go “off” faster. Some have shorter shelf lives that others. ZoomX is on the lower end, LSP is on the higher end. One of the things interesting about Puma’s A-TPU nitro foam is that it should be stable. It’s a fine line right now. Current generation foams are being pushed to their breaking points, literally. The next generation will over more avenues of development, but we are still a few years away.
Brilliant as usual Chris Sagasu. Took the AP3 for 18 yesterday, nearing 200 miles. I have felt the same way for a while and agree with you. Once a runner churns lightstrike pro into butter, the feel of the ride is magnificent and exciting, however, the bounce or energy return is a little more muted than the competition. It feels so nimble and if you’re in control that I understand why the pros love it so much. My conclusion is that it’s a training super shoe I can use for my MP long runs in the same way people train with non plated shoes , but then get a little boost on race day. The AP3 is obviously more efficient than a non plated shoe, but it’s somewhere in the middle, closer to the plated side, but not quite my race day selection. 🎯 you always nail it
Spot on. The AP3 is the best 'training' super shoe out there. I don't think I'd race in it, but use it for all my sessions. So incredibly durable and because of that actually pretty reasonable value for money, particularly as they go on sale fairly regularly.
The AP3 is such a good long run trainer. The further forward rocker point just slowly rolls the shoe forward and LSP is magnificent for durability and recovery. Also, to be fair to the AP3, it is over 2 years old and a full generation behind the other shoes on this list. Things have changed since the shoe was designed. I also think a runner with a longer stride will benefit from the AP3 more, I'm definitely a cadence runner and respond to shoes with much earlier rocker points (VF3, Evo 1, etc).
Fully agree. 💯 LSP is also so durable the AP3 is a better "long run" or tempo shoe over the B12 for a dedicated marathoner. You get a lighter and more performance oriented shoe for the work you need to do. The B12 is great, but better for more general running.
Splitting hairs based on differences that are within the measurement error for each of these metrics seems like we're missing the forest for the trees. This data suggests to me that all of these shoes more likely than not have no measurable effect on performance and that we should probably just go with what is the most comfortable and "feels" the fastest.
Actually...I agree with you. However, I do think one can still "split hairs" to confirm those feelings. As I said in the totals...I wasn't really surprised to see the final order, it was almost exactly how I would have placed the shoes prior to looking at the data (the AP3 would have been higher up though). So it was good to see the data confirm those feelings. That is the point of all of this.
@@SagasuRunning I'd be interested to see if your ranking stays the same with a bigger sample size. When I did similar tests (fellow spreadsheet guy), I found that the shoe I was wearing just wasn't important. I got so deep in the weeds that I had a formula that controlled for wind direction, elevation gained, and my average weight over the course of the run (pre and post run weigh-ins required). I even resorted to running all my tests on a track because the gps error from my garmin was enough to skew results. Eventually I had to shift my mindset from seeing my tests as a failure to looking at the bright side that I could just wear whatever shoe I liked wearing the best : )
@Jruns With any super shoe analysis it’s always splitting hairs. The main benefit you will find is what super shoe works for you at what distance. While it was a great video you should take the thought process and idea behind it and try to find a way to analyze your own running with it. None of these shoes are going to be the best or be all end all for everyone. It will be what works best for him and his running.
@cortneywebb1677 Very much… it’s very personal…. However, larger multi user data sets do offer different and interesting insights..I often work with enough data scientists to learn that.
Wow! There is a lot of effort, work and running to put together this analysis! I was not sure I would watch this half-hour video but I gave it try and it was more than worth the time. Great job!!
Your overall conclusions are interesting. I feel there is are two sets of shoes (VF3, AF3, Edge Paris) are clearly "better" for you than say the Pro 4, Pro 3. I think the Sen 10 and Peg Plus come with asterisks as the tests are quite slanted to shortish efforts so you would kind of hope the Sen 10 would do OK especially in "sprints". Peg Plus actually did better relatively lot better than you would think. Reminds me of when did 20x400 in the Next Nature. I could "get going" but maybe 1-2 secs slower than say Vaporflys but hardly matters other than vanity in training. Sometimes with these tests I feel that it is almost a validation for what you like to run in the most and also got to think what race distance am I actually looking to. When I was doing marathon training, I found doing a "shoeoff" of repeat 5K ish reps at MP in different shoes back to back not only is a very good training session but gave a little bit better insight into what I feel would best for me when tired deep into a race.
@@markoljustina ha ha. These days I only tend to comment on the channels I like or at least the ones that actually want to engage in the comments. This is an interesting video especially as quite similar to the "shoeoff" ones I have done myself. It is harder now in these sort of tests as the differences between shoes are quite small. The results think largely align with that I might have expected from the review videos in terms of what Chris likes to run in.
Pretty interesting to see this data. Didn’t think I would be into it, but it was actually fun to see how each shoe made your workouts and running feel experience different. Thanks for the video.
The second I hit play on this video, I could already guess what the comments would be. Something like, “But there’s not enough data, and you didn’t control for the environment!”-as if this is going into a peer-reviewed journal and not just a chat about shoes and their use. As an econometrician, I love seeing people think critically about data, but it drives me nuts when the nitpicking gets out of hand. That said, I really enjoyed the video.
It work with enough data scientists and analysts to know that data always has value, it just matters on context. I appreciate the kind words though and glad that the video was enjoyable and I hope useful.
Fascinating video and loved the detailed comments on each shoe for each workout. I took screenshots throughout and even programmed this workout in my watch to test some shoes in my current rotation. The SL2 will be my control shoe. Keep up the great content!
@@SagasuRunning I’m curious as well. I will be doing the test on the following: Magic Speed 3, Takumi Sen 8 (300+ miles) Atreyu Race Model, and Endorphin Pro 3. I will be starting my data set next week after my 10K race this Saturday for attempt #2 to go sub 42’ using the Endorphin Pro 3.
Generally speaking the fastest shoe for you is going to be the least fatiguing and BPM over 2K is an insufficient sample size to gauge that assessment over 26.2 and only one factor in a much more complex equation. What shoe can you ball strike a 5K or 10K? What's your unique metatarsal stress threshold point in a specific shoe? The carbon plate might accelerate that but it's going to take much more than 2K to get a sense of what shoe you and all runners have the skill & strength to handle for the intended design and use of the shoe. Adidas Energyrods 2.0 is technology not made to deliver its benefit/value in a 2K measurement. The whole idea is the torsion flex will stave off the metatarsal & leg stress you'll absorb in the carbon plated shoes - over distance - and your test here does not address where Adidas shoes are designed to shine (staving off metatarsal & leg stress over distance). Hence in reality the Adidas Adios Pro 3 might be the fastest shoe for you in the actual use of the shoe per what it's designed for, road racing distance.
Well...the AP3 and TS10 have the same rods, though the AP3 is carbon and the TS10 is not. I find the rod feel between them very comparable. The issue with the AP3, that I've know for some time, for me is the geometry. I don't respond to late stage rockers well. Luckily the Evo 1 solved this and the AP4 will have that geometry (plus some other tech) As for the short distance... while I agree to REALLY know (as a half or full marathoner) you need to do the long runs and compare the data....I have many in both the VF3 and AP3 over the years and know, very clearly, the VF3 is much faster for me...though the AP3 is a more stable long run shoe. I also disagree that Adidas's rods need a long distance to work. The TS10 is designed for short distances, with the same rod alignment and structure as the AP3/Evo1/B12 (jsut not carbon). I do agree that the benefits of rods or a plate compound the longer the distance, you don't need to ONLY run long ditances in them to find out what is working for you, at least enough to validate your feelings on the shoes and then work with a smaller subset to really fine tune and look at the data.
In this dataset, no. In my general running and what I know will work for me over 16k, 21k and 42k…. Yes. I know a more flexible carbon fiber plate (like what Nike uses, compared to say Asics) will work for me better than the Energy Rods 2.0 Adidas uses. For race the torsional rigidity of the plate works better for me. In training however, I like having some torsion flex in the shoe, especially in the forefoot, which is where I like Adidas’s system.
This is SO helpful. I had not really used the ground contact time metric from my Stryd pod to compare shoes… and I had not compared shoes on similar workouts. I am more of a short trail runner but I have a road half-marathon coming in Montreal. Will use your approach. Will test them at half-marathon pace for at least 2km + 2km each.
I'm glad it was helpful. Definitely try that to get a good baseline for yourself and understand your gear...then gather some data on longer runs to see how it scales over longer distance. I'm not running much volume right now as I'm focusing on shorter distances but if I was preparing for a half/full my next step would be to take the VF3 & AF3 out on long runs with pace and look at how the data differes, the assumption is the AF3 would allow me to hold pace longer before mechanics break down and I see a drop in ground contact time and stride length.
Great content, it was about time to see something like this from you. I didn't expect the VF3 to do so well and Adios 3 so bad but here you go. Also some AF3 results were completely unexpected.
The AF3 is the big shocker here. As i said, I kind of knew the AP3 wasn't fast for me, but the AF3 working as well as it is in very unexpected. Much more to do there...
This is really fascinating stuff! Certainly not an exact science but I think it's a really cool way to aggregate and analyze data to start to come to an educated decision as to which shoe works best for an individual. I'm in week 5 of a half marathon build and I'm debating between two shoes. I'll either run it in the AP3 or Cielo X1. I'm going to alternate workouts between each shoe. I'm going to use a track and pace as my controls. I'm looking to see which shoe allows me to run the similar paces at the lowest heart rate, over time. I figured the shoe that does this is probably the one that I'm most efficient in and will allow me to run the fastest over 13.1 miles. Thanks for the video and inspiration!
Yes, you understood the purpose. Doing that comparison on a track definitely takes many variables out of the mix, which is good. Once you do that I would find a route on the road and do a similar study there. Road results in those two shoes, especially the Cielo x1 will different on the harder road surface. Tracks can be overly forgiving in super shoes.
Thank you for doing this, really interesting. If I my suggest 2 calculations that would formalize this analysis a bit: First, Average Speed (in meters/minute)/ Average HR. Second, Average Speed (in meters/minute)/ Average Power (Watts). Would be very interesting to see the shoes ranked based on these two metrics for each of the distances.
Interesting. What are you expecting to see from both? Especially Watts, which I don't find that useful in running (as I find it very variable). I come from cycling originally, where watts ins the gold standard metric above all...but in running, not so much.
They will both give you a measure of efficiency: how far you go per heartbeat and the unit of power( watt) respectively, which since it is you doing all the running will give you a measure of which shoe you run in most efficiently. Say for instance, and I'm using number from memory so they may be a little off, that your marathon pace is 4.40/ km and your heart beat is 186 for two different shoes, but the average power required to achieve it is higher for one than the then, voila, one shoe is more efficient [for you] than the other b/c it requires less effort. Keep in mind these test runs were short and if one shoe requires more effort [power] than the other, then running an entire marathon at that higher power output is less efficient -- it wastes energy which you could have saved for the finish. The ratios of these two divisions should be quite interesting and useful.
I’ve run in all of the same shoes you tested (bought them all too) here. I’ve commented in other videos that of all of the shoes the alphafly 3 has just felt the best. Honestly I can’t get into the edge Paris, I’ve run some big workouts on the track with them. Did 21 ks this week (20 mins at marathon pace, 20 mins at HMP and 10 mins as hard as I could go after those reps) I just don’t love the edge Paris. Yes it’s fast, but I feel like I’m working so hard in the shoe to make it move the way I want it to. The alphafly 3 for me was my BQ shoe and I just feel like it’s the best shoe on the market for the marathon. like you I was a vaporfly guy since I absolutely hated the Alphafly 2, I’m super happy I gave the 3 a chance.
I’m also super happy I persevered to actually finally get a AF3…. It is the real deal. Agree on the MSEP…. It’s so much work, attention to mind the stability and recovery is still rough for my legs (though much better than FFTurbo).
In my opinion the control metric to compare shoes is watts. Your watts between shoes were too varied for a fair comparison. I would like to have seen how each shoe performed at the same wattage point-that would be a controlled test. Bicyclist understand this concept very well. Runners should as well
I come from cycling and lived and died by watts. However, I’ve never found them useful in running. I tried to train by them years ago and didn’t find them as valuable as I did in cycling. My watch reports them so I look at them as a secondary metric or curiosity.
Great vidi. Takumi and Vapour doing what is says on the tin, as shorter distance specialists. With Saucony outright as a Marathon distance option. Adios P3 showing its age
Interesting review-it's great to see some data, even though the sample size is anecdotal. I’ve conducted similar tests (1K repeats) with more advanced running gait sensors and can share some insights that align with your observations. Regarding the AF3, you mentioned "floating more," and that's absolutely correct. My tests also show that the bounce (a.k.a. "flight time") is the highest I've seen compared to other racing shoes. This shoe is specifically designed for "stride runners" or "aerial runners" like myself, which naturally results in a slightly lower cadence and a longer stride length. The fact that you like the AF3 likely stems from Nike's efforts to bridge the gap between the VF3 and AF3. Given your short stride and high cadence at speed, you're clearly more of a ground/cadence runner, as you correctly pointed out. As for the AP3, it’s not performing as well as you expected, and I'm not surprised. I’ve had similar results and ended up selling the shoe because it didn’t feel right for me, unlike your experience. While some pro runners might favor it, my results align with yours-I believe this shoe is overrated, though this is just my personal opinion as a non-pro.
I started that "float" work in the Metaspeed Sky+ last fall, but it never felt right, I did see a few cms in my stride length but the shoe beat my feet up too much to be useable for anything other than drills for me (I documented this on the channel back then). I've been working on it since though with drills, mechanics and strength building. Once I (finally) got into the AF3 I could immediately feel it paying off. Also, oddly, the Takumi Sen 10 is another shoe I really feel it in when I am hitting a clean strike with fast feet... partly due to how direct that shoe is and how light weight it is... it feels great when I get there in them. All this data was gathered with an Apple Watch Ultra 2, which is what I mainly run with. I do have a Coros Pace 2 and Coros Pod 2 for more data points... I may break that out later this fall on the track to do a smaller followup to this video. We will see. As for the AP3... I've known it's not fast for me for a while. The main issue is the late stage rocker. I respond to a rocker point much earlier in the midsole. When I tried on the Evo 1s and jogged in place in them...I immediately noticed the difference and understood why. The Boston 12 has a rocker point similar to the Evo 1 and that shoe feels much more natural to me. Nike's rocker points always have worked well for me since the 4%, though the Evo 1 pushed it back even more.
I love the topic and the approach you took. I would have liked to see a "fitness factor" where you adjusted for your calculated fitness on the day of each workout. I would also caution anyone drawing firm conclusions from N=1 runner and M=1 workout for each shoe.
Ya, done over this longer period of time I agree it was needed. it would have been minor, but needed. Interestingly, during this data set I was going through a transition in my running, lowering weekly volume. So according to Strava my fitness was dropping (which is why I don't pay attention to Strava's metrics) though my own performance was feeling faster and faster...
Great video that then prompted me to go and look at my data. I am comparing a 4x 4k LR workout @ MP in AF3 vs 6,5,4,3,2,1k LR workout @ MP in AP3. WU, Recovery, CD and course are the same in both. Workouts were 1 week apart; the 4x 4km came first. For me, I achieve better pace, longer stride, lower GCT in Adizero Pro 3. And my legs feel much less beat. I guess the ‘for me’ is doing a lot of work in these comparisons -different biomechanics suit different shoe mechanics.
Totally, this stuff is deeply personal and all around your own biomechanics. If I had to guess I'd say you respond better to a later stage rocker in the forefoot and a lower (more stable) plate configuration.
These are always so interesting for me that so many ppl have their range so close together. Like you mentioning 3:30 pace is almost a full out sprint, and a 4:14 is marathon pace. While my 200 rep pace is 2:40-45 pace but my marathon pace is 5:09 ( which should have been about 25-30 sec faster if I hadn't cramped up in the last 6k) but even than my 5k pace and my marathon pace would be 50 seconds apart, and my speed work would be another minute faster than my 5k pace. (And I'm still feeling slow on the faster efforts)
I’m working to getting my 200/400/800 pace more in range. I’ve never trained speed before. Until recently I rarely ran under 4:00/km, even in marathon training. As I mentioned at the end of these I’ll have a video out in the coming week or two talking about my plans for the fall/winter and what my focus will be. I’m excited!
Thank you for making this video, it's insightful to see your data on all these shoes. I feel a better way of ranking the shoes for Effort would be a variable that took into account speed and heart rate over just heart rate, as in some case you rated shoes that were slower but had less heart rate over shoes that were faster with slightly highere heart rate, seems like thats just a case of being faster on the reps making the heart rate higher than the shoe making any difference. Also measuring heart rate on 2x200m reps seems so completely meaningless, as the measure will lag behind for most of the rep, which means the heart rate at the start of the rep skews the data massively. I very much doubt running at the same RPE you have a difference of roughly 20bpm over 200m just because of the shoes. I also have some questions about the accuracy of your data entry and data in general. Some of the stride length and cadence numbers really dont match the speed, an obvious example of this is rep 2 of 200s, 202 cadence with 1.5m stride length is slower than 195 cadence with 1.4m stride length. There's some rounding the the stride lenth but I would've put in the addition significant figure as the difference look way bigger than they probably are when you only use 2 instead of 3 in this case. Another questionable bit of data is the Peg plus stride length of 1.1m in the marathon intervals, with a cadence of 184 basically running the same pace as the alphafly with a 1.4m stride length and 181 cadence, so roughly 25% (?!) longer stride length in once shoe for roughly 2% cadence loss equals less than a 1% difference in speed? These are just the two main examples that stood out to me but there's quite a few others that seem to be borderline on not making sense. The scoring system is another thing I would change, I would use something that scales based off the performace of the best performing shoe, probably a percentage based system. In cases where many shoes rank closely to each other this means they score nearly even points where as in your system there can be a big gap in points for minimal performance gains.
Fully agree on the HR for 200s, it definitely lags. But was trying to be consistent. It's an ok "ballpark" figure. I do agree that the proper way to do this would be blending some of the metrics and weighting them. Ultimately that is what would have been done in a proper study. However, for the point I was trying to make in this video, on how people can begin to disect their own metrics, I wanted to pull out pure metrics to do that. GC/SL was averaged. I run with a Apple Watch Ultra 2 and the Ui for GC/SL in the Workout's app isn't the most precise. I do have a Coros Pace 2 and Coros Pod 2 which would be much more precise. I will likely revisit this type of video in the late fall/winter, on the track and will consider using some other data loging, as well as a stopwatch and track markings.
I think those who are talking about the data "splitting hairs" are missing the point that these small variences over the course of a marathon, 5k, 10k, half etc is the difference between a PR or not. course condition, weather conditions, fuel stations, pacers etc are all variables that the runner can't control but we can control the shoes we lace up on race day, why wouldn't we want that extra edge if we can get it? Is this data going to end up in a peer reviewed study? No, but it is good enough for a runner to make a solid decision on which shoes are race day worthy for them. Great video as always!
Well said. It very much is a personal baseline that is used to validate what you are feeling. From there, in a smaller subset, you can be much more detailed and exacting (if you require that). Using data to validate what you think you feel is very much overlooked.
I just searched for "stride length" in your channel and didn't see a video on that specifically. I'd love to watch a video on why and how you've lengthened your stride. I'm currently at 1.2 during marathon pace efforts with a cadence of 185-190.
I don't have a video on it specifically. I talked about it a bit in my NYC Marathon training series last fall, ep 1 & 2 and then ultimately the problem I ran into with the Asics Metaspeed Sky +... but beyond that I've not done any specific content on it. Mostly for the past year I've been doing simple things…. High knees walks and strides, A skips, hill running (& trail running for strength), box jumps, focus on fast feet especially with 200s and 400sand general foot strike and mechanics mindfulness (up on my toes). That has made the difference really.
The "200" reps seems all over the place for distance and all well over 200m. In fact one is almost 300m. So are all these in different locations? Were they in a straight line? If so would wind (or even graident) have effected the times. So bit confused what we are actually comparing here.
The data in general is all over place. I had a quick scan over the stride length and cadence numbers, they just don't match up to the paces. How can one shoe have a higher cadence and stride length yet be slower in pace??
I think because the watch calculates these metrics and they all have significant error bars. The precision of the number obscures the actual blurriness of the measurement.
@@brianreiter5572 I think he might be using a stryd pod or something similar, I think I would discount all the data gathered by it in them categories as it seems all over the place and can’t really draw any meaningful conclusions from it. I feel like he should’ve used a heart rate to pace scaling metric to determine how efficient the shoes are for him. even then some of the heart rate data appears to be junk too, I very much doubt there’s a difference of ~20 heart rate between shoes running 200m reps. I appreciate he’s put a lot of effort into this video but any conclusions drawn from the small amount of data and quality of data is going to mean very little.
Thanks for posting this data, surprising to see how hard you work in the takumi sen in terms of steps and stride length in order to achieve the speed, which obviously reflects in the heart rate. I really want to see Adidas use a more efficient foam, whether it be a different TPEE or Peba going forward. Lightstrike pro is reliable and trustworthy but it’s really showing its age this year.
LSP Evo foam (whats in the Evo 1, the core in the PX2s and likely in the upcoming AP4) is a PEBA foam...so it's coming. LSP, supercritical TPEE, is very good and very fast. It's more the geometry of the AP3 that doesn't work for me. The TS10 is basically a modern racing flat, so it also make sense it requires more work, less geometry "helping" you.
Very interessting approach... So far I used an entire differet approach... I used the same route and compared the shoes by pace/heartrate/distance- pace drop+feeling (for example fit gets lose performence decrease a lot) Since my approach is not to find the fastes shoe its worked out kinda good so far. In my approach I try to find the best shoe to get me through the marathon that holds me fast while keeping the heartrate low. And to my suprise compared to your video the EP4 is one of the worst there, after 15km it drops in all metrics - in my case. Followed, as in your example with he AP3. I'm gonna do the same with the Li-Ning Feidian Ultra 4, Asics MSP and VF3 the following two weeks to find my marathon shoe for Berlin.
This data was all gathered on the same route, same warmup. same cool down. Though, I often approach validating and testing shoes they way you do. Interesting about the EP4. Any idea why that is the case for you?
@@SagasuRunning Im guessing it has something to do with the very late rocker combined with the very little bounce (compared to other racing shoes). The AP3 has in my case the same problem as the Feidian ultra 4, when the fit gets some how lose the Performance drops massively. I could explain thah with the rods and the huge cutout of the AP3 and the arrowhead shaped carbon plate in the ultra 4. So far the SC Elite v4 worked the best for me even tho I actually dont really like that shoe, which supports the claims that the SC elite v4 is very suitable for slower runners. Going to test the VF3 today, I have very high hopes 😊
Yup, you nailed it. The geometry of the AP3 has always been an issue. I don’t do well with late stage rocket. Luckily they resolved this with the Evo 1 and the AP4 looks to have that new geometry as well. The SCEv4 sounds a lot like the EP4 for me. It works but I don’t always enjoy it. Good luck with the VF3… prepare for the soft foam… you’ll see what I mean.
@@SagasuRunning I crawl back to you after a 25km run and have to confirm... The VF3 is THE GOAT. Such an amazing shoe... Pushing you but doesnt force you to push, way more stable than I expected... Not super soft, but far from being firm, maintains a very great cadence. Absolute amazing, maybe there are faster shoes out there, but in terms of versatility this is THE shoe. The last 5km felt horrible and really taxed my ankles, also walking feels arkward in it, but that is my Fitness Level and not the shoe.
@Chungdol See…. I told you! Stick with it. Since there is so much sink in ZoomX is your ankles and calves are not used to that they need time to adjust to all the micro movements/balance they do. Also agree on the VF3’s versatility. They may be faster specialist shoes but overall you are never going to be unprepared in a Vaporfly.
My advice is to avoid Strava's numbers and go straight into your watches's app and work from there. Strava often tweaks and averages things I have noticed.
Simple things…. High knees walks and strides, A skips, hill running (& trail running for strength), box jumps, focus on fast feet especially with 200s and 400sand general foot strike and mechanics mindfulness (up on my toes). The AF3 (high plate configuration) definitely helps. I’m now curious about the Metaspeed Sky Paris for the same reasons…
All of the early testing with super shoes found that there’s no running economy difference between a shoe w/ a plate and w/o a plate given the same resilient foam. The reason a plate makes a difference as a stiffening agent is that they are designed for a marathon. The stability issue with soft shoes and lower leg stabilization/hip stabilization won’t be noticed in short reps of 5-10 minutes. This is why the Pegasus plus is performing well in these tests. To be true data, these should all be done closer together, and more than once. So doing it all over again and changing the order of the shoes would be the best. The Vaporfly studies were done in multiples
I remember the 4% studies back in 2018/19. Things have changed. Super foams have improved significantly. Yes, the plate IS a stabilizing agent but in a 2024 super shoe it is doing more than it did back then. The spoon of the plate, especially a Nike fly plate (which is more flexible than others) is adding to the propulsion. The magic of a super shoe is how the plate AND foam work together. The efficiency gains of a plate are over the longer distance. For 2k of MP I’m not really surprised that the Pegasus Plus holds its own…. Over 42k it would be a different story. If I was publishing this data in a proper study I would take many of the variable out of the equation and do more rounds in each shoe. That’s not my purpose here. I’m showing how anyone can use their own data to validate what they feel when running…. Or find gaps they didn’t realize. From there you can narrow down the shoes and really look at the data if you wish to fully confirm.
I don't have supershoes to compare as you did, but I think if I ever go, I will go by your data. I am a high cadence ~ 180-185 runner with forefoot-midfoot strike with a mild supination while running,, so You is me, me is you. Haha
Amazing video! Interesting numbers you have between those videos. Surprised there is a power reading (I assume you use Stryd?) but no LSS or Impact loading rate. The recent collaboration between Stryd and Luis Orta showed that these two metrics can also add to support the perceived effect from shoes. I understand these kinds of videos are hard to make, but hopefully there is a follow up with additional metrics.
Apple Watch Ultra 2, which has onboard power, though not as accurate as what Stryd does or how Coros uses their foot pod. I come from cycling, where watts are the gold standard for performance...but in running I've never found them as useful. I tried training by power years ago (when I first came over from cycling) as I understood it well and it removed climate from it all (NYC winters)...but it just never worked for me like cycling. SO I glance at it as a curiosity and secondary metric but don't use it for anything really. I haven't watched Luis in a while...gotta go check what he is up too. Thanks for the reminder.
summarizing the data By intended use, (i.e. Marathon vs 5K), seems like it would give you a different shoe for each use. The summary you gave would say that if you could only have one super shoe,, for you, it would be the vaporfly, but what would it be if you could have 2, 1 for the Marathon & one for a 5K?
The summary was abrogating across all sets…. Looking for what was best across everything. That being said. I know I can do marathon pace and distance in a VF3, but the AF3 will be even better for me. For a 5k I would absolutely go for a VF3, even a VF2 if I could get my current size (the one I have is a half size too big). For intervals, 200,400,800…. I’d likely opt for the TS10 for the shorter ones and the VF3 for 800s/half mile. I’m actually going to talk more about all of this in the next vide on the channel August’s “Shoe Talk” as well as the other shoes I’ll be using this fall winter. Stay tuned.
The Peg Plus has replaced the Streakfly for me. However, when we have the Streakfly 2 I'm very curious where that sits for the 5K pace and 200s... I can't wait to test that shoe.
Would be interesting adding some racing flats in there 🤔 my fastest 5k parkrun times are in saucony fastwitch 9, endorphin pro 3, and puma deviate nitro elite 2. Not much different in times, all around 20:45, sometimes it feels like the high stack gets in the way and I miss the minimal feel of the fastwitch 9’s at short distances. I’ve just got some Takumi sen 10’s to see how I get on with them.
That is coming....I'll be adding in a few this fall/winter to a revised data set I'll make a video on over the winter. It's going to get so interesting.
I have VF3, AF3, AP3, TS9. Gotta say, the VF3 might be the 21k and down shoe. The AP3 I ran my first and only marathon in last fall (2:52xx) I’ve done 2 runs in the AF3…. Today was a 10k tempo but was too hot to give a fair judgement, I did do some vo2 max reps last week in them and they were really good… will decide over the rest of this block if I use the VF or AF for this fall marathon.
Before I did this the VF3 was my all road racing shoe, but now its likely 21k and under. AF3 is very obviously better for me for the full. Under the 5k though....I'm not sure, Vf3 or TS10... or likely the upcoming Streakfly 2...
This is really helpful and I'm going to look at similar data for myself to see how my own shoes grade out. IIRC though you have a lot of prior experience with the Nike shoes. Do you think perhaps your running has adapted to their properties creating an unintentional bias of some sort? They're obviously awesome shoes, so I don't mean to say anything is wrong here. I'm interested in whether it's a thing one should consider for themselves.
Absolutely! 100% I've adapted to the "Nike feel" which is very specific in my mind (I'll have a video about it in Sept). The more universal issue is I respond to rocker points pushed well behind the ball of the foot. The AP3, MSEP, EP4 all have later rocker points compared to the VF3/AF3... and the TS10/PP don't have much rocker at all so that is a more direct drive anyway. When I tried on the Adidas Evo 1 in store and jogged in place I immediately found that that shoe would work much better for me as the rocker point is pushed way back compared to the AP3.
What workouts did you do to lengthen your stride? I feel like that's the next thing I need to work on before these super shoes even make any sense for me. My cadence is good, just my stride length is very short.
Simple things…. High knees walks and strides, A skips, hill running (& trail running for strength), box jumps, focus on fast feet especially with 200s and 400s and general foot strike and mechanics mindfulness (up on my toes).
@@SagasuRunning My takeaway from your video is that shoes are not only feeling different, but that there’s actual differences which are somewhat small yet very real. I am inspired to a similar test protocol with the different shoes I have; let’s see how much these differ!
It’s faster than either for me. In one of the earlier data sets I ran a 8 1k @ 3:50/km and 4 x 200m @ 3:30/km… and it was a second or two faster with similar metrics otherwise. It is the fastest option….
Fully a geometry thing, which I’ve known for a while. I don’t respond well to late stage rockers, the AP3’s is very late. The TS10 is a modern race flat, that emphasis clean foot strike much more.
Making a few comments as I watch. So Set 1 - is the notional 2K the exact same circuit from same start and finish as the distances vary especially for the AF3 which is actually 2nd slowest on time but you have ordered it 2nd fastest on GPS pace? Also with the HR how much running did you do before as presume "fresh" on a 2K, HR is going to start low and rise but maybe at MP which just flatten out.
All of these were running the exact same route, flat, along a river…. Nearly same time of day, same weather, same 2.6k WU (to get to the river paths). The didn’t distances is me thinking on starting stopping my watch at the same light pole…. Turns out I’m not. Workout was always done in the same order as well.
@@SagasuRunning So I am just thinking when you timed yourself at 8:45 in the AF3 but say 8:35 in the Edge Paris it is not necessarily exactly the same "route" ? When I did this sort of thing I tended to focus on the actual times as feel I can time myself more accurately "point to point" than my watch could measure distance. Especially when trying to infer things from rather small relative variances.
Awesome stuff! It looks like the shape of the plate (thinking about VF3 vs AF3 having relatively similar balance of speed and effort, with different plate; while VF3 and Paris having similar plate but very different outcome in effort) might not make that much difference after all? I'm under the impression AF3 is perceived as a shoe better suited for powerful push-off and VF3 for faster turnover. On a separate note, Deviate Nitro Elite 3 might be an interesting review candidate. I've watched your review on the DN3 and understand puma isn't high on the list for you, but looks like the Elite 3 very similar to VF3 and looks like a good fit for your running style.
@@SagasuRunning Thanks! This solved my long-time AP3 puzzle. It looks like we have very similar cadence and landing pattern from your comments in the other video (you are def faster than me :). I've been using AP3 as my main quality training shoe, enjoyed it a lot, but data always showed they are not faster or more efficient than more rockered shoes. Might have to rethink whether shoes like AP3 and Endorphin Elites are suited for me (AP4 looks like they would have a rocker that's closer to VF3 though).
AP4 has the same rocker as the Evo 1… which is similar to the VF3/AF3… if anything it’s even earlier. The Evo 1 rocker felt very natural to me when I tried them on… I’m really looking forward to the AP4. Can’t wait to test them.
Hello - like this evaluation as data based. Much said about loss of energy return of shoes with increased use. So wondering if here all were similar with low miles when testing? Thanks
This data set doesn't factor that it at all, the shoes range from close to 75miles to under 10miles. However...I have three pairs of VF3s, ranging from 52miles to close to 200miles....I could compare those. I'll add it to the video idea list, no promises and it won't be soon...but maybe someday. Great idea. 🤙🏻
I'm likely going to revisit this later in the fall/winter, on the track, with a small subset of shoes in much more controlled conditions. But yes,I agree.
Also not sure if I missed anything, but if we look at the 2k comparison between Paris and AF3, the cadence is ~3% difference (181 vs 187) but stride length difference is much larger (~9% even if we take the smallest gap from rounding, i.e. 1.24m vs 1.35m). Given the pace between the two are virtually identical, would have expected the SPM and stride length gap to be similar in scale in different directions? Not to nitpick anything but just wondering if the tracker apps or something else could be making smoothening adjustments on some of these data that could impact the results
All data was gathered with a Apple Watch Ultra 2 and puleld from the Workouts app. GC & SL are averaged per rep as there is not precise metric given...so yes, it will be a bit inconsistent here and there. I did take the lowest and highest reading I saw in each rep in averaged from there. So yes, it's not a precise as a specific metric the software is reporting like SPM.
Hmm, I have a feeling the measurement error is comparatively large, enough to affect the data and some of the conclusions drawn from it. Furthermore, as you mentioned, depending on the order in which you tried each shoe, you may have gotten a fitness boost for a particular set. Still, at the end of the day, perhaps it's not surprising that the VF3 is the shoe that is best tailored to your biomechanics and (after a surprisingly good, albeit short, tempo run on the good ol' Peg 39 yesterday) I'm almost tempted to buy a pair and see how they compare to the AP3. I'll try to hold off until the Adios 9.😅
Ya, as I said, this is data set is not totally "lab certified" and there are plenty of variables in play. But it's enough to validate some impressions in order to focus in on more specifics. Interesting, the VF3, AF3, PP workouts were all done back-to-back-to-back over 10 days (with rest). The MSEP came next. the AP3 is the oldest and the EP4 was the last one I did. The VF3 and AP3 are very different shoes. I do love the AP3, but I've known for a long time the VF is just faster for me. Alot of that is familiarity and my own biomechanics adapting to the VF feel, which is still unique in super shoes.
@@SagasuRunning I agree, I suspect familiarity plays a no small part in the biomechanics equation, to the point I have to wonder about the efficiency impact when using different types of shoes (perhaps minor if you run enough to adapt to each?). This plays into a comment I think you made about keeping a rotation within brand; if each brand considers their line of products holistically, it simply makes sense to use a same-brand rotation (tempted as I am to mix and match😄). On that note, which brands would you say excel at this, and which have room for improvement?
@OdanUrr87 Adidas’s entire Adizero range works together nearly perfectly from trainer to racer. It’s by far the best one. Nike has the Pegasus/VF combo, which is nearly perfect…. V17 to AF3 is close, the PP fills in the gap well. Honestly, Saucony also does well with the T22/ES4/EP4. Other brands are much more hit or miss.
The 1K reps all seem to have the same recorded distance so seems a more obvious ordering by time or pace could be done. Re the HRs (especially Peg Plus) was that day you ran in the Peg Plus somehow "harder"? It can be difficult to be sure on HR as some days I find it is low, normal or high for reasons other than just the pure effort. Cadence looks a lot more even in the 1K. I find the AF is the only shoe where typically I find mine is a bit lower (and hence longer stride) but it always feels like it wants you to "bound" along. I also am not sure how long I can actually sustain that. "Striding out" I always found easier when fresher. Bit surprised the Peg Plus has the lowest cadence but it is quite close and hard to be sure of any real differences in these 1K ones.
The 1k reps are done on a segment I have measured many times (also it's known Strava segment in the area). Thus the start/stop are marked on the path (multiple run and triathlon clubs use the section of bike path I did these on for training)...1k seems to be their shortest mark though. I'm eventually going to go mark a proper 200/400/800 and 1/2mile... The AF3 defintley prefers me to "bound" along... The PP cadence isn't a huge surprise to me, that shoe wants me to run flatter to load the foam more. I've noticed my cadence is always lower in that shoe.
i try AF3 again, do “run” again in store today and i have feeling that i can’t find transitions between super soft heel and air pods..by the way, good video, keep going
The heel sink and then secondary sink into the mid/forefoot is very ZoomX and a huge reason many don't like the way they feel. As a forefoot striker I never notice it, but when i do heel strike it's very apparent. But compared to the previous AFs the continuous midsole makes a massive difference in how the shoe transitions. It's much much better. I'm overall really impressed with the redesign of the shoe. As always, thank you for the support.
Because I'm running on bike/running paths along a river and don't always start/stop my watch at the right line, sometimes due to people traffic. None of these were done on a track.
interesting comparison, I just wonder what comparing BPM at the same speeds on a treadmill would look like, this would take the speed variances out of the equation. I personally have to admit I go mostly by feel, but I am also much less tolerant and have to exclude a lot of shoes because of incompatibility with my biomechanics and anatomy. I would currently go VF3 as well, but as you say everyone needs to find whatvworks for them, and the Adios Pro 3 obviosly still works for a lot of elites that can do lab tests to choose their shoes.
A 'mill would make it much more exacting. Back in my cycling days I did nearly all of my important training indoors on rollers or trainer as I could control all the variables. However, I don't have access to a 'mill here in Taipei, refuse to do gyms and would rather deal with variables outside as a runner. Also...I'm not saying the AP3 is bad....it's a great shoe. What doesn't work for me in it is the geometry. Late stage rockers just don't gel with my run form. The Evo 1 fixed this and the AP4 has that geometry...so I expect it to be far faster for me. I can't wait to test them....
The Edge Paris is stiffer due to the plate configuration. FFTurbo+ is a firmer foam, even after it breaks in and Asics uses VERY stiff carbon fiber plates in their shoes.
Depends on if you are looking for a new shoe. The VF3 is much softer and a very different type of ride than the MSEP, even though they are both low plate configuration shoes. Personally, I do think everyone should experience a Vaporfly. It is the OG super shoe and still has a unique ride in the space. Though it can be hit or miss and is a big investment for a shoe one might not like, especially if they started with the current style of super shoes from another brand. The MSEP is fantastic though, so you aren't lacking anything with it.
Very much so... but I don't have access to a mill...not do I think I could stand running on one (unless it was middle of winter and freezing cold outside). Even a track would be better for this. There are plenty of ways to make this more controlled, but all workouts were done on the exact same section of road (straight bike path on a river) with the same warmup and cooldown. It's enough to get some basic understandings, validate some feelings and impressions and work from there (which I'll be talking about in my next video).
I think you are exaggerating the role of carbon fiber plate when you are comparing PegPlus and Vaporfly. The ZoomX on PegPlus is a bit different than Vaporfly/Alphafly and the amount of foam in Vaporfly is significantly more than PegPlus. The weight of the PegPlus is also a drag. In other words, I am saying there are more factors beyond CF plate in contribution to the performance. Anyway, it does look like Vaporfly 3 is the shoe for you overall if you were to have just one shoe but you might benefit more from say AF3 on a marathon since you already have it. Question: What did you (exercises, types of run, stretching, etc.) did to increase your stride length? I can benefit from your experience if you share. We are in the same age group and I am a very high cadence runner (200+ on VO2max and sprint type intervals) but I think if I can cut back on cadence and increase my stride length a bit, I would be more efficient.
I agree that it's more than the CF plate...but that is the thing people ALWAYS talk about and the general thinking out there in running presently is that you NEED a plate (super trainer, super shoe) to go fast and the new class of performance trainers coming out (of which the Peg Plus is one of the first) is somehow not capable. I've been finding the peg Plus beyond capable as a training option when I don't want a plate, which is often. Honestly...alot of it was form work. Making sure my foot strike, forefoot, is clean and powerful AND letting my body "float" before I try to drive my next leg forward. Alot of repeating "strike and float" over and over doing strides and faster running. Spending time in the Metaspeed Sky+ last year also helped. Drills have been A-skips and plyo (box jumps up, box jumps down with a explosive hop using your feet and legs only - no knees hips - as soon as you contact the ground). Fast feet/high knees hill running as well, knee drive work as well.
Interesting stuff. I have the feeling that the AP3 resists acceleration period and is hard to take above threshold pace but feels good for cruising. That’s a lot of feels. But empirically any super shoe I have tried is lower HR to maintain a pace than a daily trainer. I wonder how much correlation there is to simple mass/inertia effects especially at faster paces. For example at the risk of oversimplifying, the TS10 is essentially the same materials as a AP3 but it performs a lot better at faster paces for you. Lighter is better. Foam, plate, and geometry have to be very good to overcome any weight penalty. Data is great but also it is important to keep in mind the error in the measurement - something pounded into me in college. I’m dubious about the ground contact time measurement as a direct performance score. The really meaningful measurement you are trying to get is effort to maintain a pace or pace efficiency or velocity at VO2 consumed. HR is a slightly indirect measurement of aerobic effort. Pace is pace. Does GCT really mean mechanical efficiency especially within the precision and accuracy of a watch to capture? I just looked this up for Garmin and they state a mean absolute error of 5.4%. That suggests the range of your measurement distribution is within the error of the measurement. I also am curious how you are capturing HRM. If it is an ECG HRM then I’m sure it’s very accurate but if it is captured from the wrist then its error range is going to be large. How did you capture pace? Capturing average pace over 1km or 2km with a multi band GPS is probably pretty good for this propose but for 200m it’s definitely not good enough but lines on a track would be. The absolute number of test runs is pretty low and the differences are also not really big so any measurement error could have large effects. Everything is also driven by RPE which is kind of a squishy concept that is hidden by the precision of the measurements that you end up capturing. I’m not actually suggesting that your results are wrong at all. I think that you are validating that the AF3 is very good for you across a range of paces and the AP3 is decent for marathon and bad for 200s. But I am saying that interpreting the huge amount of data we capture so easily is hard.
I've known the AP3 isn't "fast" for me for a while. The issue is the late stage rocker on it. I respond much better to early rockers. When I tried the Evo 1 on and jogged in the store (which shocked the Addias rep as they were demo pairs) I immediately noticed the rocker point much further back on the shoe and it felt more natural (like a VF). The AP3 is great for locking into a pace and looping along in it, for me at least, but accelerations take alot of effort. I captured all of this data with an Apple Watch Ultra 2... so dual-frequency GPS (which I find very accurate for my needs). HR data is from the wrist sensor, which i know is not as accurate as a chest strap...but after YEARS of chest straps in my cycling days and then moving to wrist top as a running I find wrist top sensors that Apple and Coros use (not Garmin) are "close enough" for me. I know they aren't dead accurate, but I have worn the Apple Watch since it's day of release (and other Apple watches and Coros prior) and trust the consistent of the readings for the data. For anything under 1k I definitely thing a GPS is not ideal. Track mode on the Apple Watch Ultra 2 and Coros Pace 2 is decent...but not as good as an old stopwatch to the lines on the track. I'm going to come back to this later this fall/winter with a smaller subset of shoes on the track and will likely use a mix of GPS data and stopwatch data...as that will be a 200/400/800 data set.
Given this I think there is not a large enough data set to overcome the measurement error of oHR, pace/distance, and GCT. I am pretty sure the measurements all fall within the error bars of the tools doing the measurement which means and you don’t have a large enough sample to make robust conclusions from the data. The average error for each measurement is probably +-5% or more which is more than the difference between the measurements. The watch is averaging a lot of individual measurements and in the case of an Apple Watch applying an ML smoothing model. The most accurate measurement is probably the average HR and pace over the 2km distance because there are the most data points being average. In principle the measurements should tend to cluster around the correct values so when you average a lot of measurements you are converging on a good number. If you divide the distance covered on each “2km” test (averaging a 500 or so GPS points 1s apart) by average HR (averaging 500 or so oHR measurements) you can calculate a distance covered for effort value. At 2 significant digits all of the test runs come out to 0.013 km/heart beat/min - which is to say they were all the same effort. (Coincidentally that kind of validates the RPE concept.) To put it another way these are all really good, even the “control” shoe, and the differences are hard to tease out without lab equipment. Even the control shoe is too good. The conclusions are hard to separate from confirmation bias especially since the testing was driven by RPE. I think they are valid opinions but I also think the degree to which the conclusion is strongly supported by the data is overstated. You would need multiple tests at longer sessions simulating the actual target race to develop a data set that is rigorous enough to convincingly tease these shoes apart in a way that is improbably impractical to actually go out and do.
@@brianreiter5572 If I was going for lab quality data and a double blind test for publishing I would worry about all of this. To gather a data set to validate what I feel, even if not 100% accurate, but consistent with the same watch/course/weather/time of data and overall effort...this works for me and I would encourage anyone else to look at their data like this. At least to inform themselves of what they feel in order to narrow down on a subset of shoes and really focus on them...longer runs, more varied data, etc.
@@SagasuRunning 🤣 RPE, "comfort filter", and pace for effort can get you very far in evaluating products that work better or worse for you, absolutely. I have a personal have a bugbear about watch metrics, which is not about you. I'll offer a couple of examples here which are my own grievances and specifically with Garmin because that's the devil I know best. The basics have more noise in the measurement than the watches absolutely misrepresent with false precision. GPS points are each only accurate to about 2m in good conditions but the distribution averages out around the correct point and as a result Garmin's current pace field is very poor but 1km pace is quite good. The Garmin VO2Max estimate is super-terrible. Garmin published their validation and claims +- 9% accuracy under ideal conditions with correct weight and good GPS and pavement when used with an ECG HRM. That means if you get 50 it is between 55 and 45 probably but in reality it is much worse than that in many comparisons for lab tests. There is an entire suite of Garmin physiology metrics that are derivative of that number but including the 10ths of a point which are in the FIT file but not displayed. That's basically all nonsense and literally generating recommendations on noise. When I realized this is just a silly number generator, I found the way to turn off the display of these metrics because they are objectively unhelpful. Anyway, I did enjoy the video and your effort to take an empirical approach. It's all good and I know it is a ton of work.
You know, it’s interesting that you mention Garmin. I left the Garmin ecosystem years ago for many of the same issues you highlighted. it started with cycling head units (moved the Wahoo) and when I started running more seriously I noticed a lot of the data inconsistencies from Garmin, port GPS mapping and all their “training analytics” always seemed wonky compared to what I knew about myself from years of cycling. The whole Garmin Connect platform going down for 90+ days back in 2020 was the last straw (I mean who has their production environment connected directly to anything that would allow a “hack” to hit it from the support team, 🤔). So I left Garmin and never looked back and have never had issues with watch data from Coros or Apple. I do look at some of the blended and averaged metrics with a grain of salt…. But overall I trust what it’s reporting to me. At least inasmuch as it is consistent day-to-day…. So even if it’s off it’s off consistently and I’m working from the same point.
I ran by RPE. So the goal was to run said pace, but running at the RPE turned out to be faster. Fitness was better than I expected, at least for a shorter workout like this.
Your 200m time seems very slow compared to 5K and MP, not sure they are accurate based on that and therefore wouldn't this throw off the intended prediction of the reps. Also I'm always surprised when people think marathon shoes would be slow at any pace when pros are at 4.40 per mile
I use the VDOT Calculator to determine paces when training. 44secs in a "interval" pace for a 200 per my MP and 5K pace. So yes, slow...but in the general range. 200/400/800s are something I'm going to be spending much more time with this fall/winter, mostly on the track. There I'm looking more for 37-41secs on the 200, but that is in spikes. I'm not surprised that a shoe like the VF3, designed as an all-rounder, covers this work...but I am for the bigger and bulkier shoes optimized around the Marathon.
Can somebody please explain to me how you can have a higher spm and a longer average stride but run slower? I think it’s safe to assume that the measuring equipment doesn’t have the precision required for this testing.
Stride length was an average. It is imprecise but gives a general ballpark. There was a range of SL across the entire rep. I took the shortest value and longest value I saw in the rep and averaged it. I will go back in a few mo the with a foot pod that will be more accurate.
@@SagasuRunningcould you also include more info about when each shoe was tested because like you said your fitness improved over the 4 months. Also how many miles are on each pair of shoes. I’m sure you know some shoes are better right out of the box vs some need a little break in time to be their best.
I still love it...but it's not a racer for me. I'm so curious how the AP4 will suit me. There should be some changes to that one that I will respond too well....
@@SagasuRunning On the Ascis Metaspeed Edge Paris, I am very tempted to try it as the racing shoes for my Taipei Half. What do you think? So far, I've only 1 Ascis shoes, i.e. Superblasts.
I expect some news in the coming weeks and possibly a very limited release at Berlin (like they did for the Evo 1 last year) but I’m consistently hearing Jan/Feb 2025 now from all sources.
Coming from the AP3 you will notice ALOT of instability… and the Taipei course has some iffy pavement (especially the bus lanes on Renai Road)… but form a foam standpoint you will find FFTurbo+ very familiar coming from LSP. So you have a pair do MSEP yet?
It gives reliable results for me, which was the point I was highlighting in this video. How to use your own data to look at what works best for you as a running. I explained my data, how I structure it and analyze it to help other do the same for themselves.
"This video is going to be mostly me voicing over spreadsheet data. So, if that's not your thing, I totally understand." Clearly, you don't understand the audience you've cultivated... 😅 ALRIGHT, LET'S GO!!
Yup, I've been based here for nearly 4 years... I never plan to put a AF2 on my foot again, that shoe just makes me think of plantar fasciitis and I never want to go back there.
I really like your approach and thorough testing. However I think some of the data is so close that the statistical variance is way too low to draw any conclusions except for some obvious. I run both Nike VF, Asics MSP and Adios Pro 3 and yes, the VF and MSP are very fast and aggressive but I think the Pro 3 will come to be a benefit after 20K into a marathon. Keep the video's coming!
I agree. Data set is too small. Conditions variable. Fitness evolving. Etc etc…. But it is what it is. You can still make some informed judgements.
The AP3 is a disappointment…. I’ve done a lot of long runs in them, but the AF3 is hands down the better option for me. I’m also so curious how the AP4 will fair… especially given the tech they I think is in it. I can’t wait to do those tests.
@@SagasuRunning Curious of what's your AP3 mileage?
@q3bing On the pair in the video, about 115miles all told, they are also well over a year old.
@@SagasuRunning what's your view on the aging of super critical foams. One insider in the industry told me that it's more the aging of the foam that makes a super shoe as these perform less than wear and tear
It’s definitely a thing. Aging of any foam is a thing, super foams go “off” faster. Some have shorter shelf lives that others. ZoomX is on the lower end, LSP is on the higher end. One of the things interesting about Puma’s A-TPU nitro foam is that it should be stable.
It’s a fine line right now. Current generation foams are being pushed to their breaking points, literally. The next generation will over more avenues of development, but we are still a few years away.
Brilliant as usual Chris Sagasu. Took the AP3 for 18 yesterday, nearing 200 miles. I have felt the same way for a while and agree with you. Once a runner churns lightstrike pro into butter, the feel of the ride is magnificent and exciting, however, the bounce or energy return is a little more muted than the competition. It feels so nimble and if you’re in control that I understand why the pros love it so much. My conclusion is that it’s a training super shoe I can use for my MP long runs in the same way people train with non plated shoes , but then get a little boost on race day. The AP3 is obviously more efficient than a non plated shoe, but it’s somewhere in the middle, closer to the plated side, but not quite my race day selection. 🎯 you always nail it
Spot on. The AP3 is the best 'training' super shoe out there. I don't think I'd race in it, but use it for all my sessions. So incredibly durable and because of that actually pretty reasonable value for money, particularly as they go on sale fairly regularly.
The AP3 is such a good long run trainer. The further forward rocker point just slowly rolls the shoe forward and LSP is magnificent for durability and recovery. Also, to be fair to the AP3, it is over 2 years old and a full generation behind the other shoes on this list. Things have changed since the shoe was designed. I also think a runner with a longer stride will benefit from the AP3 more, I'm definitely a cadence runner and respond to shoes with much earlier rocker points (VF3, Evo 1, etc).
Fully agree. 💯
LSP is also so durable the AP3 is a better "long run" or tempo shoe over the B12 for a dedicated marathoner. You get a lighter and more performance oriented shoe for the work you need to do. The B12 is great, but better for more general running.
@tomverdich184 Agree 💯
Definitely agree. just go mine to 115 miles and besides some dirt they look exactly the same!!
Splitting hairs based on differences that are within the measurement error for each of these metrics seems like we're missing the forest for the trees. This data suggests to me that all of these shoes more likely than not have no measurable effect on performance and that we should probably just go with what is the most comfortable and "feels" the fastest.
Actually...I agree with you. However, I do think one can still "split hairs" to confirm those feelings. As I said in the totals...I wasn't really surprised to see the final order, it was almost exactly how I would have placed the shoes prior to looking at the data (the AP3 would have been higher up though). So it was good to see the data confirm those feelings. That is the point of all of this.
@@SagasuRunning I'd be interested to see if your ranking stays the same with a bigger sample size. When I did similar tests (fellow spreadsheet guy), I found that the shoe I was wearing just wasn't important. I got so deep in the weeds that I had a formula that controlled for wind direction, elevation gained, and my average weight over the course of the run (pre and post run weigh-ins required). I even resorted to running all my tests on a track because the gps error from my garmin was enough to skew results. Eventually I had to shift my mindset from seeing my tests as a failure to looking at the bright side that I could just wear whatever shoe I liked wearing the best : )
@Jruns With any super shoe analysis it’s always splitting hairs. The main benefit you will find is what super shoe works for you at what distance. While it was a great video you should take the thought process and idea behind it and try to find a way to analyze your own running with it. None of these shoes are going to be the best or be all end all for everyone. It will be what works best for him and his running.
Ha! Way more analytical that I am. I’m a designer…. I like broad brush strokes. 😂 I’d be happy for the data here to be used in a larger study though.
@cortneywebb1677 Very much… it’s very personal…. However, larger multi user data sets do offer different and interesting insights..I often work with enough data scientists to learn that.
Wow! There is a lot of effort, work and running to put together this analysis! I was not sure I would watch this half-hour video but I gave it try and it was more than worth the time. Great job!!
Ha! Thank you for making it through. It's doing surprisingly well. Hopefully people are learning things they can apply to their own running. 🤙🏻
Your overall conclusions are interesting. I feel there is are two sets of shoes (VF3, AF3, Edge Paris) are clearly "better" for you than say the Pro 4, Pro 3. I think the Sen 10 and Peg Plus come with asterisks as the tests are quite slanted to shortish efforts so you would kind of hope the Sen 10 would do OK especially in "sprints". Peg Plus actually did better relatively lot better than you would think. Reminds me of when did 20x400 in the Next Nature. I could "get going" but maybe 1-2 secs slower than say Vaporflys but hardly matters other than vanity in training. Sometimes with these tests I feel that it is almost a validation for what you like to run in the most and also got to think what race distance am I actually looking to. When I was doing marathon training, I found doing a "shoeoff" of repeat 5K ish reps at MP in different shoes back to back not only is a very good training session but gave a little bit better insight into what I feel would best for me when tired deep into a race.
Tim Gröse comment, every shoe tuber nightmare 😂
@@markoljustina ha ha. These days I only tend to comment on the channels I like or at least the ones that actually want to engage in the comments. This is an interesting video especially as quite similar to the "shoeoff" ones I have done myself. It is harder now in these sort of tests as the differences between shoes are quite small. The results think largely align with that I might have expected from the review videos in terms of what Chris likes to run in.
Tim is my inspiration for doing data videos like this!
🙏🏻
@@TimGrosehowdy Tim 👋
Pretty interesting to see this data. Didn’t think I would be into it, but it was actually fun to see how each shoe made your workouts and running feel experience different. Thanks for the video.
I’m glad it was interesting. Thank you for watching! 🤙🏻
The second I hit play on this video, I could already guess what the comments would be. Something like, “But there’s not enough data, and you didn’t control for the environment!”-as if this is going into a peer-reviewed journal and not just a chat about shoes and their use. As an econometrician, I love seeing people think critically about data, but it drives me nuts when the nitpicking gets out of hand. That said, I really enjoyed the video.
It work with enough data scientists and analysts to know that data always has value, it just matters on context. I appreciate the kind words though and glad that the video was enjoyable and I hope useful.
Fascinating video and loved the detailed comments on each shoe for each workout. I took screenshots throughout and even programmed this workout in my watch to test some shoes in my current rotation. The SL2 will be my control shoe. Keep up the great content!
I’m curious what some of your data looks like. It’s not a very difficult workout but it is one that will test the ranges of shoes well.
@@SagasuRunning I’m curious as well. I will be doing the test on the following: Magic Speed 3, Takumi Sen 8 (300+ miles) Atreyu Race Model, and Endorphin Pro 3. I will be starting my data set next week after my 10K race this Saturday for attempt #2 to go sub 42’ using the Endorphin Pro 3.
@mattbeier Good luck this weekend. 💪🏻💪🏻💪🏻
@@SagasuRunning Thank you! I’m feeling strong and the speed is there. Hoping the 74 degree projected temps feel cooler.
@mattbeier man, 74F would feel like goodie weather to me right now. 😂 I can’t even imagine it right now. Haven’t felt that cold since February.
Generally speaking the fastest shoe for you is going to be the least fatiguing and BPM over 2K is an insufficient sample size to gauge that assessment over 26.2 and only one factor in a much more complex equation.
What shoe can you ball strike a 5K or 10K? What's your unique metatarsal stress threshold point in a specific shoe? The carbon plate might accelerate that but it's going to take much more than 2K to get a sense of what shoe you and all runners have the skill & strength to handle for the intended design and use of the shoe.
Adidas Energyrods 2.0 is technology not made to deliver its benefit/value in a 2K measurement. The whole idea is the torsion flex will stave off the metatarsal & leg stress you'll absorb in the carbon plated shoes - over distance - and your test here does not address where Adidas shoes are designed to shine (staving off metatarsal & leg stress over distance).
Hence in reality the Adidas Adios Pro 3 might be the fastest shoe for you in the actual use of the shoe per what it's designed for, road racing distance.
Valid points.
Well...the AP3 and TS10 have the same rods, though the AP3 is carbon and the TS10 is not. I find the rod feel between them very comparable. The issue with the AP3, that I've know for some time, for me is the geometry. I don't respond to late stage rockers well. Luckily the Evo 1 solved this and the AP4 will have that geometry (plus some other tech)
As for the short distance... while I agree to REALLY know (as a half or full marathoner) you need to do the long runs and compare the data....I have many in both the VF3 and AP3 over the years and know, very clearly, the VF3 is much faster for me...though the AP3 is a more stable long run shoe.
I also disagree that Adidas's rods need a long distance to work. The TS10 is designed for short distances, with the same rod alignment and structure as the AP3/Evo1/B12 (jsut not carbon). I do agree that the benefits of rods or a plate compound the longer the distance, you don't need to ONLY run long ditances in them to find out what is working for you, at least enough to validate your feelings on the shoes and then work with a smaller subset to really fine tune and look at the data.
@@SagasuRunning >> I also disagree that Adidas's rods need a long distance to work.
In this dataset, no. In my general running and what I know will work for me over 16k, 21k and 42k…. Yes. I know a more flexible carbon fiber plate (like what Nike uses, compared to say Asics) will work for me better than the Energy Rods 2.0 Adidas uses. For race the torsional rigidity of the plate works better for me. In training however, I like having some torsion flex in the shoe, especially in the forefoot, which is where I like Adidas’s system.
This is SO helpful. I had not really used the ground contact time metric from my Stryd pod to compare shoes… and I had not compared shoes on similar workouts. I am more of a short trail runner but I have a road half-marathon coming in Montreal. Will use your approach. Will test them at half-marathon pace for at least 2km + 2km each.
I'm glad it was helpful. Definitely try that to get a good baseline for yourself and understand your gear...then gather some data on longer runs to see how it scales over longer distance. I'm not running much volume right now as I'm focusing on shorter distances but if I was preparing for a half/full my next step would be to take the VF3 & AF3 out on long runs with pace and look at how the data differes, the assumption is the AF3 would allow me to hold pace longer before mechanics break down and I see a drop in ground contact time and stride length.
You deserve so many more views and subscribers. You are something very special! It appeals to me a great great deal anyway!
Working on that! Thank you for the kind words. 🤙🏻
Excellent for all the shoe and data nerds 😊👏🏾
I'm happy to hear this. Thank you for watching! 🤙🏻
Great content, it was about time to see something like this from you. I didn't expect the VF3 to do so well and Adios 3 so bad but here you go. Also some AF3 results were completely unexpected.
The AF3 is the big shocker here. As i said, I kind of knew the AP3 wasn't fast for me, but the AF3 working as well as it is in very unexpected. Much more to do there...
This is really fascinating stuff! Certainly not an exact science but I think it's a really cool way to aggregate and analyze data to start to come to an educated decision as to which shoe works best for an individual.
I'm in week 5 of a half marathon build and I'm debating between two shoes. I'll either run it in the AP3 or Cielo X1.
I'm going to alternate workouts between each shoe. I'm going to use a track and pace as my controls. I'm looking to see which shoe allows me to run the similar paces at the lowest heart rate, over time. I figured the shoe that does this is probably the one that I'm most efficient in and will allow me to run the fastest over 13.1 miles.
Thanks for the video and inspiration!
Yes, you understood the purpose.
Doing that comparison on a track definitely takes many variables out of the mix, which is good. Once you do that I would find a route on the road and do a similar study there. Road results in those two shoes, especially the Cielo x1 will different on the harder road surface. Tracks can be overly forgiving in super shoes.
Thanks, great content as always.
Thank you for watching! 🤙🏻
Thank you for doing this, really interesting. If I my suggest 2 calculations that would formalize this analysis a bit:
First, Average Speed (in meters/minute)/ Average HR.
Second, Average Speed (in meters/minute)/ Average Power (Watts).
Would be very interesting to see the shoes ranked based on these two metrics for each of the distances.
Interesting. What are you expecting to see from both? Especially Watts, which I don't find that useful in running (as I find it very variable). I come from cycling originally, where watts ins the gold standard metric above all...but in running, not so much.
They will both give you a measure of efficiency: how far you go per heartbeat and the unit of power( watt) respectively, which since it is you doing all the running will give you a measure of which shoe you run in most efficiently.
Say for instance, and I'm using number from memory so they may be a little off, that your marathon pace is 4.40/ km and your heart beat is 186 for two different shoes, but the average power required to achieve it is higher for one than the then, voila, one shoe is more efficient [for you] than the other b/c it requires less effort.
Keep in mind these test runs were short and if one shoe requires more effort [power] than the other, then running an entire marathon at that higher power output is less efficient -- it wastes energy which you could have saved for the finish.
The ratios of these two divisions should be quite interesting and useful.
Got it. Interesting. Makes sense. That’s very much something I would have done on the bike years ago looking at climb data.
I’ve run in all of the same shoes you tested (bought them all too) here.
I’ve commented in other videos that of all of the shoes the alphafly 3 has just felt the best. Honestly I can’t get into the edge Paris, I’ve run some big workouts on the track with them. Did 21 ks this week (20 mins at marathon pace, 20 mins at HMP and 10 mins as hard as I could go after those reps) I just don’t love the edge Paris. Yes it’s fast, but I feel like I’m working so hard in the shoe to make it move the way I want it to. The alphafly 3 for me was my BQ shoe and I just feel like it’s the best shoe on the market for the marathon. like you I was a vaporfly guy since I absolutely hated the Alphafly 2, I’m super happy I gave the 3 a chance.
I’m also super happy I persevered to actually finally get a AF3…. It is the real deal.
Agree on the MSEP…. It’s so much work, attention to mind the stability and recovery is still rough for my legs (though much better than FFTurbo).
In my opinion the control metric to compare shoes is watts. Your watts between shoes were too varied for a fair comparison. I would like to have seen how each shoe performed at the same wattage point-that would be a controlled test. Bicyclist understand this concept very well. Runners should as well
I come from cycling and lived and died by watts. However, I’ve never found them useful in running. I tried to train by them years ago and didn’t find them as valuable as I did in cycling. My watch reports them so I look at them as a secondary metric or curiosity.
Now we just need a repeat of all these tests for a better sample. great video, loved the data.
Community project! It’s not a bad idea…. It could be interesting 🤔
Once again very good video. I really like how you think and your methodology.
It can definitely be tighter but it answers the questions I was trying to get to the bottom of.
Great vidi. Takumi and Vapour doing what is says on the tin, as shorter distance specialists. With Saucony outright as a Marathon distance option. Adios P3 showing its age
Excellent summary.... agree fully.
Interesting review-it's great to see some data, even though the sample size is anecdotal. I’ve conducted similar tests (1K repeats) with more advanced running gait sensors and can share some insights that align with your observations.
Regarding the AF3, you mentioned "floating more," and that's absolutely correct. My tests also show that the bounce (a.k.a. "flight time") is the highest I've seen compared to other racing shoes. This shoe is specifically designed for "stride runners" or "aerial runners" like myself, which naturally results in a slightly lower cadence and a longer stride length. The fact that you like the AF3 likely stems from Nike's efforts to bridge the gap between the VF3 and AF3. Given your short stride and high cadence at speed, you're clearly more of a ground/cadence runner, as you correctly pointed out.
As for the AP3, it’s not performing as well as you expected, and I'm not surprised. I’ve had similar results and ended up selling the shoe because it didn’t feel right for me, unlike your experience. While some pro runners might favor it, my results align with yours-I believe this shoe is overrated, though this is just my personal opinion as a non-pro.
I started that "float" work in the Metaspeed Sky+ last fall, but it never felt right, I did see a few cms in my stride length but the shoe beat my feet up too much to be useable for anything other than drills for me (I documented this on the channel back then). I've been working on it since though with drills, mechanics and strength building. Once I (finally) got into the AF3 I could immediately feel it paying off. Also, oddly, the Takumi Sen 10 is another shoe I really feel it in when I am hitting a clean strike with fast feet... partly due to how direct that shoe is and how light weight it is... it feels great when I get there in them.
All this data was gathered with an Apple Watch Ultra 2, which is what I mainly run with. I do have a Coros Pace 2 and Coros Pod 2 for more data points... I may break that out later this fall on the track to do a smaller followup to this video. We will see.
As for the AP3... I've known it's not fast for me for a while. The main issue is the late stage rocker. I respond to a rocker point much earlier in the midsole. When I tried on the Evo 1s and jogged in place in them...I immediately noticed the difference and understood why. The Boston 12 has a rocker point similar to the Evo 1 and that shoe feels much more natural to me. Nike's rocker points always have worked well for me since the 4%, though the Evo 1 pushed it back even more.
I love the data analysis!
Thank you for watching. 🤙🏻
I love the topic and the approach you took. I would have liked to see a "fitness factor" where you adjusted for your calculated fitness on the day of each workout. I would also caution anyone drawing firm conclusions from N=1 runner and M=1 workout for each shoe.
Ya, done over this longer period of time I agree it was needed. it would have been minor, but needed. Interestingly, during this data set I was going through a transition in my running, lowering weekly volume. So according to Strava my fitness was dropping (which is why I don't pay attention to Strava's metrics) though my own performance was feeling faster and faster...
Great video that then prompted me to go and look at my data. I am comparing a 4x 4k LR workout @ MP in AF3 vs 6,5,4,3,2,1k LR workout @ MP in AP3. WU, Recovery, CD and course are the same in both. Workouts were 1 week apart; the 4x 4km came first. For me, I achieve better pace, longer stride, lower GCT in Adizero Pro 3. And my legs feel much less beat. I guess the ‘for me’ is doing a lot of work in these comparisons -different biomechanics suit different shoe mechanics.
Totally, this stuff is deeply personal and all around your own biomechanics. If I had to guess I'd say you respond better to a later stage rocker in the forefoot and a lower (more stable) plate configuration.
These are always so interesting for me that so many ppl have their range so close together. Like you mentioning 3:30 pace is almost a full out sprint, and a 4:14 is marathon pace. While my 200 rep pace is 2:40-45 pace but my marathon pace is 5:09 ( which should have been about 25-30 sec faster if I hadn't cramped up in the last 6k) but even than my 5k pace and my marathon pace would be 50 seconds apart, and my speed work would be another minute faster than my 5k pace. (And I'm still feeling slow on the faster efforts)
I’m working to getting my 200/400/800 pace more in range. I’ve never trained speed before. Until recently I rarely ran under 4:00/km, even in marathon training.
As I mentioned at the end of these I’ll have a video out in the coming week or two talking about my plans for the fall/winter and what my focus will be. I’m excited!
Thank you for making this video, it's insightful to see your data on all these shoes.
I feel a better way of ranking the shoes for Effort would be a variable that took into account speed and heart rate over just heart rate, as in some case you rated shoes that were slower but had less heart rate over shoes that were faster with slightly highere heart rate, seems like thats just a case of being faster on the reps making the heart rate higher than the shoe making any difference. Also measuring heart rate on 2x200m reps seems so completely meaningless, as the measure will lag behind for most of the rep, which means the heart rate at the start of the rep skews the data massively. I very much doubt running at the same RPE you have a difference of roughly 20bpm over 200m just because of the shoes.
I also have some questions about the accuracy of your data entry and data in general. Some of the stride length and cadence numbers really dont match the speed, an obvious example of this is rep 2 of 200s, 202 cadence with 1.5m stride length is slower than 195 cadence with 1.4m stride length. There's some rounding the the stride lenth but I would've put in the addition significant figure as the difference look way bigger than they probably are when you only use 2 instead of 3 in this case. Another questionable bit of data is the Peg plus stride length of 1.1m in the marathon intervals, with a cadence of 184 basically running the same pace as the alphafly with a 1.4m stride length and 181 cadence, so roughly 25% (?!) longer stride length in once shoe for roughly 2% cadence loss equals less than a 1% difference in speed? These are just the two main examples that stood out to me but there's quite a few others that seem to be borderline on not making sense.
The scoring system is another thing I would change, I would use something that scales based off the performace of the best performing shoe, probably a percentage based system. In cases where many shoes rank closely to each other this means they score nearly even points where as in your system there can be a big gap in points for minimal performance gains.
Fully agree on the HR for 200s, it definitely lags. But was trying to be consistent. It's an ok "ballpark" figure.
I do agree that the proper way to do this would be blending some of the metrics and weighting them. Ultimately that is what would have been done in a proper study. However, for the point I was trying to make in this video, on how people can begin to disect their own metrics, I wanted to pull out pure metrics to do that.
GC/SL was averaged. I run with a Apple Watch Ultra 2 and the Ui for GC/SL in the Workout's app isn't the most precise. I do have a Coros Pace 2 and Coros Pod 2 which would be much more precise. I will likely revisit this type of video in the late fall/winter, on the track and will consider using some other data loging, as well as a stopwatch and track markings.
I think those who are talking about the data "splitting hairs" are missing the point that these small variences over the course of a marathon, 5k, 10k, half etc is the difference between a PR or not. course condition, weather conditions, fuel stations, pacers etc are all variables that the runner can't control but we can control the shoes we lace up on race day, why wouldn't we want that extra edge if we can get it? Is this data going to end up in a peer reviewed study? No, but it is good enough for a runner to make a solid decision on which shoes are race day worthy for them. Great video as always!
Well said. It very much is a personal baseline that is used to validate what you are feeling. From there, in a smaller subset, you can be much more detailed and exacting (if you require that). Using data to validate what you think you feel is very much overlooked.
I just searched for "stride length" in your channel and didn't see a video on that specifically. I'd love to watch a video on why and how you've lengthened your stride. I'm currently at 1.2 during marathon pace efforts with a cadence of 185-190.
I don't have a video on it specifically. I talked about it a bit in my NYC Marathon training series last fall, ep 1 & 2 and then ultimately the problem I ran into with the Asics Metaspeed Sky +... but beyond that I've not done any specific content on it.
Mostly for the past year I've been doing simple things…. High knees walks and strides, A skips, hill running (& trail running for strength), box jumps, focus on fast feet especially with 200s and 400sand general foot strike and mechanics mindfulness (up on my toes).
That has made the difference really.
The "200" reps seems all over the place for distance and all well over 200m. In fact one is almost 300m. So are all these in different locations? Were they in a straight line? If so would wind (or even graident) have effected the times. So bit confused what we are actually comparing here.
The data in general is all over place. I had a quick scan over the stride length and cadence numbers, they just don't match up to the paces. How can one shoe have a higher cadence and stride length yet be slower in pace??
I think because the watch calculates these metrics and they all have significant error bars. The precision of the number obscures the actual blurriness of the measurement.
@@brianreiter5572 I think he might be using a stryd pod or something similar, I think I would discount all the data gathered by it in them categories as it seems all over the place and can’t really draw any meaningful conclusions from it. I feel like he should’ve used a heart rate to pace scaling metric to determine how efficient the shoes are for him. even then some of the heart rate data appears to be junk too, I very much doubt there’s a difference of ~20 heart rate between shoes running 200m reps. I appreciate he’s put a lot of effort into this video but any conclusions drawn from the small amount of data and quality of data is going to mean very little.
This...
Apple Watch Ultra 2. Though I do have Coros Pace 2 and Coros Pod 2 I may use in a followup data set later this fall/winter.
Thanks for posting this data, surprising to see how hard you work in the takumi sen in terms of steps and stride length in order to achieve the speed, which obviously reflects in the heart rate. I really want to see Adidas use a more efficient foam, whether it be a different TPEE or Peba going forward. Lightstrike pro is reliable and trustworthy but it’s really showing its age this year.
LSP Evo foam (whats in the Evo 1, the core in the PX2s and likely in the upcoming AP4) is a PEBA foam...so it's coming. LSP, supercritical TPEE, is very good and very fast. It's more the geometry of the AP3 that doesn't work for me. The TS10 is basically a modern racing flat, so it also make sense it requires more work, less geometry "helping" you.
@@SagasuRunninghow do you find the deviate elite 3’s TPEE compare to LSP?
Puma’s sc TPEE foam isn’t as resilient and is more tuned to durability. LSP is much more performance oriented. Both require similar break-in periods.
Very interessting approach...
So far I used an entire differet approach... I used the same route and compared the shoes by pace/heartrate/distance- pace drop+feeling (for example fit gets lose performence decrease a lot)
Since my approach is not to find the fastes shoe its worked out kinda good so far.
In my approach I try to find the best shoe to get me through the marathon that holds me fast while keeping the heartrate low.
And to my suprise compared to your video the EP4 is one of the worst there, after 15km it drops in all metrics - in my case.
Followed, as in your example with he AP3.
I'm gonna do the same with the Li-Ning Feidian Ultra 4, Asics MSP and VF3 the following two weeks to find my marathon shoe for Berlin.
This data was all gathered on the same route, same warmup. same cool down. Though, I often approach validating and testing shoes they way you do. Interesting about the EP4. Any idea why that is the case for you?
@@SagasuRunning Im guessing it has something to do with the very late rocker combined with the very little bounce (compared to other racing shoes).
The AP3 has in my case the same problem as the Feidian ultra 4, when the fit gets some how lose the Performance drops massively.
I could explain thah with the rods and the huge cutout of the AP3 and the arrowhead shaped carbon plate in the ultra 4.
So far the SC Elite v4 worked the best for me even tho I actually dont really like that shoe, which supports the claims that the SC elite v4 is very suitable for slower runners.
Going to test the VF3 today, I have very high hopes 😊
Yup, you nailed it. The geometry of the AP3 has always been an issue. I don’t do well with late stage rocket. Luckily they resolved this with the Evo 1 and the AP4 looks to have that new geometry as well.
The SCEv4 sounds a lot like the EP4 for me. It works but I don’t always enjoy it.
Good luck with the VF3… prepare for the soft foam… you’ll see what I mean.
@@SagasuRunning I crawl back to you after a 25km run and have to confirm... The VF3 is THE GOAT.
Such an amazing shoe... Pushing you but doesnt force you to push, way more stable than I expected... Not super soft, but far from being firm, maintains a very great cadence. Absolute amazing, maybe there are faster shoes out there, but in terms of versatility this is THE shoe.
The last 5km felt horrible and really taxed my ankles, also walking feels arkward in it, but that is my Fitness Level and not the shoe.
@Chungdol See…. I told you! Stick with it. Since there is so much sink in ZoomX is your ankles and calves are not used to that they need time to adjust to all the micro movements/balance they do.
Also agree on the VF3’s versatility. They may be faster specialist shoes but overall you are never going to be unprepared in a Vaporfly.
Your reviews keep shoe talk interesting.
I appreciate hearing that! Thank you for watching. 🤙🏻
Fun with numbers. I need to spend more time in Strava to analyze my running data.
My advice is to avoid Strava's numbers and go straight into your watches's app and work from there. Strava often tweaks and averages things I have noticed.
@@SagasuRunning Aha. That's good to know. Thanks, Chris.
Great vid, interested in the stride lengthening drills you did to improve, area I’ve been looking to improve.
Simple things…. High knees walks and strides, A skips, hill running (& trail running for strength), box jumps, focus on fast feet especially with 200s and 400sand general foot strike and mechanics mindfulness (up on my toes).
The AF3 (high plate configuration) definitely helps. I’m now curious about the Metaspeed Sky Paris for the same reasons…
@@SagasuRunning excellent thank you for the info
All of the early testing with super shoes found that there’s no running economy difference between a shoe w/ a plate and w/o a plate given the same resilient foam. The reason a plate makes a difference as a stiffening agent is that they are designed for a marathon. The stability issue with soft shoes and lower leg stabilization/hip stabilization won’t be noticed in short reps of 5-10 minutes. This is why the Pegasus plus is performing well in these tests. To be true data, these should all be done closer together, and more than once. So doing it all over again and changing the order of the shoes would be the best. The Vaporfly studies were done in multiples
I remember the 4% studies back in 2018/19. Things have changed. Super foams have improved significantly. Yes, the plate IS a stabilizing agent but in a 2024 super shoe it is doing more than it did back then. The spoon of the plate, especially a Nike fly plate (which is more flexible than others) is adding to the propulsion. The magic of a super shoe is how the plate AND foam work together.
The efficiency gains of a plate are over the longer distance. For 2k of MP I’m not really surprised that the Pegasus Plus holds its own…. Over 42k it would be a different story.
If I was publishing this data in a proper study I would take many of the variable out of the equation and do more rounds in each shoe. That’s not my purpose here. I’m showing how anyone can use their own data to validate what they feel when running…. Or find gaps they didn’t realize. From there you can narrow down the shoes and really look at the data if you wish to fully confirm.
I don't have supershoes to compare as you did, but I think if I ever go, I will go by your data. I am a high cadence ~ 180-185 runner with forefoot-midfoot strike with a mild supination while running,, so You is me, me is you. Haha
Ha...yes, we are one. 🤙🏻
Amazing video! Interesting numbers you have between those videos.
Surprised there is a power reading (I assume you use Stryd?) but no LSS or Impact loading rate. The recent collaboration between Stryd and Luis Orta showed that these two metrics can also add to support the perceived effect from shoes.
I understand these kinds of videos are hard to make, but hopefully there is a follow up with additional metrics.
Apple Watch Ultra 2, which has onboard power, though not as accurate as what Stryd does or how Coros uses their foot pod. I come from cycling, where watts are the gold standard for performance...but in running I've never found them as useful. I tried training by power years ago (when I first came over from cycling) as I understood it well and it removed climate from it all (NYC winters)...but it just never worked for me like cycling. SO I glance at it as a curiosity and secondary metric but don't use it for anything really.
I haven't watched Luis in a while...gotta go check what he is up too. Thanks for the reminder.
summarizing the data By intended use, (i.e. Marathon vs 5K), seems like it would give you a different shoe for each use. The summary you gave would say that if you could only have one super shoe,, for you, it would be the vaporfly, but what would it be if you could have 2, 1 for the Marathon & one for a 5K?
The summary was abrogating across all sets…. Looking for what was best across everything.
That being said. I know I can do marathon pace and distance in a VF3, but the AF3 will be even better for me. For a 5k I would absolutely go for a VF3, even a VF2 if I could get my current size (the one I have is a half size too big). For intervals, 200,400,800…. I’d likely opt for the TS10 for the shorter ones and the VF3 for 800s/half mile.
I’m actually going to talk more about all of this in the next vide on the channel August’s “Shoe Talk” as well as the other shoes I’ll be using this fall winter. Stay tuned.
Would have been interesting to see where the StreakFly would sit in that matrix. Thanks for the detail!
The Peg Plus has replaced the Streakfly for me. However, when we have the Streakfly 2 I'm very curious where that sits for the 5K pace and 200s... I can't wait to test that shoe.
Would be interesting adding some racing flats in there 🤔 my fastest 5k parkrun times are in saucony fastwitch 9, endorphin pro 3, and puma deviate nitro elite 2.
Not much different in times, all around 20:45, sometimes it feels like the high stack gets in the way and I miss the minimal feel of the fastwitch 9’s at short distances.
I’ve just got some Takumi sen 10’s to see how I get on with them.
That is coming....I'll be adding in a few this fall/winter to a revised data set I'll make a video on over the winter. It's going to get so interesting.
I have VF3, AF3, AP3, TS9.
Gotta say, the VF3 might be the 21k and down shoe.
The AP3 I ran my first and only marathon in last fall (2:52xx)
I’ve done 2 runs in the AF3…. Today was a 10k tempo but was too hot to give a fair judgement, I did do some vo2 max reps last week in them and they were really good… will decide over the rest of this block if I use the VF or AF for this fall marathon.
Before I did this the VF3 was my all road racing shoe, but now its likely 21k and under. AF3 is very obviously better for me for the full. Under the 5k though....I'm not sure, Vf3 or TS10... or likely the upcoming Streakfly 2...
This is really helpful and I'm going to look at similar data for myself to see how my own shoes grade out. IIRC though you have a lot of prior experience with the Nike shoes. Do you think perhaps your running has adapted to their properties creating an unintentional bias of some sort? They're obviously awesome shoes, so I don't mean to say anything is wrong here. I'm interested in whether it's a thing one should consider for themselves.
Absolutely! 100% I've adapted to the "Nike feel" which is very specific in my mind (I'll have a video about it in Sept). The more universal issue is I respond to rocker points pushed well behind the ball of the foot. The AP3, MSEP, EP4 all have later rocker points compared to the VF3/AF3... and the TS10/PP don't have much rocker at all so that is a more direct drive anyway.
When I tried on the Adidas Evo 1 in store and jogged in place I immediately found that that shoe would work much better for me as the rocker point is pushed way back compared to the AP3.
What workouts did you do to lengthen your stride? I feel like that's the next thing I need to work on before these super shoes even make any sense for me. My cadence is good, just my stride length is very short.
Simple things…. High knees walks and strides, A skips, hill running (& trail running for strength), box jumps, focus on fast feet especially with 200s and 400s and general foot strike and mechanics mindfulness (up on my toes).
@32:05 That type of video is definitely interesting to me!! 🥇
Ha! I'm glad...thank you for making it all the way to the end. I hope you learned something that you can apply to your own running!
@@SagasuRunning My takeaway from your video is that shoes are not only feeling different, but that there’s actual differences which are somewhat small yet very real. I am inspired to a similar test protocol with the different shoes I have; let’s see how much these differ!
@Anza_34832 very much true. Also shoes are deeply personal. How you will respond will be very different than me.
Interesting, thanks for the video. Where do you think the Vaporfly 2 would fall for the 5k pace category? Thanks.
It’s faster than either for me. In one of the earlier data sets I ran a 8 1k @ 3:50/km and 4 x 200m @ 3:30/km… and it was a second or two faster with similar metrics otherwise. It is the fastest option….
@@SagasuRunning I see, thanks. It's a shame that it's so difficult to buy these days
Worthy of comment that the TS10 looked better for you than the AP3 across all paces. Do you think it is a geometry thing?
Fully a geometry thing, which I’ve known for a while. I don’t respond well to late stage rockers, the AP3’s is very late. The TS10 is a modern race flat, that emphasis clean foot strike much more.
Making a few comments as I watch. So Set 1 - is the notional 2K the exact same circuit from same start and finish as the distances vary especially for the AF3 which is actually 2nd slowest on time but you have ordered it 2nd fastest on GPS pace? Also with the HR how much running did you do before as presume "fresh" on a 2K, HR is going to start low and rise but maybe at MP which just flatten out.
All of these were running the exact same route, flat, along a river…. Nearly same time of day, same weather, same 2.6k WU (to get to the river paths). The didn’t distances is me thinking on starting stopping my watch at the same light pole…. Turns out I’m not.
Workout was always done in the same order as well.
@@SagasuRunning So I am just thinking when you timed yourself at 8:45 in the AF3 but say 8:35 in the Edge Paris it is not necessarily exactly the same "route" ? When I did this sort of thing I tended to focus on the actual times as feel I can time myself more accurately "point to point" than my watch could measure distance. Especially when trying to infer things from rather small relative variances.
you know its fast when it is the metaspeed sky paris
Based on how well the AF3 is working for me AND my issues with the MSEP exploding on me...I'm seriously considering a Sky Paris...
Awesome stuff! It looks like the shape of the plate (thinking about VF3 vs AF3 having relatively similar balance of speed and effort, with different plate; while VF3 and Paris having similar plate but very different outcome in effort) might not make that much difference after all? I'm under the impression AF3 is perceived as a shoe better suited for powerful push-off and VF3 for faster turnover. On a separate note, Deviate Nitro Elite 3 might be an interesting review candidate. I've watched your review on the DN3 and understand puma isn't high on the list for you, but looks like the Elite 3 very similar to VF3 and looks like a good fit for your running style.
Rocker point is the bigger one for me. Both the VF3 and AF3 have their rocker point much further toward the midfoot than the AP3/EP4/MSEP.
@@SagasuRunning Thanks! This solved my long-time AP3 puzzle. It looks like we have very similar cadence and landing pattern from your comments in the other video (you are def faster than me :). I've been using AP3 as my main quality training shoe, enjoyed it a lot, but data always showed they are not faster or more efficient than more rockered shoes. Might have to rethink whether shoes like AP3 and Endorphin Elites are suited for me (AP4 looks like they would have a rocker that's closer to VF3 though).
AP4 has the same rocker as the Evo 1… which is similar to the VF3/AF3… if anything it’s even earlier. The Evo 1 rocker felt very natural to me when I tried them on… I’m really looking forward to the AP4. Can’t wait to test them.
Hello - like this evaluation as data based. Much said about loss of energy return of shoes with increased use. So wondering if here all were similar with low miles when testing? Thanks
This data set doesn't factor that it at all, the shoes range from close to 75miles to under 10miles. However...I have three pairs of VF3s, ranging from 52miles to close to 200miles....I could compare those. I'll add it to the video idea list, no promises and it won't be soon...but maybe someday. Great idea. 🤙🏻
Would be interesting to see the data of this workout performed after a 10k in a control shoe - I have a feeling the adios would do better
I'm likely going to revisit this later in the fall/winter, on the track, with a small subset of shoes in much more controlled conditions. But yes,I agree.
Also not sure if I missed anything, but if we look at the 2k comparison between Paris and AF3, the cadence is ~3% difference (181 vs 187) but stride length difference is much larger (~9% even if we take the smallest gap from rounding, i.e. 1.24m vs 1.35m). Given the pace between the two are virtually identical, would have expected the SPM and stride length gap to be similar in scale in different directions? Not to nitpick anything but just wondering if the tracker apps or something else could be making smoothening adjustments on some of these data that could impact the results
All data was gathered with a Apple Watch Ultra 2 and puleld from the Workouts app. GC & SL are averaged per rep as there is not precise metric given...so yes, it will be a bit inconsistent here and there. I did take the lowest and highest reading I saw in each rep in averaged from there. So yes, it's not a precise as a specific metric the software is reporting like SPM.
Hmm, I have a feeling the measurement error is comparatively large, enough to affect the data and some of the conclusions drawn from it. Furthermore, as you mentioned, depending on the order in which you tried each shoe, you may have gotten a fitness boost for a particular set. Still, at the end of the day, perhaps it's not surprising that the VF3 is the shoe that is best tailored to your biomechanics and (after a surprisingly good, albeit short, tempo run on the good ol' Peg 39 yesterday) I'm almost tempted to buy a pair and see how they compare to the AP3. I'll try to hold off until the Adios 9.😅
Ya, as I said, this is data set is not totally "lab certified" and there are plenty of variables in play. But it's enough to validate some impressions in order to focus in on more specifics.
Interesting, the VF3, AF3, PP workouts were all done back-to-back-to-back over 10 days (with rest). The MSEP came next. the AP3 is the oldest and the EP4 was the last one I did.
The VF3 and AP3 are very different shoes. I do love the AP3, but I've known for a long time the VF is just faster for me. Alot of that is familiarity and my own biomechanics adapting to the VF feel, which is still unique in super shoes.
@@SagasuRunning I agree, I suspect familiarity plays a no small part in the biomechanics equation, to the point I have to wonder about the efficiency impact when using different types of shoes (perhaps minor if you run enough to adapt to each?). This plays into a comment I think you made about keeping a rotation within brand; if each brand considers their line of products holistically, it simply makes sense to use a same-brand rotation (tempted as I am to mix and match😄). On that note, which brands would you say excel at this, and which have room for improvement?
@OdanUrr87 Adidas’s entire Adizero range works together nearly perfectly from trainer to racer. It’s by far the best one. Nike has the Pegasus/VF combo, which is nearly perfect…. V17 to AF3 is close, the PP fills in the gap well.
Honestly, Saucony also does well with the T22/ES4/EP4.
Other brands are much more hit or miss.
The 1K reps all seem to have the same recorded distance so seems a more obvious ordering by time or pace could be done. Re the HRs (especially Peg Plus) was that day you ran in the Peg Plus somehow "harder"? It can be difficult to be sure on HR as some days I find it is low, normal or high for reasons other than just the pure effort. Cadence looks a lot more even in the 1K. I find the AF is the only shoe where typically I find mine is a bit lower (and hence longer stride) but it always feels like it wants you to "bound" along. I also am not sure how long I can actually sustain that. "Striding out" I always found easier when fresher. Bit surprised the Peg Plus has the lowest cadence but it is quite close and hard to be sure of any real differences in these 1K ones.
The 1k reps are done on a segment I have measured many times (also it's known Strava segment in the area). Thus the start/stop are marked on the path (multiple run and triathlon clubs use the section of bike path I did these on for training)...1k seems to be their shortest mark though. I'm eventually going to go mark a proper 200/400/800 and 1/2mile...
The AF3 defintley prefers me to "bound" along...
The PP cadence isn't a huge surprise to me, that shoe wants me to run flatter to load the foam more. I've noticed my cadence is always lower in that shoe.
i try AF3 again, do “run” again in store today and i have feeling that i can’t find transitions between super soft heel and air pods..by the way, good video, keep going
The heel sink and then secondary sink into the mid/forefoot is very ZoomX and a huge reason many don't like the way they feel. As a forefoot striker I never notice it, but when i do heel strike it's very apparent.
But compared to the previous AFs the continuous midsole makes a massive difference in how the shoe transitions. It's much much better. I'm overall really impressed with the redesign of the shoe.
As always, thank you for the support.
🙂
third test. If you want to run 2 x 200, how is it possible to do an interval of (sometimes) 280 meters?
Because I'm running on bike/running paths along a river and don't always start/stop my watch at the right line, sometimes due to people traffic. None of these were done on a track.
(typo in the title) thanks for the videos
Thanks for that. Got it. 🤙🏻
interesting comparison, I just wonder what comparing BPM at the same speeds on a treadmill would look like, this would take the speed variances out of the equation. I personally have to admit I go mostly by feel, but I am also much less tolerant and have to exclude a lot of shoes because of incompatibility with my biomechanics and anatomy. I would currently go VF3 as well, but as you say everyone needs to find whatvworks for them, and the Adios Pro 3 obviosly still works for a lot of elites that can do lab tests to choose their shoes.
A 'mill would make it much more exacting. Back in my cycling days I did nearly all of my important training indoors on rollers or trainer as I could control all the variables. However, I don't have access to a 'mill here in Taipei, refuse to do gyms and would rather deal with variables outside as a runner.
Also...I'm not saying the AP3 is bad....it's a great shoe. What doesn't work for me in it is the geometry. Late stage rockers just don't gel with my run form. The Evo 1 fixed this and the AP4 has that geometry...so I expect it to be far faster for me. I can't wait to test them....
I ran a 10km with the Sky Paris but felt stiff so you think Edge is much softer?
The Edge Paris is stiffer due to the plate configuration. FFTurbo+ is a firmer foam, even after it breaks in and Asics uses VERY stiff carbon fiber plates in their shoes.
can you add one more for percdntage performance from base
I'm not sure I understand the metric?
I've never tried the Vaporfly 3 but if I like the Edge Paris is there any benefit to trying it or should I just stick to the Edge?
Depends on if you are looking for a new shoe. The VF3 is much softer and a very different type of ride than the MSEP, even though they are both low plate configuration shoes. Personally, I do think everyone should experience a Vaporfly. It is the OG super shoe and still has a unique ride in the space. Though it can be hit or miss and is a big investment for a shoe one might not like, especially if they started with the current style of super shoes from another brand.
The MSEP is fantastic though, so you aren't lacking anything with it.
Wouldn't doing these tests on a treadmill be much more consistent? Great video though!
Very much so... but I don't have access to a mill...not do I think I could stand running on one (unless it was middle of winter and freezing cold outside). Even a track would be better for this. There are plenty of ways to make this more controlled, but all workouts were done on the exact same section of road (straight bike path on a river) with the same warmup and cooldown. It's enough to get some basic understandings, validate some feelings and impressions and work from there (which I'll be talking about in my next video).
I think you are exaggerating the role of carbon fiber plate when you are comparing PegPlus and Vaporfly. The ZoomX on PegPlus is a bit different than Vaporfly/Alphafly and the amount of foam in Vaporfly is significantly more than PegPlus. The weight of the PegPlus is also a drag. In other words, I am saying there are more factors beyond CF plate in contribution to the performance. Anyway, it does look like Vaporfly 3 is the shoe for you overall if you were to have just one shoe but you might benefit more from say AF3 on a marathon since you already have it.
Question: What did you (exercises, types of run, stretching, etc.) did to increase your stride length? I can benefit from your experience if you share. We are in the same age group and I am a very high cadence runner (200+ on VO2max and sprint type intervals) but I think if I can cut back on cadence and increase my stride length a bit, I would be more efficient.
I agree that it's more than the CF plate...but that is the thing people ALWAYS talk about and the general thinking out there in running presently is that you NEED a plate (super trainer, super shoe) to go fast and the new class of performance trainers coming out (of which the Peg Plus is one of the first) is somehow not capable. I've been finding the peg Plus beyond capable as a training option when I don't want a plate, which is often.
Honestly...alot of it was form work. Making sure my foot strike, forefoot, is clean and powerful AND letting my body "float" before I try to drive my next leg forward. Alot of repeating "strike and float" over and over doing strides and faster running. Spending time in the Metaspeed Sky+ last year also helped.
Drills have been A-skips and plyo (box jumps up, box jumps down with a explosive hop using your feet and legs only - no knees hips - as soon as you contact the ground). Fast feet/high knees hill running as well, knee drive work as well.
@@SagasuRunning Thanks for the info on lengthening the stride. I should get some regularity on doing drills before workouts.
@TolunayOrkun It definitely makes a difference. Muscle activation and things you become aware of on the run. Highly recommend.
I think that if you were going to use a non-plated control shoe, the Superblast would have been a better pick.
Why? It's not meant for speed, beyond marathon pace and isn't a pure super foam.
If you could only have 1 race shoe for HMs and FMs, for a not so fast runner (5-7 min/km), the adidas AP3 or saucony EP4?
EP4...
You should really try the puma fast r nitro elite 2 im really interested in it and think the geometry and foam beats the vf3 and ts9
I refuse to run in a shoe with a decoupled heel and forefoot. That’s a hard pass for me…. Won’t even consider it.
Interesting stuff. I have the feeling that the AP3 resists acceleration period and is hard to take above threshold pace but feels good for cruising. That’s a lot of feels. But empirically any super shoe I have tried is lower HR to maintain a pace than a daily trainer.
I wonder how much correlation there is to simple mass/inertia effects especially at faster paces. For example at the risk of oversimplifying, the TS10 is essentially the same materials as a AP3 but it performs a lot better at faster paces for you.
Lighter is better. Foam, plate, and geometry have to be very good to overcome any weight penalty.
Data is great but also it is important to keep in mind the error in the measurement - something pounded into me in college.
I’m dubious about the ground contact time measurement as a direct performance score. The really meaningful measurement you are trying to get is effort to maintain a pace or pace efficiency or velocity at VO2 consumed. HR is a slightly indirect measurement of aerobic effort. Pace is pace. Does GCT really mean mechanical efficiency especially within the precision and accuracy of a watch to capture? I just looked this up for Garmin and they state a mean absolute error of 5.4%. That suggests the range of your measurement distribution is within the error of the measurement.
I also am curious how you are capturing HRM. If it is an ECG HRM then I’m sure it’s very accurate but if it is captured from the wrist then its error range is going to be large.
How did you capture pace? Capturing average pace over 1km or 2km with a multi band GPS is probably pretty good for this propose but for 200m it’s definitely not good enough but lines on a track would be.
The absolute number of test runs is pretty low and the differences are also not really big so any measurement error could have large effects.
Everything is also driven by RPE which is kind of a squishy concept that is hidden by the precision of the measurements that you end up capturing.
I’m not actually suggesting that your results are wrong at all. I think that you are validating that the AF3 is very good for you across a range of paces and the AP3 is decent for marathon and bad for 200s. But I am saying that interpreting the huge amount of data we capture so easily is hard.
I've known the AP3 isn't "fast" for me for a while. The issue is the late stage rocker on it. I respond much better to early rockers. When I tried the Evo 1 on and jogged in the store (which shocked the Addias rep as they were demo pairs) I immediately noticed the rocker point much further back on the shoe and it felt more natural (like a VF). The AP3 is great for locking into a pace and looping along in it, for me at least, but accelerations take alot of effort.
I captured all of this data with an Apple Watch Ultra 2... so dual-frequency GPS (which I find very accurate for my needs). HR data is from the wrist sensor, which i know is not as accurate as a chest strap...but after YEARS of chest straps in my cycling days and then moving to wrist top as a running I find wrist top sensors that Apple and Coros use (not Garmin) are "close enough" for me. I know they aren't dead accurate, but I have worn the Apple Watch since it's day of release (and other Apple watches and Coros prior) and trust the consistent of the readings for the data.
For anything under 1k I definitely thing a GPS is not ideal. Track mode on the Apple Watch Ultra 2 and Coros Pace 2 is decent...but not as good as an old stopwatch to the lines on the track.
I'm going to come back to this later this fall/winter with a smaller subset of shoes on the track and will likely use a mix of GPS data and stopwatch data...as that will be a 200/400/800 data set.
Given this I think there is not a large enough data set to overcome the measurement error of oHR, pace/distance, and GCT. I am pretty sure the measurements all fall within the error bars of the tools doing the measurement which means and you don’t have a large enough sample to make robust conclusions from the data.
The average error for each measurement is probably +-5% or more which is more than the difference between the measurements. The watch is averaging a lot of individual measurements and in the case of an Apple Watch applying an ML smoothing model.
The most accurate measurement is probably the average HR and pace over the 2km distance because there are the most data points being average. In principle the measurements should tend to cluster around the correct values so when you average a lot of measurements you are converging on a good number. If you divide the distance covered on each “2km” test (averaging a 500 or so GPS points 1s apart) by average HR (averaging 500 or so oHR measurements) you can calculate a distance covered for effort value. At 2 significant digits all of the test runs come out to 0.013 km/heart beat/min - which is to say they were all the same effort. (Coincidentally that kind of validates the RPE concept.)
To put it another way these are all really good, even the “control” shoe, and the differences are hard to tease out without lab equipment. Even the control shoe is too good. The conclusions are hard to separate from confirmation bias especially since the testing was driven by RPE. I think they are valid opinions but I also think the degree to which the conclusion is strongly supported by the data is overstated.
You would need multiple tests at longer sessions simulating the actual target race to develop a data set that is rigorous enough to convincingly tease these shoes apart in a way that is improbably impractical to actually go out and do.
@@brianreiter5572 If I was going for lab quality data and a double blind test for publishing I would worry about all of this.
To gather a data set to validate what I feel, even if not 100% accurate, but consistent with the same watch/course/weather/time of data and overall effort...this works for me and I would encourage anyone else to look at their data like this. At least to inform themselves of what they feel in order to narrow down on a subset of shoes and really focus on them...longer runs, more varied data, etc.
@@SagasuRunning 🤣
RPE, "comfort filter", and pace for effort can get you very far in evaluating products that work better or worse for you, absolutely.
I have a personal have a bugbear about watch metrics, which is not about you.
I'll offer a couple of examples here which are my own grievances and specifically with Garmin because that's the devil I know best. The basics have more noise in the measurement than the watches absolutely misrepresent with false precision.
GPS points are each only accurate to about 2m in good conditions but the distribution averages out around the correct point and as a result Garmin's current pace field is very poor but 1km pace is quite good.
The Garmin VO2Max estimate is super-terrible. Garmin published their validation and claims +- 9% accuracy under ideal conditions with correct weight and good GPS and pavement when used with an ECG HRM. That means if you get 50 it is between 55 and 45 probably but in reality it is much worse than that in many comparisons for lab tests. There is an entire suite of Garmin physiology metrics that are derivative of that number but including the 10ths of a point which are in the FIT file but not displayed. That's basically all nonsense and literally generating recommendations on noise. When I realized this is just a silly number generator, I found the way to turn off the display of these metrics because they are objectively unhelpful.
Anyway, I did enjoy the video and your effort to take an empirical approach. It's all good and I know it is a ton of work.
You know, it’s interesting that you mention Garmin. I left the Garmin ecosystem years ago for many of the same issues you highlighted. it started with cycling head units (moved the Wahoo) and when I started running more seriously I noticed a lot of the data inconsistencies from Garmin, port GPS mapping and all their “training analytics” always seemed wonky compared to what I knew about myself from years of cycling.
The whole Garmin Connect platform going down for 90+ days back in 2020 was the last straw (I mean who has their production environment connected directly to anything that would allow a “hack” to hit it from the support team, 🤔).
So I left Garmin and never looked back and have never had issues with watch data from Coros or Apple. I do look at some of the blended and averaged metrics with a grain of salt…. But overall I trust what it’s reporting to me. At least inasmuch as it is consistent day-to-day…. So even if it’s off it’s off consistently and I’m working from the same point.
I do not get the first test. you say you run 2k's at 4,14. but you ran the 2ks in different times? pls help :D
I ran by RPE. So the goal was to run said pace, but running at the RPE turned out to be faster. Fitness was better than I expected, at least for a shorter workout like this.
For science!!
… and finding what is fastest!
where is puma deviate nitro elite 3 that shoe is beast
I won't be covering that shoe...it's not for me.
Your 200m time seems very slow compared to 5K and MP, not sure they are accurate based on that and therefore wouldn't this throw off the intended prediction of the reps. Also I'm always surprised when people think marathon shoes would be slow at any pace when pros are at 4.40 per mile
I use the VDOT Calculator to determine paces when training. 44secs in a "interval" pace for a 200 per my MP and 5K pace. So yes, slow...but in the general range.
200/400/800s are something I'm going to be spending much more time with this fall/winter, mostly on the track. There I'm looking more for 37-41secs on the 200, but that is in spikes.
I'm not surprised that a shoe like the VF3, designed as an all-rounder, covers this work...but I am for the bigger and bulkier shoes optimized around the Marathon.
Can somebody please explain to me how you can have a higher spm and a longer average stride but run slower?
I think it’s safe to assume that the measuring equipment doesn’t have the precision required for this testing.
Stride length was an average. It is imprecise but gives a general ballpark. There was a range of SL across the entire rep. I took the shortest value and longest value I saw in the rep and averaged it.
I will go back in a few mo the with a foot pod that will be more accurate.
@@SagasuRunningcould you also include more info about when each shoe was tested because like you said your fitness improved over the 4 months.
Also how many miles are on each pair of shoes. I’m sure you know some shoes are better right out of the box vs some need a little break in time to be their best.
I’ll find a way to include that in my next data video later this fall/winter.
Absolutely sad to see Adidas AP3 is at the bottom among your six chosen super shoes 😞
I still love it...but it's not a racer for me. I'm so curious how the AP4 will suit me. There should be some changes to that one that I will respond too well....
@@SagasuRunning Do you have any idea when the AP4 will be launched?
@@SagasuRunning On the Ascis Metaspeed Edge Paris, I am very tempted to try it as the racing shoes for my Taipei Half. What do you think? So far, I've only 1 Ascis shoes, i.e. Superblasts.
I expect some news in the coming weeks and possibly a very limited release at Berlin (like they did for the Evo 1 last year) but I’m consistently hearing Jan/Feb 2025 now from all sources.
Coming from the AP3 you will notice ALOT of instability… and the Taipei course has some iffy pavement (especially the bus lanes on Renai Road)… but form a foam standpoint you will find FFTurbo+ very familiar coming from LSP. So you have a pair do MSEP yet?
AP3 only 5 points ahead of peg plus, I wonder where the Boston gonna land, hahah😂
The B12 is great...but it is well out of it's league in this data set....it would be quite far away.
But you used yourself as the only tester, not sure if that would give reliable results
It gives reliable results for me, which was the point I was highlighting in this video. How to use your own data to look at what works best for you as a running. I explained my data, how I structure it and analyze it to help other do the same for themselves.
@@SagasuRunning fair enough
👍
🤙🏻
How on earth you guys so fast? :)
It's all relative... keep at it 🤙🏻
"This video is going to be mostly me voicing over spreadsheet data. So, if that's not your thing, I totally understand."
Clearly, you don't understand the audience you've cultivated... 😅
ALRIGHT, LET'S GO!!
Hahaha...I think I'm beginning too...though was trying to warn those that are new here. 🤣
Surprising that you are in Taiwan , I wan to know how Alphafly2 work with these data tks for sharing
Yup, I've been based here for nearly 4 years... I never plan to put a AF2 on my foot again, that shoe just makes me think of plantar fasciitis and I never want to go back there.
@@SagasuRunning cheers