Pitching Nerd Alert on a Shoe- and Foot-Related Podcast

On Slack, I pitched James and colleagues on doing a Shoecast. I don’t know how long the Slack post will persist, so I write here for posterity. I would like a deep dive on feet and how they fit in shoes. This is inspired because I have some difficulty finding a perfect fit. I recall a 2020 or 2021 podcast that Velonews did with the guys from Giro, and Giro said they aimed to design their shoe lasts to fit as many people as possible, and to minimize movement inside the shoe. OK, fine, but aren’t Giros shaped relatively narrow? What is a last, anyway? Here are some questions I thought about.

  1. First of all, how do feet even vary in shape? Everyone knows length. I assume that many of us probably more or less fit standard width shoes. If our feet are wider than average, we’d probably figure it out from foot pain. Alternatively, if you got a measurement taken on a Brannock device and the operator was competent, you’d probably get a sense of how wide your feet are. Are there more shape parameters apart from length and width?

For example, I have heard some people say they have high insteps. Is that a thing, i.e. feet vary in height along the Z-axis as well as X and Y? Do heels vary in size? When you look down at your feet, you see that your toes form a curve - does the amount of curvature vary between people? Again, are there other parameters?

  1. Do different genders, racial/ethnic groups, or other groups vary in average parameters? For example, many manufacturers offer women’s shoes that are narrower than a given unisex size. However, Specialized’s anthropometric review argued (based on a Retul dataset) that women’s feet are about as wide as men’s on average, as shown below. However, do note that they are measuring circumference around the ball of the foot; it’s possible this measurement might vary depending on foot volume, although it may not (high volume feet might be bigger further up, more in the instep)

(Title: When to Share Product Platforms: An Anthropometric Review, referenced in this CT article by Sarah Lukas, most recently downloaded here)

As a less well-known example, Bont offer an Asian fit shoe, which they characterize as (relative to a standard fit for a given size) wider in the forefoot, narrower in the heel, and “flatter instep/arch”, which I think means lower volume.

  1. If I’m designing a shoe, where do I get lasts? Are there a few industry standard sets? What data are they based on and how current is that data? How do you select the number of steps between sizes?

  2. Some companies have reputations for relatively narrow shoes (e.g. Sidi, some (?) Giro models), and some for relatively wide (e.g. Bont, Lake). How do these differences originate?

  3. Are last shapes refined as companies go along? Shimano seems to allude to doing this when they released the 902 model of the S-Phyre shoe, and James reported a perceptible difference in fit.

  4. How can consumers reliably select a shoe? We are often advised to go to a store and try on a large number of shoes, preferably with your preferred insole brand. Are there particular things to look for? E.g. tightness around the forefoot, slop in the heel area? Most of us learned in street wear to make sure you have about a thumb’s width of excess length in front of your longest toe. Is that an accurate heuristic for cycling shoes? If you have slightly wide feet, you may have learned to size up, and in fact Shimano’s guidelines specifically recommend doing this. Is that a good heuristic? I worry that the fit elsewhere on the foot can become sloppy. Is this a good heuristic? Last, you often develop problems with a shoe fit later on, maybe after a long ride, maybe after some time. How reliably can consumers tell in the store if a shoe will work?

  5. Somewhat related to #6. Do Brannock devices give good information to begin with? Do their measurements vary significantly if different people conduct a measurement (if you’ve ever heard the term inter-rater reliability, this is what I’m talking about)? Most Brannock devices I’ve seen output a nominal size (e.g. Euro size 40), whereas if you check, different manufacturers’ claimed last lengths may vary for a given nominal size, e.g. Fizik size 40 is listed as 257mm, Shimano is 252mm from my memory, I think Giro claimed 255mm but I can’t be 100% sure and they definitely round to the nearest 5mm which I look down upon. Anyway, are there issues like this in finding your shoe size from a Brannock device?

  6. But related to #7: is there a standard way to measure shoe length? Most manufacturers list their last lengths, even if you have to dig (e.g. Shimano) or they’re rounded to the nearest 5mm (e.g. Giro, although they seem to have taken the length charts off their website) or they’re shown on a ruler instead of given in mm (e.g. Shimano, again, although I think the shoe boxes list the actual length). I know that Lake specifically says mark on paper the heel and the furthest toe, then measure just the Y dimension like you’d measure bike reach. However, at least a couple manufacturers I’ve seen (e.g. 45 North and Quoc Pham) have you measure the length from heel to big toe, as if you were measuring the actual TT length on a compact frame.

  7. By now, that may be a full podcast episode. We still haven’t covered arch support. It seems that most shoe companies put a crummy stock insole that’s usually for a low arch in their shoes. There may be an unstated expectation that you replace it, but do most people? How can you select a good insole?

Insoles seem to vary somewhat in their actual height. For example, Steve Hogg contended in 2011 that Specialized’s insoles were relatively low, whereas Icebug Slim and Sofsole had considerably taller arches than any Specialized.

Moreover, some insoles offer metatarsal support (e.g. Specialized, Icebug). Some don’t (e.g. Bontrager). How beneficial is metatarsal support? Does it vary by rider? For example, some people I’ve talked to on forums don’t like Icebugs with their fairly prominent metatarsal bump, even though it doesn’t really stick in my foot when I am riding. Are there other differences between insoles apart from arch height and metatarsal support?

  1. How much difference does the construction of the uppers make in shoe fit? For example, I believe that I would not normally fit Giro’s lace shoe lasts, which I have heard are narrow (but I don’t know that the E70 uses the same last as the rest of the Empire line). Based on James’ review of the E70 knit and laced shoe, however, I decided to try it, and it was pretty comfortable at first, especially after I skipped a lace hole to relieve pressure. James did say that the knit uppers seemed to have more conformability, which could enhance comfort.

For the record, after 2 years on my E70s, I believe that they are slightly on the narrow side compared to what I would want if I could have a custom shoe made entirely for me. That is, they are within my fit envelope but on the narrow side. Moreover, when I tighten the laces enough to stop all foot movement during very hard pedaling, they’ll hurt my feet, especially along the outsides. Right now, even during fast group rides, I feel like my feet don’t move around so much that I want a different size or a different shoe entirely. If I were still road racing, I might feel differently.


Yeah but did Hogg mention anything about the polarity of the insoles? Dude is a very sharp, interesting quack.

But for real, I would definitely listen to shoecast. (Podiacast?)

1 Like

I’d be very interested in a pod/ped-cast on this too. I have a wide forefoot and often struggle to find cycling shoes that fit (Lake wide fit are the best I’ve found).

One thing that frustrates me is when shoe companies equate a higher performance fit as being narrower, with only casual/entry level shoes being wider. I understand getting a good lock-down is important, but the implication that you must have narrower feet to use a stiffer/lighter/higher efficiency shoe is ludicrous.

Love this idea. Good off-season fodder for us to dig into.

1 Like

You know the funny thing? Examine Lake’s very detailed sizing charts. They are one of the companies that do this. If I measured my feet correctly, their race last shoes are too narrow for me. In their case, that’s not a deal-breaker if I wanted Lakes, as there are many good shoes in their second-tier last that fit well. (That said, if I absolutely wanted 4-hole Speedplay shoes, those are only offered in their narrower race last.)

Hogg is unconventional. I would hesitate to call him a quack, but yeah, I am under the impression that his championed midfoot position doesn’t work for everyone … that said, whatever your judgment, I’m focusing on his assertion that the arch heights differ among insoles. That is, a low arch Specialized may not equal a low arch Icebug.

Alpine ski boots have paved a path. Most brands offer the same level of performance ( for ski boots its stiffness) in 2 or 3 different volumes. Volume is usually a combination of width and instep height. Feet aren’t 2 dimensional but 3. Then we can talk about footbeds :wink:

1 Like

All of this would be much easier if shoe companies would adopt ISO - mondopoint sizing. I assume most of them don’t want the complication of admitting that one width doesn’t fit all.
I use a Brannock Mondo tool which gives length in cm, but unfortunately still width in letters (so I pull out the calipers to get mm).
This would be much easier if companies would use mondo sizing, but my efforts to get that information from brands have been futile. So only Lake and Bont are any use.
It’s very useful to be able to explain to a fit client using the Lake sizing chart why they are getting numb feet with the standard narrow lasts from other brands.

Using the Ansur 88 dataset (nearly 4000 individuals) gives averages of (length/width/circ):
M 270/101/249
F 244/90/223

If we take the average female ratios and apply to the male average length we’d get
So there isn’t a great of difference on average between boys and girls feetsies.

Averages aren’t much help to an individual, in fact they’re hindrances as brands use them as shortcuts. So we’ll look at the variation.

Taking the standard deviations of the circumference and width ratios gives 1sd of ~10.5 and ~4.5 respectively (there are slight differences in sexes but those aren’t big in this context).

Refreshing your memory of high school stats - this means that 68% of the population fall within the 4.5mm band either side of the average width. For our average 270mm foot length male this means basically 96-106mm foot width.

Compare that to the Lake sizing chart and you see that the 43.5 size in standard width is 96mm, so you’re relying on material stretch to make it snug for the average or less width foot (which is fine). But you’re looking at the Wide or more for maybe 40-45% of the population if shoes are sized correctly.

Needing a high volume shoe could be a result of a large circumference or a high instep.

For selecting insoles there are two factors - comfort and alignment. Comfort measurement based on pressure sensors or rider feedback. Alignment requires measuring knee track and hip motion while pedalling. I’ll stick an insole in and check whether it improves knee alignment.


As someone who has had a terrible time lately finding an appropriate pair of shoes, I’d be interested in this.

After our new puppy savaged my Shimanos, which were always a bit tight but ok up to about 3.5 hours, I thought I’d get a decent replacement pair.

I decided to go for Lakes, on the recommendation of many, and religiously took measurements: I have a length of 268mm and a width of 99 (interestingly almost bang on average). So after carefully correlating this with Lake’s chart, I ordered (so I thought) the right pair… and they were both too short and too narrow.

I then went up half a size AND to the wide fit: STILL a bit short and narrow.

Now either the sizing chart is essentially a work of fiction, or there’s something else going on here. I know I have a high arch/instep, but now looking at ordering my 3rd pair in a fortnight, I’m pretty baffled, tbh…

Using the Ansur 88 dataset (nearly 4000 individuals) gives averages of (length/width/circ):
M 270/101/249
F 244/90/223

If we take the average female ratios and apply to the male average length we’d get
So there isn’t a great of difference on average between boys and girls feetsies.

It may be better to plot the data. Thanks for mentioning the ANSUR 88 dataset; I wasn’t aware that it was publicly available. In the graph below, the X axis is foot length. The Y axis is foot breadth. There are some brief descriptions on the ANSUR site at Penn State University if you click the Visual Guide button. The red dots are males, the greenish dots are females, and each dot represents one person in the dataset. There’s a region where men and women overlap, probably from foot length (as measured in ANSUR) 245mm to 265mm or thereabouts. That’s the small side of the men’s range and the high side of the women’s.


Anyway, for reference, I believe that Euro size 40 is around a foot length of 250-255mm. I don’t think sizes under 40 are commonly stocked. Specialized seems to offer its shoe lineup as unisex, and their site mentions sizes down to EU 36. I am pretty sure Shimano’s shoes are offered in men’s and women’s sizing; on Competitive Cyclist, men’s S-Phyres are available down to 40 and women’s down to 36.

Basically, at any given length, the vertical height of the scatterplot represents the potential variance in foot breadth. You will get similar results if you display foot circumference.

Now, if you fit a simple linear model where breadth is the dependent variable and length and female gender are the independent variables, it tells you that the average breadth for women is about 5mm less than men at the same foot length. But again, that’s the average, and there are people above and below the average. So, for me, this would indicate that I could offer a wide range of unisex sizes and multiple widths at each size.

And also, this appears to contradict what Specialized found in its own dataset from Retul. I think that Specialized may have said that they suspect measurement error in the ANSUR data. It wasn’t discussed in the CT article I linked, and it might have been in a different CT article or on another site entirely. I recall that Specialized argued that its own Retul used pretty precise measurement tools, whereas they thought the ANSUR data were less precisely measured.

My position is even if the the ANSUR data were correctly measured and the average woman has narrower feet than the average man, controlling for foot length, you may still be better able to serve the needs of the overall population by a unisex lineup going down to size 35 or so and up to whatever it goes up to, with multiple widths. Now, the thing is, it appears that “width” may not be interpreted by most players as the width of the carbon outsole; they may just offer higher or lower volume uppers. I didn’t discuss in my initial post, but the issue was discussed in a 2018 BikeRadar article. The author basically claimed that only Lake and Bont actually offer wider outsoles; everyone else just offers higher volume uppers. Giro appears to state this upfront (it designates some shoes as HV), and if you dig through Shimano’s literature they also show that their ‘wide’ shoes really are more like higher volume.

I’m also highly interested in this. Specifically, in foot volume. I’ve dealt with movement in the shoe + arch pain for many years. Gone through several pairs of shoes. My hypothesis is that I have VERY low volume feet. The only shoe that ever truly made my forefoot feel secure was the Ares I tried on (and damn near bought). The rep I was talking to has the same issue as I do and said he’d be on the ares if he could afford it. Apparently Sam Bennett had the same issue, hence the Ares.

And here’s where I get kooky. I think arch support is fundamentally flawed. People get arch pain and think they should fill in the arch. But that defeats the purpose of an arch. It’s not supposed to bear load in the middle - only the ends. I tried arch support and it felt like I was foam rolling my arches. My hypothesis is that what really causes arch pain is arch collapse, which only happens when your forefoot is allowed to slide in the shoe. If you could pin the forefoot down hard enough, not only would you eliminate vertical slop, you’d also eliminate fore-aft sliding, which would prevent arch collapse.

I would be overjoyed if CT talked to experts and covered this crazy hypothesis of mine in a podcast.


Seems made for Nick Squillari :slight_smile:

1 Like

Another big issue I found is toe box shape. Even if your shoes are wide enough at the metatarsals, often immediately forward of this, the shape of the toe box tapers aggressively and crushes your toes together or curls them under.

Spent a lot of time with wedges and whatnot before realising that if your foot is squished and toes aren’t grounded properly, you have no idea whats really flat/level/neutral for making adjustments, on top of the discomfort. Bonts are better at this but I had issues with them elsewhere.

In the end I went off the deep end and make my own custom fit shoes from scratch :upside_down_face:

1 Like

Cycling in Alignment with Colby Pearce did a few episodes like this:
Ep. 8 Shoes

Ep. 11 Orthotics