The recent explosion of interest in minimalist running shoes has brought along with it an interest in specific shoe design features that probably rarely crossed the mind of most runners just a year ago (before the publication of Born to Run by Christopher McDougall got people thinking about these kind of things). One of these is the concept of heel-to-toe drop, sometimes also referred to as heel-toe offset, heel-toe differential, or heel-toe lift (you sometimes also see “forefoot” substituted for “toe”). So what exactly is heel-toe drop, and why are people interested in it as a shoe design element?
Heel-Toe drop/offset/differential as defined by Brooks Running is “the difference between (midsole + outsole) heel height and (midsole + outsole) forefoot height” (see picture above from New Balance if you’re not clear what the midsole and outsole are). Thus, a drop of zero would mean that when seated in the shoe, the heel and ball of the forefoot would be at exactly the same height off of the ground. A drop of 12mm would mean that the heel sits 12mm higher off the ground than the forefoot. The importance of the HT drop value is that it’s thought that the lower it is, the easier it will be to land on your midfoot or forefoot while running. I’m not sure if there have been published studies confirming this, but my personal experience running in shoes of varying HT drop values, as well as a few of my informal laboratory attempts to correlate heel height in shoes with footstrike, seem to suggest that this relationship is likely real. You can check out these posts for more:
Vibram Fivefingers and Barefoot Running: Does Removing Heel Cushion Change Footstrike?
Given growing interest among runners in how a given shoe might affect their footstrike, some running shoe manufacturers and running shoe sellers have begun reporting this value as a standard spec on each shoe. For example, Brooks Running has posted HT drop data on their blog for their more “minimalist” shoes, and Running Warehouse lists midsole height for the heel and forefoot for many of the shoes that it sells. I like this trend a lot, as it allows runners to consider one more piece of data to help them make a more informed choice of shoe to match their running style (e.g., it allows midfoot/forefoot runners to more easily find shoes with a lower drop if that’s what they prefer).
Despite the seemingly simple definition, I still find a few points confusing about the concept of heel-toe drop. Some places refer only to midsole height (e.g., Running Warehouse), whereas others (Brooks) include both midsole and outsole height in their calculation. Including the outsole does make a difference in some cases – as an example here are the stats reported on the Brooks blog for the Brooks Launch (one of my favorite shoes):
Midsole Height: Heel (22 mm), Forefoot (10 mm)
Outsole Height: Heel (3 mm), Forefoot (5.5 mm)
Heel-to-Toe Offset: 9.5 mm
Tooling Height: Heel (25 mm), Forefoot (15.5 mm)
You can see from the above example that the Launch actually has a slightly lower offset when the outsole is included in the calculation, due to the fact that the outsole is actually a bit thicker in the forefoot (tooling height, another new term to me, is oustole + midsole thickness). What confuses me is why you would ever exclude the outsole, and why the insole never seems to be included in these calculations? Seems to me that the goal is to determine how high off the ground the heel sits relative to the forefoot, and thus all three elements separating the foot from the ground should be included: insole (insert), midsole, and outsole.
Given this, I thought I’d do a little home experiment and measure the heel-toe drop in some of the shoes in my shoe rack. The challenge in doing this is that you can’t simply measure heel or forefoot height from the side of the shoe because the midsole frequently curls up around the sides above the level where the heel sits, such that you might overestimate thickness if measuring in this manner (not to mention that you wouldn’t be including the insole thickness). Rather, what I did was to use a bar clamp that could be cinched up inside the shoe on the insole and on the bottom of the outsole – in other words, I clamped from the top of the insole to the bottom of the outsole. By comparing the length of the clamp bar when fully closed to the length when cinched to the shoe (see photos below), I could calculate the difference and thereby estimate heel or forefoot thickness.
I measured heel thickness at the center of the heel, and forefoot thickness where the ball of the foot (i.e., metatarsal heads) would rest. The latter was accomplished by sliding the clamp arm through the laces between the tongue and the eyerow of the upper, so that no upper fabric was included in the measurement (this couldn’t be done in the Vibram Fivefingers KSO’s or Nike Sneakerboat, so I estimated the thickness of the forefoot fabric and subtracted it from the measured height). All measurements were taken on both the right and left shoes and averaged (measurements were remarakably consistent between sides on most shoes, giving me greater confidence in my data).
Clearly the measurements I provide below aren’t perfect, and a far superior way to do this would be to cut the shoe in half lengthwise and measure thickness with a calipers (which for obvious reasons, I can’t justify doing to my own shoes – but Newton Running has a picture posted of one of their shoes cut in this manner – see below).
Furthermore, all of these shoes have been worn for varying amounts of time, so I have no idea how much of an effect midsole compaction or outsole wear with increasing mileage might have on these measurements (that would be another interesting home experiment to do, and I hypothesize that if anything, compaction would decrease drop by compacting the heel cushion in a heel-striker like me; regarding outsole wear, most of mine is on the lateral corner of the heel, and was probably not a major factor in these measurements).
Despite the above considerations, I’ve made an academic living out of measuring microsopic parts of animal skeletons, so I’m pretty careful when it comes to taking measurements, and thus I’m confident that these numbers are at least reasonably accurate (and the method used was the same on all shoes).
Update 6/08/09: Ian Adamson from Newton Running was kind enough to leave several comments on this post, including one in which he lists the heel and forefoot thickness for each line of Newton shoes. Since my numbers for the Sir Isaac don’t match his, I thought I’d try to remeasure my Sir Isaac’s using another technique that is also easy to do at home. All that it entailed was measuring the height of a skewer stick, and then measuring the height to which it extended when lifted by the heel or forefoot of the shoe. You can see how I did this in the pictures below:
Below is what the resulting marks looked like on the paper:
You can see from the above that this method yielded a heel height of 33mm, and a forefoot height of 25mm, which gives a heel-toe drop of 8mm. These numbers are very close to those that I measured using the clamp method (see below – the fact that they are 0.5-1mm larger might be due to the clamp slightly compressing the insole when cinched), and I am confident in their accuracy. Not sure why there is a discrepancy with the numbers that Ian reported below (28mm heel, 23mm forefoot, 5mm drop).
Here are my numbers for heel height, forefoot height, and heel-toe drop using the clamp method:
Saucony Progrid Guide: 31mm heel, 18mm forefoot = 13mm drop
Brooks Launch: 28.5mm heel, 18mm forefoot = 10.5mm drop
Brooks Green Silence: 25mm heel, 16mm forefoot = 9mm drop
Newton Sir Isaac: 32mm heel, 24.5mm forefoot (measured at middle actuator lugs) = 7.5mm drop
Nike Lunaracer: 24mm heel, 18mm forefoot = 6mm drop
Brooks Mach 11 XC Flat: 16mm heel, 12mm forefoot = 4mm drop
Saucony Kilkenny 3 XC Flat: 17mm heel, 13mm forefoot = 4mm drop
Nike Free 3.0 v2: 23mm heel, 19mm forefoot = 4mm drop
Vibram Fivefingers KSO: 8mm heel, 8mm forefoot = 0mm drop
Nike Sneakerboat II: 9mm heel, 11mm forefoot = -2mm drop
Saucony Kinvara: 23mm heel, 19mm forefoot = 4mm drop
Nike Free Run+: 26mm heel, 19mm forefoot = 7mm drop
GoLite Amp Lite (w/ thickest forefoot insert): 19mm heel, 19mm forefoot = 0mm drop
-All men’s shoes above are size 10 except the Nike Lunaracer and Saucony Kilkenny, which are 10.5, and the Vibram Fivefingers KSO’s, which are size 42)
Asics 2130: 31.5mm heel, 17mm forefoot = 14.5mm drop
New Balance 768: 34.5mm heel, 20.5mm forefoot = 14mm drop
Brooks Adrenaline 9: 32mm heel, 19mm forefoot = 13mm drop
Nike Free 5.0: 29mm heel, 19mm forefoot = 10mm drop
-All women’s shoes above are size 9 except the New Balance 768’s, which are 8.5
Asics 2130 (Kid’s 12): 26.5mm heel, 18.5mm forefoot = 8mm drop
A couple of thoughts on the numbers above:
1. First, how do my measurements match up with those published for some of these shoes? It’s difficult to compare absolute heel and forefoot thickness since I included the insole, so instead I’ll focus on the numbers reported for heel-toe drop in a few shoes for which I could find published data on-line (insole height might explain some of the discrepancies):
Brooks Launch: my measured drop = 10.5mm, mfr. published drop = 9.5mm
Brooks Green Silence: my measured drop = 9mm, mfr. published drop = 8mm
Brooks Mach 11: my measured drop = 4mm, mfr. published drop = 6.7mm (after running almost 50 miles in the Mach 11’s, I suspected that the drop was lower than the 6.7mm reported by Brooks)
Newton Sir Isaac: my measured drop = 7.5mm, mfr. published drop = 5m
My conclusion is that it is often hard to know what the thickness numbers being reported by manufacturers include – i.e., just midsole, midsole + outsole, or insole + midsole + outsole. Seems to me the latter is the most accurate representation of where the foot would sit relative to the ground in any given shoe, and that’s what I have reported here (with potential inaccuracies noted above). I should point out that a 1mm discrepancy is very small, and probably negligible from a performance standpoint (and not to mention, also probably well within the margin of error).
2. You can clearly see my preference for minimalist shoes by looking at the drops on most of the shoes in my collection. A more typical trainer like the Saucony Guide (or most of my wife’s shoes in the Women’s section) has a drop of 12-14mm, whereas most of my shoes have a drop of less than 10mm.
3. The thickness of the heel in the Newton Sir Isaac (see picture below) surprised me. At 32mm, it has a heel thickness that is right in line with most of the traditional training shoes included here. This likely explains why I heel strike in them despite the fact that they are designed to be a shoe to help runners transition into a midfoot/forefoot gait. They have a lower drop than traditional training shoes, but this appears to mostly be accomplished by a thickening of the forefoot and addition of the actuator lugs.
4. Comparing my Nike Free 3.0 (see picture below) to my wife’s Nike Free 5.0 shows clearly that the Free 3.0 is the flatter soled shoe. The heel thickness in the Free 5.0 is not far off what you would find in a traditional training shoe.
5. Spikeless cross country flats like the Brooks Mach 11 (see picture below) or Saucony Kilkenny are truly very flat (4mm drop for both). They’re also very cheap, which makes them good choices if you can handle a narrow shoe and want to experiment with a low-drop shoe.
6. The Vibram Fivefingers KSO’s (see below) are a true zero-drop shoe.
7. The Nike Sneakerboat II (picture below) appears to have a negative drop, meaning that the forefoot is thicker than the heel. I’m not 100% certain on the accuracy for this one as it doesn’t have a tongue and I had to subtract an estimate of upper fabric thickness for the forefoot, but I need to take these on a run and see how they perform.
8. My wife, who I’m almost certain is a natural forefoot striker based on her shoe wear patterns (need to confirm this with video), clearly needs some more appropriate shoes for her gait. Speculation altert!!! – maybe this is why she’s had continual problems with running injuries and I haven’t?
9.My 6yo son’s Asics 2130’s have a lower drop than my wife’s 2130’s. This is good, but there’s still an awful lot of probably unnecessary cushion on the sole of that shoe. I’d still like to find him some sneakers with a flatter sole, but thankfully he spends most of his time barefoot or in Crocs.
Hope you found this helpful – I’ll probably add to this list as new shoes find their way into my house. I’d also encourage you to give this a try yourself, and feel free to report numbers for your shoes in the comments. If you have any of the shoes that I measured, I’d be interested to see how your measurements compare.