Ken Roberts - - Cross Country Skiing

what's here

see also:  more ideas + reports | public discussion | more on XC skiing

pole length for different skiing motions + situations


I started writing this because I thought I had some simple insights about pole length. Then the analysis turned out to be trickier than I thought, and some of the simple things turned out to be unimportant (or wrong). But then I was into the midst of it, so I just continued and wrote down here what I found instead.

what's here

"pole angle"

This concept is so central to everything here that I need to take the time to give its definition. Basically it just means "how far the pole shaft is slanted away from vertical. If the pole is pointing straight down to the snow, then the angle is zero. If the pole tip is pointing way back from the hand, then the "pole angle" is large. So normally the pole angle is small when you first set down the tip onto the ground, and gets larger and larger as the skier moves forward and the pole-push continues.

More carefully: I define "pole angle" as the angle between (1) the line from the tip of the pole to the effective point of force transfer between the skier's arm and the pole; and (2) the line thru the pole tip which is perpendicular to the surface of the ground where the skier's pole tip and feet are contacting.  So if the skier is on flat ground and the pole is hanging vertically down from the skier's hand, then the "pole angle" is zero. The closer the skier's hand is to the ground in front of the pole tip, the greater the "pole angle".

For now I will assume that in forward-propulsion situations (b) + (c), the effective point of force transfer is in the plane defined by the line perpendicular to the ground surface and the line of the skier's overall direction of motion.

limiters on pole too long

If the only goal is most effective forward-motion propulsion . . .

The critical limit is that the pole tip not lose its grip against the ground at the finish of the pole-push.

It might have been thought that the maximum length is determined by the skier's position at the start of the pole-push. But applying force and power at the start of the push always benefits from using a longer pole and just landing its tip farther back relative to the foot. Physics says that the more the pole is slanted away from vertical, the higher percentage of applied force and power gets translated into forward-motion propulsion (instead of just pounding down into the snow).

So why don't the top racers actually just land the pole tip farther behind? Because as the skier glided forward and continued pushing on the pole, the pushing hand would more downward and the longer pole would be aimed too much directly backward and then lose its grip in the snow and just skid away before the pole-push was finished.

I think many cross-country skiers feel that the problem with a pole too long is because it makes them reach too high at the start. But that's only because unconsciously they have learned that they cannot set down the tip too far back at the start because of losing grip at the finish.

Q: What if the pole-push is shorter, so it doesn't have to reach the maximum forward or backward arm positions? A: But (other things being equal) it's better to use the section of the arm's effective range-of-motion which goes to its point lowest to the ground, because then the average pole angle will be larger, so the average percentage of power translated into forward-motion propulsion will be larger. 

How this critical limit determines maximum pole length is based on two geometric quantities: maximum pole angle and height of the hand at the finish. Once those are known, it's simple trigonometry to get the maximum pole length: Divide the hand height by the "sine" of the pole angle.

maximum pole angle: a key factor on poles too long is how slanted away from vertical the pole can be without without losing its grip into the snow and just skidding away instead of transmitting pushing force.

This maximum angle of slant is a little different for different snow (or non-snow) conditions and pole-tip designs. For very hard snow with a dull pole-tip, the maximum pole angle would be smaller. Very soft (or ungroomed) snow might have a smaller maximum angle than a "normally firm" groomed snow surface. Rollerskiing on concrete has a lower maximum pole angle than rollerskiing on asphalt with a coarser-stone composition.

Surfaces that easily maintain grip of the tip tend to favor longer poles.

height of the hand: the other key factor is how high the hands are above the ground at the finish of the effective pushing thru the pole (not including "follow through" motion of the hand where it is not actually pushing any more).

How high off the ground hands are (or ought to be) at the finish of the pole-push depends on several things: the lengths of different bones of the skier's body (which tend to be proportional to the skier's height), the postural configuration of the skier's body at the finish (more bent or more straight), and the poling motion style technique and how well-trained are different specific muscles.

Note: the most effective classic striding technique for speed requires a lower hip position than most other skiing techniques. This results in a lower hand position, and by the trigonometric calculation leads to selecting a shorter pole. Interesting, more athletic skiers who tend to use a lower hip position might be using shorter poles. On the other hand, more athletic skiers are more likely to use other classic techniques such as pure double-poling which benefit from longer poles.

Usually guidelines for pole length are based on the skier's height. But a skier with unusual proportions of leg or arm bones might want to adjust the guideline.

not precise: Actually there's significant room for variation on when is a pole too long. Because the skier can modify the "technique" of how the poling motion goes at the finish of the push -- which would result in a different height of the hand. Once the skier trains a lot with a particular pole length, their muscular development pattern becomes optimized for that poling motion. Even if you don't get pole length exactly optimal for the finish, your muscles adapt to make it better. Switching to a more "correct" pole length is not likely to immediately improve speed until after several weeks of retraining and adjusting the pattern of muscular development to exploit that length.

So "trying out" different pole lengths to see which one feels better for you might not be the best approach for long-term speed. Might be better to choose the best pole length based on "theory", and over time the pole itself will mold your technique and feel for it.

How "manage around" poles too long? - If you use multiple techniques or encounter very different situations in one session of skiing with the same poles, you're going to have to deal with pole length not being optimal, which might be on the side of "too long". -- Two methods:

(1) Cut off your push early at the finish, so there's little risk of the tip losing grip on the ground and skidding away. You miss out on a little of the propulsive contribution from the arm muscles, but as compensation you get to start recovery toward the next pole-push quicker -- so can increase the turnover frequency of poling.

(2a) Raise your hand a little higher before setting down the pole tip into the ground. (I like this less than the first method.) If there is the normal "crash" of the upward body, the higher arm-hand position is going to be less effective at "rigidly" transmitting that kinetic energy into the poles -- more likely to be some "collapse" of the arms which absorbs some of that energy - (if you could reliably maintain rigid transmission of the normal "crash" energy with that arm-hand position, then that would just be your normal starting position, and this pole length would not be "too long" for you). Or you raise your arm up enough so it reduces the effective distance of the dropping of the weight of the torso, so it does not have time to accumulate so much kinetic energy, so there's less impact -- enough less so that you can "rigidly" transmit it even with this less effective higher arm position. (I don't like this because I think the "crash" strategy is an important way for bigger muscles contribute propulsive power to poling).

(2b) If you're using a technique or situation with little or no "crash" -- usually this would be with single-poling at higher turnover frequency -- then transmitting the impact isn't a problem. Instead you're just starting the pole-push with your arm in a biomechanical configuration where it's significantly less effective for generating power - (because if your arm muscles were pretty effective in that higher hand position, then this pole length would not be "too long" for you). The downside is that including a less effective phase in the stroke-cycle reduces the turnover frequency (more than people think, because ineffective phases also tend to be slow), and turnover frequency is one of the three key drivers of power, and reducing the frequency poling might also reduce the turnover frequency of the leg muscles.

 limiters on pole too short

There is a different key limiter (if propulsion is the only goal) for avoiding choosing a pole too short:  If the skier tries to maximize vertical potential energy by raising the weight of the torso + shoulders high as possible, then with short poles they must crash down too far before being "caught" by the pole-push and transmitted and converted into forward-propulsive work. "Too far" because the impact is too much for the hands and arms to transmit without "absorbing" too much of it.

If try to minimize the crash downward distance by landing the pole tip farther forward, then the pole is too vertical at the impact of the weight up the upper body -- pole angle too small, so too high a percentage of the kinetic energy just goes into pounding down into the snow, not enough percentage goes into forward-propulsive work.

The other way to minimize the crash downward distance is to have the arm more extended (straighter) and aimed more downward at the moment the pole tip is set down.

How "manage around" poles too short? - If you use multiple techniques or encounter very different situations in one session of skiing with the same poles, you're going to have to deal with pole length not being optimal, which might be on the side of "too short". -- Two methods:

(1) Just don't "unbend" your body so much just before landing the pole tip into the ground. Don't get your shoulders up so high. This implies that you're getting less contribution to propulsive power from your back-extension and leg-extension muscles in between pole-pushes. On the hand you can increase the turnover frequency of the pole-pushes.

(2) Have the arm more extended (straighter) and aimed more downward at the moment the pole tip is set down. This approach also requires less "isometric" arm strength for catching and transmitting the "crash" impact -- good if you haven't trained your arms for it. The disadvantage is that you give up some range of motion in using your arm muscles to actively push in a later phase of the pole-push, after they've finished their key role of transmitting the energy of the "crash".

trade-offs too long versus too short

fast versus slow skiing: At higher speeds the finish of the push tends to be a little less important, because the arm muscles are not able to deliver power (i.e. power as defined in physics and measured on Watts) as effectively at higher muscle speeds. At the start of the pole push, the speed of the hand can be significantly slower than the overall velocity of the skier's body relative to the ground, so muscle speed is not a problem. Near the finish the speed of the hand must match the skier's overall speed over the ground, so difficulty of the muscles delivering power at higher speed can become a factor.

Therefore the finish phase is a little less important for higher skiing speeds. So skiers who tend to be in situations that emphasizes higher speeds might want a little less worried about the pole tip skidding away just before the finish of the stroke -- or a little more likely to finish pushing a little earlier so that skiing away would be less likely to be a concern. Therefore they might want to "err" on the side of choosing a pole a little longer.

Skiers at slower speeds tend to benefit from continuing the push farther back, because the percentage of power which is translated into forward-motion propulsion is higher when the pole angle is larger. So for them it's more important to choose a pole length that assures that the tip will maintain grip and not skid away just before the end of their push. Therefore they want to "err" on the side of choosing a pole a little shorter.

high-force versus low-force situations: The usual high-force situations are climbing up a steep hill or accelerating at the start of a race. There's two concerns for applying high forces with poling: recruiting more muscles for direct push, and transmitting vertical kinetic energy from dropping body weight onto the poles. Recruiting muscles (e.g. abdominal crunch) other than the obvious arm muscles tends to work best earlier in the pushing stroke. "Catching" and transmitting vertical kinetic energy can only happen at the start of the push. Both of these tend to favor a configuration that translates a higher percentage of power into forward-motion propulsion earlier in the stroke, which happens when the pole angle is larger, which tends to be easier with longer poles.

Now the longer poles which benefit the early phase are more likely to have grip problems at the finish of the pole-push. That's a trade-off. The point is that a skier concerned about high-force situations is more likely to want to "err" on the side of favoring the starting phase, even if that might result in neglecting the finish a little.

low turnover versus high turnover frequency of poling: Lower frequency of poling tends to allow more time in between pushes to use the back-extension muscles to raise the weight of the torso and shoulders and head higher. So there can be more gravitational potential energy accumulated in this "recovery" phase between pushes. When this potential energy is converted into kinetic energy by dropping the weight of the torso onto the poles, it adds force and power to the early phase of the pole-push. (This kinetic energy can also be increased by starting to contract the abdominal "crunch" muscles before the pole tips land on the ground).

This "catching" of the kinetic energy of dropping can only happen at the start of the pole-push. The higher percentage of this downward energy gets translated into forward-motion propulsion the better. This percentage is determined by how much the pole angle is slanted away from vertical: larger angle implies larger percentage. Larger angle is easier to obtain with longer poles.

Actually for vertical kinetic energy and force, the increasing percentage translated into forward-motion propulsion is limited to a poling angle of 45 degrees (or a little more for climbing a hill). Above that angle the percentage decreases. But since grip of the tip at the finish typically is lost around 55-60 degrees pole angle away from vertical, and the total range of pole angle is around 35-40 degrees for the full stroke, then the starting pole angle must be around 15-25 degrees -- so there's no danger of using a poling motion configuration or pole length which mostly results in reducing the translation percentage.

single-poling motion techniques tend to have a higher turnover frequency of poling -- "single-poling" means pushing with only one pole at at time on alternate side of the body. Typical techniques which use single-poling are: classic striding ("diagonal") and "herringbone" climbing on cross-country skis, single-poling ("herringbone") skate on cross-countryskis, walking with poles on dry land, and ski mountaineering climbing.


I made some indoor measurements of my pole positions which felt "normal" for classic striding, using a pole length just a little more than more normal classic-technique pole and just socks on my feet.

Starting position: hand-pole connection around 48 inches above floor and 14 inches forward from pole tip basket on floor. (so that's a pole angle of about 16 degrees from vertical).

Finishing position: hand-pole connection around 32 inches above floor and 41 inches forward from pole tip basket on floor. (so that's a pole angle of about 52 degrees from vertical).

So the distance that the hand-pole connection moved forward during the pole-push was about 27 inches. But the distance my foot moved forward was about 39-40 inches. That's because my foot moves forward relative to my hand. A typical walking stride might be 24 inches, not more than 30 inches. So there is no way that anything like normal walking (with no glide) can use the full range-of-motion of one single-pole-push for each step by each foot (which is two single-pole-pushes for each full-stroke-cycle of one step each by both feet). Perhaps that's why some walkers sometimes use one double-pole-push for each full-stroke-cycle of one step each by both feet.

Lowest finishing position available: hand-pole connection around 27 inches above floor and 43 inches forward from pole tip basket on floor. (so that's a pole angle of about 58 degrees from vertical).

different motions and situations

classic versus skating:  Classic striding uses a shorter pole than skating or pure double-poling, because the big performance limiter on power from the legs in striding is in the length of the leg-push -- tends to be too short, because the foot must be held fixed on the ground during the leg-push. To make the time + distance of the leg-push longer, try to keep the foot in contact with the ground longer. This is accomplished by bringing the hip lower thru most of the push. If the hip is lower, then the shoulder tends to be lower, and the hand tends to be closer to the ground at the finish -- so you end up with a shorter pole length.

The skating leg-push does not require the foot to be fixed in contact with the ground, so its push can be spread over a longer time without dropping the hip lower -- instead just glide longer while pushing.

Also, classic striding tends to have higher frequency of poling and lower overall speed of skiing than most skating, and both of these factors tend to favor shorter poles - (see above under trade-offs).

"classic" motion techniques:

classic striding ("diagonal") tends to be higher-frequency, higher-force, lower-speed than kick double-pole, and like kick double-pole requires a low hip position in order to achieve a longer more effective leg-push. All but one factor indicates shorter poles - (see above under trade-offs)

herringbone climbing: does not require low hips to achieve a long leg-push like classic striding, so that indicates longer poles than classic striding. On the other hand, herringbone climbing tends to be higher-frequency, higher-force, lower-speed than pure double-poling (or skating) -- 2 out of 3 favor shorter poles than for pure double-poling or skating.

pure double-poling: tends to be higher-speed, lower-frequency, lower-force than all other classic techniques. And does not require low hips to achieve a longer leg-push. Tends to favor longer poles like skating.

kick double pole: Requires  a low hip position in order to achieve a longer more effective leg-push, which indicates significantly shorter poles than for pure double-poling. But the lower frequency than classic striding tends to indicate poles not as short.

special racing situation: A course which is mostly flat with a few short steep climbing sections could be handled with minimal grip wax and double-poling the flat sections and herringbone for the steep climbs -- and using long poles.

skating motion techniques:

V1 (a.k.a "offset", "paddle-dance") : tends to be lower-frequency, higher-force, lower-speed than V2 skate -- 2 out of 3 favor longer poles (see above under trade-offs) than V2 - (unless you favor a high-frequency style of V1 "light and quick").

V2 (a.k.a "1-skate", "double-dance") : tends to be higher-frequency, lower-force, higher-speed than V1 skate -- 2 out of 3 favor shorter poles than V1 - (unless you favor a lower-frequency style of V2).

open field skate (a.k.a "2-skate", "V2 alternate", "single-dance") : tends to be lower-frequency, lower-force, higher-speed compared to other skating motions -- 2 out of 3 favor longer poles.

single-poling skate (a.k.a. "herringbone skate", "diagonal skate", "coaches skate"): tends to higher-force, higher-frequency, lower-speed compared to other skating motions -- 2 out of 3 favor shorter poles.

inline skating on asphalt versus skis on snow:

Of course inline skates normally put the feet and the entire body higher off the ground, which obviously favors longer poles. Beyond that . . .
inline skates on smooth asphalt tends to be higher-speed, lower-frequency, lower-force than skating on snow -- 2 out of 3 favor longer poles.

Note that for skating on flat or gentle asphalt, there's lots of other ways to add speed than by poling. Even without poles, just swinging the arms from side-to-side adds power to skating propulsion. Typically poling gets in the way of the best rhythm of the more clever ways to use the leg muscles and other body motions to add power.

The big problem with using poling to increase speed is that it increases air resistance -- which is very very important at the speeds of fast skating.

One situation where poles could help speed is up long steep hills. Otherwise when you're not worried about optimal speed, they can be fun to play with.

ski mountaineering racing (or "ski-rando" or "randonnee" or "backcountry ski" climbing):

Tends to be slower-speed, higher-frequency, higher-force than classic striding on gentler terrrain. 2 out of 3 tend to favor shorter poles - (see above under trade-offs)

The "higher-force" factor (which favors longer poles) can be very important for the steep climbing, but soft ungroomed snow tends to make poling less effective for helping with this, and race courses often have pre-set climbing tracks which are less steep. Also the variable snow quality might not reliably allow grip of the pole tip at higher pole angles near the finish of the stroke, which favors shorter poles.

On the other hand long leg-stride length is not as critical for ski mountaineering as for classic striding, so low hips are not as important, which allows using a longer pole. (but leg-stride length is important for the very fastest ski mountaineers).

So I think that serious ski mountaineering racers tend to use pole length roughly similar to serious cross-country ski racers. Much longer poles than most ski mountaineering tourers.

Note that unlike most poles for ski mountaineering touring, ski mountaineering racing poles have a tip that is designed to grip the snow well at large pole angles.

ski mountaineering touring (or "backcountry skiing" or "randonnee" climbing):

Poling tends not to be very effective in many ungroomed backcountry snow conditions. Ski mountaineering poles are not well-designed to grip the snow at larger pole angles. Ski mountaineering poles tend not to have well-trained poling muscles or to know or care much about poling technique. They tend to be much more concerned about downhill turning performance and control in often difficult snow conditions.

Therefore ski mountaineering tourers tend to use downhill-length poles (which are much shorter than poles for ski moutaineering racing or cross-country skiing).


Rollerskis are supposed to simulate skiing on snow, so most of the considerations for different pole lenghts for different techniques are the same as for on snow.

The big difference between using poles on rollerskis versus on snow is that on asphalt it is significantly more difficult for the pole tip to reliably maintain grip thru the finish of the pole-push.  Grip tends to better on coarser asphalt or packed dirt, worse on smoother asphalt, worse on cold days, worst on smooth concrete. Keeping the tip sharp makes a difference -- normally requires a diamond file or diamond grinding wheel.

On the other hand, the usual design of a rollerski puts the foot a little higher off the ground, so that tends to call for a longer pole.

Therefore based on the calculations above under limiters of pole too long, with favorable surface conditions and recently-sharpened tips on asphalt, rollerskiing typically calls for roughly the same pole length as snow skiing.

But if pavement is smooth or cold or you haven't sharpened your tips recently and competently, then rollerskiing calls for shorter poles than snow skiing.

A subtle difference between rollerskiing and snow-skiing (especially softer snow) is that applying force thru the poles takes weight off from the skis, and that matters a lot on snow. Because snow is more compressible than asphalt, the resistance of sliding friction is significantly reduced with less weight on the ski -- especially in softer snow where the ski is somewhat "plowing" into the snow. There is some reduction of rolling friction of a rollerski wheel over asphalt (more so a softer or pneumatic wheel or on softer dirt or coarser asphalt) -- but I suspect snow often tends to be more non-linear in this, and surely softer snow is much more non-linear than rollerski wheels. Reducing weight on the rollerski does not change the friction of a variable-resistance device that acts between the ski and the wheel.

I think the danger of lots of rollerskiing practice with poles not short enough is that you tend to compensate in various ways:

  • by shortening the finishing extension of the pole-push.

  • by setting the tips down too far forward relative to the feet. So the starting pole angle is too small, so the "dropping" of upper body weight onto the poles is not effectively translated into forward-motion propulsion. So your unconscious neuro-muscular control module starts perceiving that dropping upper body weight is less important, so gives it a smaller role in the stroke-cycle, doesn't train those specific muscles as well.

  • by reaching forward with the hands just before the start. One result is to mess up the "imprinting" of neuro-muscular coordination of the timing of the arm-push phase, which is better left for the final phase.

The problems are that: (1) you can't just consciously re-adjust those things when you switch to snow-skiing. Cross-country skiing motions are much too complicated to be guided by the rational conscious mind. Maybe your conscious mind can control one or two factors (and also slow them down), but there are ten others which it must ignore at the same time. (2) repeating these motions hundreds and thousands of times during a session of skiing requires specific muscular training. If the training of specific muscles has been neglected or unbalanced because of "compensation" during two months of rollerskiing for incorrect pole length, then it's going take significant time to train those muscles up to speed for snow skiing -- and in the mean time their deficiency might hold back other muscles from their best training, or lead them to other "compensations".

The biggest danger is that lots of practice and training with incorrect pole length might lead to your muscles and your unconscious neuro-muscular control module settling into a self-reinforcing pattern which is sub-optimal -- sort of an "ecological system" which is locally optimal, so that any single change or any set of small changes makes it feel worse (at least for several weeks). So it's difficult just to find a path, and then difficult to perservere along that path long enough get "over the hump" into the long-term better pattern.


Get really good at sharpening your pole tips.  Don't train on concrete. Avoid training on smooth asphalt in cold weather -- prefer coarser asphalt even though it doesn't feel as good.  Avoid training with rubber-tip poles.

"dryland" training for cross-country ski racing:

I've heard it recommended that for training exercises and drills on "dry land" (i.e. no wheels or snow to permit any gliding), the pole length should be shorter than for snow skiing or rollerskiing.

Usually "dryland" training is done on grass, so I don't see any problem with the tip losing grip at the finish.

The feet are a little closer to the ground than for snow-skiing (and significantly closer than for rollerskiing) -- so that would be part of the reasoning.

Or maybe the reasoning is that the whole pole-push stroke tends to be shorter, and  that somehow shorter poles makes it a better simulation of poling on snow, or less interferes with the simulation other non-poling motions.

My feeling is that poling on grass is so different from poling on snow (except climbing very very steep snow) that I don't see any point in doing it or practicing with it. So I don't.

"nordic" walking with poles

I don't think of "nordic walking" as a speed thing. Rather it's more of a pleasant lower-impact exercise thing.

Using poles does burn a few more calories than just walking. Pushing with the poles could make you go a little faster than normal walking without poles. With walking there's no way the arms can add propulsive work (except with poles -- or perhaps wing-like fans). That's because if at least one foot is in effective contact with the ground at all times, any positive work added by an arm-swing move to one phase of the stroke-cycle must be exactly cancelled by negative work from the opposite arm-swing move in another phase. So in walking arm-swing is only for balance (and burning extra calories). Using poles does allow the arms to contribute to propulsion.

But if you're really serious about going faster or burning more calories per minute, it's simpler to just run (or at least jog, if "run" sounds too serious). Running or jogging does permit the arm-swing to contribute to forward-motion propulsive force and power -- provided there's a time in the stroke cycle where both feet are up off the ground in the air simultaneously.

Nordic walking with poles increases power by recruiting some additional small muscles. Jogging or running (without poles) increases power (and calorie-burning) by greatly increasing the utilization of some big muscles in the legs - (and also by starting to utilize some power from the arms). Seems clear to me that switching to running has a bigger effect than adding poles.

But running puts more impact on knee and ankle joints and various ligaments in the foot and leg. And some people just don't like the feeling of jogging or running. So for them adding poles to walking is a benefit.

running with poles

There's little evidence that using poles helps with running speed on flat or gentle terrain. At running speeds the additional pushing force from using poles just isn't very relevant. The critical bottleneck is how quickly you can more your legs back and forth to contact the ground with minimum losses. What matters is finding ways to increase muscle speed, not adding more force intensity.

The arm-swing if correctly timed can already add force to the leg-push in running (though not for walking), so there's no need use poles in order to recruit the arm muscles to help propulsion. Poles just tend to get in the way.

Possibly with lots of practice, using poles could help with running up a very steep hill, or with occasional steeper moves on irregular terrain. But usually if the terrain is steep enough to benefit from poles, it's too steep for real running to be effective. (If real running is defined to include getting both feet simultaneously up off the ground into the air for any significant time.

I do find the poles helpful for doing down steep hills -- to absorb the impact.


There are basically three different roles for using a pole in skiing:

(a) for balance

(b) to transmit direct muscular force to the ground and convert it for forward-motion propulsion

(c) to transmit gravitational force or vertical kinetic energy to the ground and convert it for forward-motion propulsion.  

simple principles for each role

(a) balance -- the usual goal of balance is to keep the skier's body from falling down onto the ground. So the forces which need to applied are roughly vertical, with pole tip roughly underneath the skier's hand. The skier's arm is held mostly static while transmitting the force. Most people find it easier to transmit a static force with their arm more extended (less bent), and for the vertical force, to extend the arm roughly downward.

So the hand-pole connection is lower, and a short pole is clearly best - (unless the skier is on a steep slope and needs to plant the pole below on the slope).

(b) transmit direct muscular force into forward propulsion. The smaller the pole angle, the lower percentage of muscular force goes into forward propulsion, and the more goes into just compressing the snow down. (The percentage which goes into foward propulsion is roughly the "sine" of the pole angle). If the pole is nearly vertical when skiing on flat ground, almost all the force goes into just pounding down into the snow. The larger the pole angle, the higher the percentage of the direct muscular force goes into forward propulsion.

This generally tends to favor a longer pole, but there are specific factors which can override this.

For now I am assuming that the direct muscular force is applied roughly in the direction of the shaft of the pole.

(c) transmit vertical force or kinetic energy into forward propulsion. On flat ground, the optimal pole angle for converting vertical force into forward horizontal force is 45 degrees -- so for vertical force you want a pole definitely longer than the shortest that would reach the ground, but not too long.

I doubt this principle has any implication for realistic skiing propulsion situations than (b) -- because it's pretty difficult to have the whole range of motion at a pole angle greater than 45 degrees -- because there wouldn't be much motion.

If climbing up a hill, the optimal pole angle is 45 degrees plus half the slope angle of the hill. So for climbing up a 10-degree hill, the optimal pole angle for converting vertical force into propulsion is 50 degrees - (which by my definition of "pole angle" is 50 degrees from perpendicular to the slope, which is only 40 degrees from vertical of gravity).

Derivation: The percentage of vertical force applied along the shaft of the pole is roughly the cosine of the pole angle minus the slope angle. Then the percentage of force along the shaft which is applied in the direction of motion (forward and slightly upward) is the sine of the pole angle. Find the maximum by dIfferentiating that product of cosine and sine and setting the derivative to zero. The derivative is the cosine of twice the pole angle minus the slope angle.


changing of angle:  The pole angle keeps changing while the pole is applying force to the ground. Because the skier is moving somewhere, but the pole tip must be held at a fixed point on the ground while it is transmitting force. So you always get a range of pole angles -- normally land the pole with a smaller starting pole angle and lift it up with a much larger finishing pole angle. Most of the pole-push is not made at the optimal angle. Instead you try to choose a range that contains the optimum somewhere near its middle.

grip puts limit on larger angle:  If the pole angle gets too large, the tip of the pole just slips away instead of transmitting force. Usually there needs to be some significant downward force to hold the tip gripping onto the ground. Or sometimes the pole tip can deform the ground so that it gets embedded somehow. So even though principle (b) says the larger pole angle transmits a larger percentage of direct muscular force, there's a limit on this.

forward skiing speed is driven by power, not force:  If the skier's goal is speed, the main driver of the physics of speed is power (measured in Watts), and power is roughly force multiplied by velocity (with all the factors carefully correctly defined and understood). The problem is that power is more complicated to analyze and understand than force. A more fundamental problem is that our muscles tend to produce their highest sustainable force at very low muscle contraction speed -- and at very high muscle contraction speed are much less effective for applying force. So there's a trade-off between the two main components of power:  force versus velocity.

changing pole angle is like shifting gears:  Small pole angle allows the arm to move more slowly than the overall speed of the skier relative to the ground, so the skier's muscles can generate more force. Large pole angle forces the skier's arm to move almost as fast as the skier's overall speed, so the skier's muscles cannot generate as much force. On the other hand, a higher percentage of that muscular force is converted into forward propulsion. So smaller pole angle is like the "high gear" on a bicycle -- good for handling higher speeds. The odd thing with poling is that the "gear" changes during the pole-push -- so we can think of only a sort of "average" gear for the push.

calculating velocity relationships

I found that with a simplified kinematics of the arm and the pole, it's possible to solve for velocity relationships (e.g. shoulder-arm angular velocity as a function of skier speed during different phases of the pole-push), even if I couldn't derive a closed-form solution of the positional values. Because the basic kinematic formulas are linear combinations of trigonometric functions and constants, so they are straightforward to differentiate to get velocities, and differentiation makes some of the constant terms go away, and then it can be easier to find a closed-form solution. Suppose the arm is one link and the and the pole is another link in two dimensions.
Define . . .

θ = pole angle

φ = arm angle (angle of line from shoulder to hand-pole connection with horizontal line thru shoulder joint)

L = fixed length of pole

r = current distance from shoulder joint to hand-pole connection.

x = how far the shoulder joint is in front of the pole-tip ground contact point in the forward horizontal direction.

y = how far the shoulder joint is above the pole-tip ground contact point in the vertical direction.

kinematic formulas

x = L sin θ - r sin φ

y = L cos θ + r sin φ

eliminate θ by substituting

L cos θ  =  sqrt [ L2 - (x - r cos φ)2]  =  y - r sin φ

which yields

2 r (y sin φ - x cos φ)  =  x2 + y2 + r2 - L2

then can differentiate with φ, x, y, r as functions of time and solve to get the angular speed dφ/dt  as a function of positions and of speeds dx/dt, dy/dt, dr/dt.

And make some simple assumptions about those speeds at different phases of the stroke-cycle: e.g. dx/dt is somewhat larger than overall skier forward speed early in the push, but tends to be roughly equal to overall forward speed later; dy/dt is positive in the first half, then zero in the second half; and dr/dt is small at the start, but large near the finish.

Which is a great oversimplication, but could produce helpful insights about "how fast do the shoulder and arm muscles have to move" in order to "keep up with" the overall skier forward speed at different phases in the stroke-cycle.

Goms - Oberwald - Reckingen (Switzerland)

(see also 2007 report)

Sharon skied from Oberwald to Reckingen on a Saturday in late March. That's in the Goms area which is the high northeast end of the Rhone river valley in Switzerland (sort of between Andermatt and Brig). It's a pretty area for cross-country skiing [ photos ]. Sharon had a good time. Ken mostly provided the car shuttle, and also some skiing of his own near Reckingen. Sharon rented classic no-wax skis in Oberwald, which worked very well for the conditions. Snow coverage was excellent.

Seems like most skiers we saw were going in the direction from Oberwald to Niederwald - (? perhaps they took the train in the uphill direction, and skied one-way in the (southwest) downhill direction ?)

Ken found it tricky to find parking near the main ski trail which was definitely legal. The main ski trail tends to go closer to the south side of the Rhone river valley (where its snow is more protected from the sun). The main road and the towns tend to be on the north side of the valley. There's an alternate "sunny" ski trail on some of the higher parts of the Rhone valley -- but the sunny trail doesn't usually go near to the main trail (though the sunny trail tends to be closer to the main road and to parking).

Our car shuttle plan ran into trouble in Ulrichen when Ken parked near the "sunny" trail waiting for Sharon, but Sharon skied from Oberwald to Ulrichen on the main trail, so she arrived nowhere near where Ken was waiting.

Here's some possible parking places Ken explored:

  • Oberwald: near the train station.

  • Ulrichen: road between Ulrichen + Geschinen parallel to the main road is not far from the "sunny" trail (but not so close to the main trail).

  • Geschinen: road between Ulrichen + Geschinen parallel to the main road is not far from the "sunny" trail.

  • Reckingen: Ken drove down to and across the old wooden bridge over the Rhone river, then parked just south of the bridge.

(We didn't look for parking further SW below Reckingen.)

two (or more) styles of V1 skate

Further thought later in 08april:

Similar thoughts about V2 skate -- that vertical position of hips does matter. For low and moderate speed V2, the same argument about how "higher hip position wins by two votes to one". Possibly at higher speeds the lower hip position might win for V2 (also because of aerodynamics) -- but not for V1 because V1 is not intended for higher speeds.

What about Open Field Skate (V2A)? Since V2A is usually meant for higher speeds, it's easy to make the argument for lower hip position on the Recovery-side, with the reasons of better high-speed performance of the Extension and Sweep-out muscle moves, and aerodynamics. On the Poling-side of Open Field Skate all those arguments apply, but there's another factor: the objective of getting a very fast pole-push, in order to "keep up with" the high speed. And it also has to be a forceful pole-push, to make a significant contribution to propulsion (as opposed to just going thru the motion). This requires a substantial "wind up" preparation move, which one might guess should involve getting up as high as possible. Perhaps if you're not going to make the high preparation for a very fast and forceful pole-push, you might as well not be making a pole-push at all -- just skate with no poling. Or if you're not going fast enough to require a "full wind up" for the pole-push to make a significant contribution to propulsion, maybe you should be doing V2 instead of Open Field Skate. Not sure what the answer is for the Poling-side.

Further thought later in 08march:

Now I'm thinking that vertical position of hips does matter. Lower hip is better for direct propulsion from the Extension phase of the skate-push. But higher hip is better for lower-speed higher-force-intensity use of two other muscle moves: (1) raising the weight of the hips and upper body to build potential energy for the pole-push; and (2) medial hip-knee rotation in phase 1 of the skate-push. So higher hip position wins by two votes to one.

08march posted to xcskiforum, see also new stuff on V1 skate page

In the last week I've started thinking there's two different styles of V1 skate, after I had assumed that there was only one that mattered. And I'm thinking now that one is more effective for climbing up a hill - (not the one I assumed). The two styles are:

(A) skating focus:  compromise poling to maximize effectiveness of leg skate-pushes.

(B) poling focus:  compromise leg skate-pushes to maximize effectiveness of poling.

I think the big leg-extension muscles (e.g. gluteus and quadriceps) supply the majority of the power for both, so I can see it as a question about how to use those muscles. Skating leg-push "aims" the leg-extension push more outward sideways and backwards, while double-poling "aims" the leg-extension push downward. So you can't maximize both at once. (and which one gets maximized has implications for some other moves + positions)

Which is more effective?
Here's my analysis:

  • Double-poling reduces gliding friction during the poling phase, because it takes some weight off the ski (but increases it during the recovery phase).

  • Skating leg-push increases gliding friction during its push (and doesn't decrease it ever).

  • The physics for converting muscular work of double-poling into forward-propulsion output is simple: two-step sequence (build gravitational potential energy, then convert into downward kinetic energy) with no tricky timing and then a single 2-dimensional direction change (downward into backward).

  • The physics for converting skating leg-push into propulsion is complicated: three different directions each with its own different physical principles: (1) horizontal perpendicular to ski-glide direction; (2) horizontal parallel; (3) vertical. And one dimension depends on special timing. If everything were perfectly constructed and coordinated, the losses could be minimized, but . . .

So my guess is that double-poling is 10-20% more effective in using the big leg-extension muscles (except at higher speeds). So style B wins for climbing hills. I think that's how the elite racers are doing their V1 skate.

and their "jump skate" V1 is the more radically focused version of the "style B" strategy.

I think the key observable difference is that style B has more vertical motion of the hips:

The dropping of the hips (along with the weight of the upper body) takes place mostly in the second half of the recovery-side push (maybe also a little in the start of the poling-side push). Many elite racers make their pole-plant before the set-down of the poling-side foot (not simultaneous like the normal simplistic explanation of V1). So most of the dropping of the butt and transmission into the pole-push is over before the poling-side skate-push even starts.

Then the raising of the hips (and weight of the upper body) starts before the poling-side push finishes and continues thru the first half of the recovery-side push.

Position of the hips is not what the difference is about:

Like I think I had a "low hips" version of style A, but I suspect lots more people are using a "high hips" version of style A.

Likely some elite racers do style B with hips starting "medium"-high then dropping to "low" then back up to "medium" -- while others do style B with hips starting "high" then drop to "medium" then back up to "high"

I'd be glad to hear criticisms of these two styles of V1 -- or proposals of other different styles.


Sometimes skaters have asked what sort of summer dryland exercises they can do to train for "stepping up the hill" in ski-skating. One reply I've seen: You do not "step" up the hill, you should thinking about skating up the hill. So which is the better mental image for V1 up a hill: stepping or skating? I'm saying the answer is: Neither. The most effective way for an athletic well-trained skier is to double-pole up the hill (with some assistance from skate-pushes).

Not the answer I ever expected. Not the answer I wanted: I was actually pretty good at pure Classic double-poling, but I didn't much like training for it (one reason I stopped doing Classic races).

Funny thing is I remember reading about a research study done in the 90s where they measured the forces and found out that elite XC ski racers skating up hills got more than half their power from poling (exactly what I'm saying here now). But at the time I just couldn't believe it.

Partly because I thought poling was about upper body muscles -- didn't yet know that it was a leg-muscle thing. Partly because I had just seen a guy win a big uphill rollerski race without using any poles.

This theory of V1 and the evidence supporting it is kinda old -- just not well known.

See also new stuff on V1 skate page

jump skate -- different kinds

08march posted to xcskiforum

someone wrote (in response my mention of asymmetric one-sided jump V1 skate by elite racers):

Please explain to me why Jump V2 doesn't seem to exist for explosive climbing?

I have also observed in an Olympic 2006 race video a "jump" Open Field Skate (V2A) used in starting sprints on flat terrain -- again a one-sided jump.

The place I've seen speedskaters go airborne symmetrically on both sides is on videos of starting sprints on inline skates (not skis) with No poles. I think they're simply pushing so hard in each skate-push that the result of going air-borne is an unintended side-effect.

My best guess about V2 versus V1 for jumping is that using a jump in most situations of cross-country XC ski skating with poles incurs significant "overhead expense" per jump. So if you're going to incur that per-jump overhead, better to make it one big jump than two little ones.

Second best guess is that it takes a lot of peak strength from the leg-extension muscles to launch a jump. When it's a big jump on only one side, it's easier to use both legs more to perform the launch -- to spread the peak force over a larger muscle mass. When jumping more frequently toward both sides, the launch burden falls more on each leg independently, so there's a more intense peak stress on muscles, which tires them out faster.

video of elite racer:

In the video of the 2006 Olympic sprint Men's finals, my memory is that when Bjorn Lind used Jump OFS a little later in his starting sprint, he was jumping from his recovery-side onto his poling-side, so he was launching mainly off his recovery-side leg -- but he also set down his poling-side ski (aimed roughly straight forward) just before the jump. Then he landed on his poles and onto that same poling-side ski (now angled out to the side).

Seems likely that he was using the big extension muscles of the poling-side leg to help launch the jump. [He might also have been using its "sweep inward" muscles (e.g. hip-adduction) to add some sideways motion, but I now think that's at most a secondary benefit)

classic double-poling to improve skate technique

08mar7 posted to xcskiforum

somebody posted

[ links to two videos of their skating technique -- and asked for suggestions. ]

Looks like good side-to-side weight shift. Good set down of foot close underneath, not out to the side. Solid basic side-to-side balance, not getting the hips "caught in the middle".

Hard to see the subtleties of leg-moves from those two videos, but anyway I think the "low-hanging fruit" for this skier to increase speed is in Poling.

The poling in these videos looks too arm-focused, needs more help from some abdominal crunch, and some up-and-down motion of the butt.

Bicyclists have a good excuse for not having much abdominal crunch to go down or much back-extension for getting their shoulders back up again. But bicyclists have no excuse for not getting their butt into poling, because the muscles for raising the weight of the butt (and with it the entire upper body) are exactly the big leg-extension muscles used to push down on a bicycle pedal.

I know this seems like a "classic" skier's thing, but I suggest lots of practice doing pure double-poling. Start finding out how steep a hill you can double-pole up with no use of leg-pushes. Find out what really works for delivering serious force thru your poles in a situation where there's no way to fake it.

Watch some Classic-technique race videos and see how the best racers do their double-poling. Get video of your own double-poling and compare. (? Try some Classic races yourself ?)

My advice is to make it your goal for classic double-poling to be comfortable getting your butt and upper body weight up so high and so forward just before starting the pole-push, that you're up on the balls and toes of your feet, with your heels off the groun

(then Finally get comfortable doing jump double-pole.)

Once you feel how much power and speed double-poling can really deliver on its own, start playing with how to introduce that power into your V1 and V2.

ski France


some useful websites:

  • Ski Nordique France: The source of cross-country trail + snow conditions that I check first. Also other helpful info.

  • Savoie Ski de Fond: Helpful info and map of the better cross-country ski centers in the department of Savoie.

Peisey-Nancroix: Pretty skiing in a quiet valley between Moutiers and Bourg St Maurice, between the La Plagne and Les Arcs downhill ski resorts. Pretty because the trails go near the base of big spectacular mountains (Dome des Platieres and Mont Pourri) and because of lots of "rural mountain style" stone buildings alongside the trails. Trail network is basically a loop with variations. And basically "up and down the valley", but nicely done (especially on the blue loop) to avoid long climbs on the uphill side by breaking them up with flat or downhill sections. The slope of the valley is steepest near the nordic welcome center, and got gentler as I climbed away and up the valley. So if you want to do more distance than a single loop of the blue trail, to minimize climbing use the "bouclage" cut-off trail  to stay high, instead of returning down to the nordic welcome center. For those want tougher climbs, the black loop offers them. The report we saw indicated 47km of trails, and they were all groomed and still open in early April when I visited. [website]


la Féclaz (Savoie Grand Revard): A great place gets better. Sharon + I parked at Saint François de Sales on a Saturday. Sharon skied to la Feclaz, and I skated to Refuge de la Plate. Then I noticed there was a trail I had never seen groomed before. So I tried it. It started with an unusually steep uphill -- but after that it had more typical hills, so I kept going. Then a longer downhill and longer uphill. And it took me to the Refuge du Creux de Lachat, which had a fine view of Mont Blanc and also a mountain (near Lac d'Annecy) called La Tournette (which I had once climbed up and skied down). I saw other people skating along the way. I had a snack at the refuge, then skated back to Refuge de la Plate and continued on to other trails in the center. They all seemed better than ever, so la Feclaz stands for me as one of the most fun cross-country trail networks in the world.

La Féclaz is in the Bauges mountains northeast from the city of Chambéry (which is roughly east from Lyon, south from Geneva + Annecy, west from Torino / Turin, north from Grenoble). Our normal method of driving to La Féclaz is by taking the A41 or A43 autoroute toward Chambery and taking the exit for "Massif des Bauges", then continuing to follow signs for "Massif des Bauges", roughly north thru the outskirts of the city (a little tricky at some points), and then the road climbs steeply with curves. It climbs thru St Jean d'Arvey, then just before Col de Planpalais the road curves away toward La Féclaz. Parking in several lots or on the roadside near the La Féclaz (forêt) access point where they sell tickets for use of the ski trails. Rental of skis + boots + poles at several shops in the village nearby.

see photos of la Féclaz

le Grand Coin: Sharon and I tried this cross-country area on a mid-week day. It has views of spectacular mountains: close to the Grand Coin, across the valley to the Chaine de Belledonne, and further to the southeast to the Aiguilles d'Arves. The environment closer to the trails was also nice. The trails tend to be of the up-up-up-down-down-down variety, while I usually prefer up-down-up-down-up-down for cross-country skiing.

le Grand Coin is near Montaimont (Bonvillard entry) and Montpascal (Col du Chaussy entry), which are north from La Chambre and St-Jean-de-Maurienne and south from Col de la Madeleine, all roughly east from Chambery and Lyon.

frequency of leg-pushes in V2 versus V1 skating

08feb7 posted to xcskiforum

somebody wrote that

It's strange that people doing high-frequency V1 skate are going faster than me doing V2 skate.

. . . I'm doing some "crunch", but maybe not quite a substantial butt drop.

But the problem is this: Your butt rising up and dropping down is the way that the big leg muscles contribute power to poling. How far up and down your butt goes ... how much your upper body weighs ... how many times per minute -- multiple those three together and convert to Watts and that's the power contribution of your legs - (provided your arms transmit that power to the poles).

Here's the leg-muscle game in V2: If your butt isn't moving up and down much, then the main Power contribution of your leg muscles is in the skate-pushing. So if you're doing V2 at 40 rpm, then each set of leg muscles is making its pushes at 40 rpm.

Suppose a bike racer told you he's got good cardio-vascular capacity, but he didn't know why he was getting passed by other riders. So you asked him, "What cadence frequency are you making your leg-pushes?", and he said "40 rpm".

I think you'd then say something like, "But your leg muscles cannot exploit your good cardio-vascular capacity if they're only pushing at 40 rpm. Everybody knows more like 80 rpm is better."

What the elite ski racers do is push with each leg's main muscles twice in each V2 stroke-cycle: First time to lift the mass of the upper body, then a second time to skate-push the ski out to the side. So even though their overall stroke cadence is 40 rpm, they're pushing each set of big leg muscles at a turnover frequency of 80 rpm.

So if you do a "smooth" "efficient" V2 with "well-balanced" glide at 40 rpm, with your big leg muscles making their pushes at 40 rpm, it's not a surprise that some "choppy" V1 guy who's pushing his big leg muscles at 55 rpm is going past you. (Just like if you were both on bicycles). But if you learn to use your legs to substantially drive your V2 pole-push, then you'll get 80 rpm pushes out of your biggest muscles. That's simply the known way to draw the highest sustainable aerobic power from them.

What I'm suggesting is the main reason pro-racer V2 is faster than pro-racer V1 is because it's pretty difficult to do V1 at 80 rpm on normal-length skis. Effective butt-rise-and-fall V2 is the "secret" way to achieve a sustainable 80 rpm on skis.

single-ski balance versus overlap in skating

08feb6 posted to xcskiforum

somebody wrote that

single-ski balance in Skating is roughly the same as for Classic striding

Yes it's important to be capable of balancing and gliding on one ski when skating. But unlike for Classic striding, it's not important to actually do it all the time.

Especially in V2, lots of elite racers have both skis on the snow for a portion of the stroke cycle. It's not unheard of to see a brief "overlap" of both skis on the snow in V1 either.

Elite Ice and inline speedskaters "overlap" the timing of both feet on the ground frequently. On inlines, I can move pretty fast without ever lifting either foot off the ground.

I think the main advantage in V2 skate of setting down the next foot while the previous foot is still pushing is that it allows you to support the pushing-side hip for a longer time in a lower position (without this support, the pushing hip would just fall into the ground if tried to extend its pushing time). The longer the hip is lower, the higher the percentage of the Force of the main extension push is aimed outward into immediate direct propulsive Work, instead of downward into the ground.

In V1 the same "support" point might still apply, but for a shorter time duration. Another advantage is that the two legs can work together to start raising the weight of the hips and upper body during the recovery-side push (instead of requiring the recovery-side leg to do a pure single-leg "squat" lift).

In V1 (and "free" skating with no poling) there's also the possibility of pushing outward with both legs at the same time -- so both legs are simultaneously (briefly) applying immediate direct propulsive work.

Bicycle pedaling, it's perfectly normal for elite racers to overlap direct propulsive pushing with both two legs at the same time. Why should skating be different?

For Classic striding, definitely you do want to be sure to delay setting down the next foot until after the current leg has finished pushing against the ground -- otherwise you can lose grip.

For Skating the physics implies that it's worth at least playing around with the opposite strategy: Try to set down the next ski early (and earlier) -- to increase turnover frequency, support a lower hip position, and possibly extracting simultaneous immediate propulsive work (or simultaneous lifting work) from both legs.

more . . .

see also


concept words: ski skiing cross country cross-country xc snow skier skiers skis nordic roberts report